Foundations of Pragmatics PDF

Foundations of Pragmatics
HoPs 1
Handbooks of Pragmatics
Editors
Wolfram Bublitz
Andreas H. Jucker
Klaus P. Schneider
Volume 1
De Gruyter Mouton
Edited by
Wolfram Bublitz
Neal R. Norrick
De Gruyter Mouton
ISBN 978-3-11-021425-3
e-ISBN 978-3-11-021426-0
Library of Congress Cataloging-in-Publication Data
Foundations of pragmatics / edited by Wolfram Bublitz, Neal R.
Norrick.
p. cm. (Handbook of pragmatics; 1)
Includes bibliographical references and index.
ISBN 978-3-11-021425-3 (alk. paper)
1. Pragmatics. I. Bublitz, Wolfram. II. Norrick, Neal R.
P99.4.P72F68 2011
4011.45dc22
2011013980
Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie;
detailed bibliographic data are available in the Internet at http://dnb.d-nb.de.
2011 Walter de Gruyter GmbH & Co. KG, Berlin/Boston
Cover image: Vetta Collection/iStockphoto
Typesetting: Drlemann Satz GmbH & Co. KG, Lemfrde
Printing: Hubert & Co. GmbH & Co. KG, Gttingen
Printed on acid-free paper
Printed in Germany
www.degruyter.com
Preface to the handbook series

Wolfram Bublitz, Andreas H. Jucker and Klaus P. Schneider
The series Handbooks of Pragmatics, which comprises nine self-contained volumes, provides a comprehensive overview of the entire field of pragmatics. It is
meant to reflect the substantial and wide-ranging significance of pragmatics as a
genuinely multi- and transdisciplinary field for nearly all areas of language description, and also to account for its remarkable and continuously rising popularity
in linguistics and adjoining disciplines.
All nine handbooks share the same wide understanding of pragmatics as the
scientific study of all aspects of linguistic behaviour. Its purview includes patterns
of linguistic actions, language functions, types of inferences, principles of communication, frames of knowledge, attitude and belief, as well as organisational
principles of text and discourse. Pragmatics deals with meaning-in-context, which
for analytical purposes can be viewed from different perspectives (that of the
speaker, the recipient, the analyst, etc.). It bridges the gap between the system side
of language and the use side, and relates both of them at the same time. Unlike syntax, semantics, sociolinguistics and other linguistic disciplines, pragmatics is defined by its point of view more than by its objects of investigation. The former precedes (actually creates) the latter. Researchers in pragmatics work in all areas of
linguistics (and beyond), but from a distinctive perspective that makes their work
pragmatic and leads to new findings and to reinterpretations of old findings. The
focal point of pragmatics (from the Greek prgma ,act) is linguistic action (and
inter-action): it is the hub around which all accounts in these handbooks revolve.
Despite its roots in philosophy, classical rhetorical tradition and stylistics, pragmatics is a relatively recent discipline within linguistics. C.S. Peirce and C. Morris
introduced pragmatics into semiotics early in the twentieth century. But it was not
until the late 1960s and early 1970s that linguists took note of the term and began
referring to performance phenomena and, subsequently, to ideas developed and advanced by Wittgenstein, Ryle, Austin and other ordinary language philosophers.
Since the ensuing pragmatic turn, pragmatics has developed more rapidly and diversely than any other linguistic discipline.
The series is characertised by two general objectives. Firstly, it sets out to reflect the field by presenting in-depth articles covering the central and multifarious
theories and methodological approaches as well as core concepts and topics characteristic of pragmatics as the analysis of language use in social contexts. All articles are both state of the art reviews and critical evaluations of their topic in the
light of recent developments. Secondly, while we accept its extraordinary complexity and diversity (which we consider a decided asset), we suggest a definite
structure, which gives coherence to the entire field of pragmatics and provides
vi
orientation to the user of these handbooks. The series specifically pursues the following aims:
it operates with a wide conception of pragmatics, dealing with approaches that
are traditional and contemporary, linguistic and philosophical, social and cultural, text- and context-based, as well as diachronic and synchronic;
it views pragmatics from both theoretical and applied perspectives;
it reflects the state of the art in a comprehensive and coherent way, providing a
systematic overview of past, present and possible future developments;
it describes theoretical paradigms, methodological accounts and a large
number and variety of topical areas comprehensively yet concisely;
it is organised in a principled fashion reflecting our understanding of the structure of the field, with entries appearing in conceptually related groups;
it serves as a comprehensive, reliable, authoritative guide to the central issues
in pragmatics;
it is internationally oriented, meeting the needs of the international pragmatic
community;
it is interdisciplinary, including pragmatically relevant entries from adjacent
fields such as philosophy, anthropology and sociology, neuroscience and psychology, semantics, grammar and discourse analysis;
it provides reliable orientational overviews useful both to students and more
advanced scholars and teachers.
The nine volumes are arranged according to the following principles. The first
three volumes are dedicated to the foundations of pragmatics with a focus on micro
and macro units: Foundations must be at the beginning (volume 1), followed by
the core concepts in pragmatics, speech actions (micro level in volume 2) and discourse (macro level in volume 3). The following three volumes provide cognitive
(volume 4), societal (volume 5) and interactional (volume 6) perspectives. The
remaining three volumes discuss variability from a cultural and contrastive (volume 7), a diachronic (volume 8) and a medial perspective (volume 9):
1. Foundations of pragmatics
Wolfram Bublitz and Neal R. Norrick
2. Pragmatics of speech actions
Marina Sbis and Ken Turner
3. Pragmatics of discourse
Klaus P. Schneider and Anne Barron
4. Cognitive pragmatics
Hans-Jrg Schmid
5. Pragmatics of society
Gisle Andersen and Karin Aijmer
6. Interpersonal pragmatics
Miriam A. Locher and Sage L. Graham
7. Pragmatics across languages and cultures
Anna Trosborg
8. Historical pragmatics
Andreas H. Jucker and Irma Taavitsainen
9. Pragmatics of computer-mediated communication
Susan Herring, Dieter Stein and Tuija Virtanen
vii
Acknowledgements
This initial, foundational volume in this series of handbooks represents the work of
a large number of individuals over a fairly long period of time. We must, of course,
thank our contributors for producing high quality articles reflecting scholarly rigor
and a strong sense of what our readers seek in a handbook of this scope. Without
conscientious, willing reviewers, projects like this are simply impossible. Our sincere thanks go out to all our reviewers, though we cannot identify them by name.
We editors are surrounded by competent, trustworthy staff in Augsburg and Saarbrcken, Germany. We own a debt of gratitude to Claudia Enzweiler for proofreading and Sylvia Monzon for her clerical services, to Katharina Rters, Katrin Stuis
and Claudia Rieger for their help with the desk editing and compilation of the indexes. Finally, we are grateful to the De Gruyter team: Barbara Karlson, Wolfgang
Konwitschny and Anke Beck.
Table of contents
Preface to the handbook series . . . . . . . . . . . . . . . . . . . . . . . .

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
v
ix
Introduction: the burgeoning field of pragmatics

Wolfram Bublitz and Neal R. Norrick . . . . . . . . . . . . . . . . . . . . .
Part I Conceptual foundations

1.
2.
3.
4.
Pragmatics as a linguistic concept

Anita Fetzer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
23
Micropragmatics and macropragmatics

Piotr Cap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
51
Pragmalinguistics and sociopragmatics

Sophia Marmaridou . . . . . . . . . . . . . . . . . . . . . . . . . .
77
Metapragmatics
Axel Hbler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Part II Theoretical foundations

5.
The rise of pragmatics: a historiographic overview

Wataru Koyama . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
6.
Semiotic foundations of pragmatics

Winfried Nth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
7.
Pragmatics in modern philosophy of language

Nikola Kompa and Georg Meggle . . . . . . . . . . . . . . . . . . . 203
8.
Foundations of pragmatics in functional linguistics

Saskia Daalder and Andreas Musolff . . . . . . . . . . . . . . . . . . 229
9.
Foundations: ethnomethodology and Erving Goffman

Christine Domke and Werner Holly . . . . . . . . . . . . . . . . . . 261
xii
Table of contents
10.
Pragmatics in Habermas Critical Social Theory

Maeve Cooke . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
289
Part III Key topics in pragmatic description

11.
12.
13.
14.
Deixis and indexicality

William F. Hanks . . . . . . . . . . . . . . . . . . . . . . . . . . . .
315
Reference and anaphora

Monika Schwarz-Friesel and Manfred Consten . . . . . . . . . . . .
347
Speech acts
Elena Collavin . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
373
Types of inference: entailment, presupposition, and implicature

Yan Huang . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
397
Part IV The place of pragmatics in the description of discourse

15.
16.
17.
18.
Pragmatics and grammar

Arnulf Deppermann . . . . . . . . . . . . . . . . . . . . . . . . . .
425
Pragmatics and semantics

John Saeed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
461
Pragmatics and prosody: prosody as social action

Elizabeth Couper-Kuhlen . . . . . . . . . . . . . . . . . . . . . . .
491
Pragmatics and literature

Jacob L. Mey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
511
Part V Methods and tools

19.
20.
Approaching the data of pragmatics

Monika Bednarek . . . . . . . . . . . . . . . . . . . . . . . . . . .
537
Experimental pragmatics
Richard Breheny . . . . . . . . . . . . . . . . . . . . . . . . . . . .
561
Table of contents
xiii
21.
Corpus-based pragmatics I: qualitative studies

Gisle Andersen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
22.
Corpus-based pragmatics II: quantitative studies

Christoph Rhlemann . . . . . . . . . . . . . . . . . . . . . . . . . . 629
23.
The transcription of face-to-face interaction

Roger J. Kreuz and Monica A. Riordan . . . . . . . . . . . . . . . . 657
About the authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 681

Name index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 691
Subject index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 706
xiv
Table of contents

1.
The aim of the handbook
The present volume, as the initial handbook in the nine volume series Handbooks
of Pragmatics, provides a comprehensive overview of the foundations of pragmatics. Its in-depth articles cover the roots and evolution of those central theories
and approaches as well as key concepts and topics that are characteristic of twenty-first century pragmatics as an approach to the means and ways of using language
in authentic social contexts. The articles provide reliable orientational overviews
useful to researchers, students, and teachers. They offer both state-of-the-art reviews of their topics and critical evaluations in the light of on-going developments.
Thus, topics are considered not only within their contemporary scholarly context
but are also critically evaluated from one or more current perspectives. As the
opening volume in the series, Foundations of Pragmatics provides historical, conceptual, theoretical and methodological vantage points from which the articles in
the following eight volumes can be related to each other and to the development of
the entire field.
The series as a whole seeks to be reasonably comprehensive to account for the
exceptionally vast, unusually heterogeneous and still rapidly expanding field of
pragmatics. At the same time, it sets out to give structure and coherence to the field
through its organization and choice of topics as well as by revealing traits and contours that have evolved during the past half-century. The nine volumes are meant to
reflect the substantial and wide-ranging significance of this transdisciplinary subject for nearly all areas of language description, to trace its origins and influences,
as well as to account for its remarkable and continuously rising popularity in linguistics and adjoining disciplines.
2.
The history of pragmatics
Even though its roots can be traced back to early classical traditions of rhetoric and
stylistics, to Immanuel Kants conception of pragmatics as empirical and purposive and to William James, who pointed out its practical nature, modern pragmatics
is a fairly recent discipline. Its inauguration as an independent field of study within
semiotics took place early in the 20th century by C. Morris, R. Carnap and ultimately C.S. Peirce. The classic division between syntax, semantics, and pragmatics goes back to Morris, who distinguished three separate dimensions of semiosis within his science of signs:

One may study the relations of signs to the objects to which the signs are applicable.
This relation will be called the semantical dimension of semiosis, symbolized by the
sign DSEM; the study of this dimension will be called semantics. Or the subject of study
may be the relation of signs to interpreters. This relation will be called the pragmatical
dimension of semiosis, symbolized by the sign DP; the study of this dimension will be
named pragmatics. One important relation of signs has not yet been introduced: the formal relations of signs to one another. [] This third dimension will be called the syntactical dimension of semiosis, symbolized by the sign DSYN, and the study of this dimension will be named syntactics. (Morris 1938: 2122)
Syntax studies the relations signs bear to other signs, semantics the relation between signs and objects, and pragmatics the relation between signs and their interpreters. Of course, there were and are differences of opinion on where exactly to
draw the line between semantics and pragmatics, as many of the chapters in this
volume will show.
Some thirty years elapsed before pragmatics finally made its way into modern
linguistics in the late 1960s, when linguists began to explore so-called performance phenomena. To this end, they adopted ideas developed and advanced by
L. Wittgenstein, G. Ryle, P. Strawson, J.L. Austin and other eminent (ordinary or
natural) language philosophers. It seems safe to claim that the ensuing pragmatic
turn was most notably induced by J.L. Austin, J.R. Searle and H.P. Grice, who
were interested in utterance meaning rather than sentence or word meaning, i.e. in
studying unique historical events created by actual speakers to perform linguistic
acts in actual situational contexts in order to accomplish specific goals. Other
scientific movements that nourished pragmatics include anthropology (B. Malinowski, P. Wegener, A. Gardiner), contextualism (J.R. Firth) and functionalism
(K. Bhler, R. Jakobson, D. Hymes), ethnomethodology (H. Garfinkel, E. Goffman, H. Sacks) and European sociology (J. Habermas). This volume addresses all
of these influences.
Since the pragmatic turn, pragmatics has arguably developed more rapidly and
diversely than any other linguistic discipline. Since the 1970s, the early AngloAmerican framework of pragmatic linguistic study has been immensely expanded
and enhanced by research in Continental Europe and elsewhere. With historiographic hindsight, it can be seen that the broadening, i.e. the interdisciplinary expansion, of the field of pragmatics has been a cumulative process; the broader conception of pragmatics chronologically (and causally) followed the narrower one.
We can easily detect a first shift from the binarity of early structuralist concepts
(such as lexical/word meaning versus causal/sentence meaning) to the multiplicity
of speech act related concepts. A further step towards a (conceptual as well as methodological) widening of the field took place in the 1970s and 1980s, when structure and action oriented pragmatics linked up with the emerging interactional paradigms in sociology (in general) and ethnomethodology (in particular). The ensuing
developments have seen inter alia a realignment of context (from a static and au-
tonomous to a dynamic and collaborative concept, which is cognitively, situationally and socio-culturally much more refined), and a shift from the concept of the
unilateral, i.e. the individual speakers (speech) act (as advocated by pioneers like
Austin) to the interactionally expanded concept of the bi- or (in some types of interactive computer-mediated forms of communication) multi-lateral inter-act. As
such, contemporary broad pragmatics takes account of the interactional turn
that can be observed in the most recent development of interactive (Web 2.0-based)
media formats.
Taken all in all, the establishment and development of linguistic pragmatics has
been an authentic success story.
3.
The definition of pragmatics
Despite its scientific acclaim, the notion of pragmatics remains somewhat enigmatic and is still difficult to define. This holds for its readings in everyday discourse as well as in scholarly contexts. Nonetheless, when we refer to attitudes and
modes of behavior as pragmatic, we mean that they have a factual kind of orientation in common. People who act pragmatically or take a pragmatic perspective
generally have a preference for a practical, matter of fact and realistic rather than a
theoretical, speculative and idealistic way of approaching imminent problems and
handling everyday affairs. To put it differently, they share a concrete, situation-dependent approach geared to action and usage rather than an abstract, situation-independent and system-related point of view. To assume a pragmatic stance in
everyday social encounters as well as in political, historical and related kinds of
discourse, means to handle the related affairs in a goal-directed and object-directed, common-sense and down to earth kind of way. Such an understanding of
pragmatics as an attitude in non-scientific discourse has obviously left its traces in
scientific definitions of the term. By and large we can say that in semiotics and
philosophy, pragmatic characterizes those theoretical and methodological approaches that are oriented toward use and context rather than toward some system,
and that they regard use and context as creating a high degree of analytical surplus.
While essentially the same is true for linguistics in general, there is no commonly accepted definition of pragmatics in linguistics which would refer to a
single, unified and homogeneous field of study. In contemporary linguistics, we
can identify a narrow and a broad way of delineating pragmatics (of which the
former is sometimes allocated to an Anglo-American and the latter to a Continental [European] tradition of pragmatics, cf. Huang 2007: xi). According to the
narrow view, pragmatics is understood as the systematic investigation of what and
how people mean when they use language as a vehicle of action in a particular context and with a particular goal in mind. Thus, the context-dependency of utterance
meaning is the central component of more narrowly defined accounts of prag-
matics, which focus on a few key issues that can be juxtaposed with related issues
in other modules of language theory such as grammar and semantics. Those issues
include indexicality/deixis (versus anaphora), presuppositions, implicatures (versus entailments) and speech acts (versus types of sentences), to name only the most
conspicuous topics.
In this volume, as indeed in the whole Handbooks of Pragmatics series, we
adopt a much broader point of view and understand pragmatics as the scientific
study of all aspects of linguistic behavior. In particular,
pragmatics includes patterns of linguistic actions, language functions, types of inferences, principles of communication, frames of knowledge, attitude and belief, as well as
organisational principles of text and discourse. Pragmatics deals with meaning-in-context, which for analytical purposes can be viewed from different perspectives (that of
the speaker, the recipient, the analyst, etc.). It bridges the gap between the system side of
language and the use side, and relates both of them at the same time. Unlike syntax, semantics, sociolinguistics and other linguistic disciplines, pragmatics is defined by its
point of view more than by its objects of investigation. The former precedes (actually
creates) the latter. Researchers in pragmatics work in all areas of linguistics (and
beyond), but from a distinctive perspective that makes their work pragmatic and leads
to new findings and to reinterpretations of old findings. The focal point of pragmatics
(from the Greek prgma act) is linguistic action (and inter-action); it is the hub around
which all accounts in these handbooks revolve. (Preface to the handbook series)
Pragmatics is fundamentally concerned with communicative action in any kind of

context. The multifaceted research paradigm of pragmatics has provided new directions and perspectives in the arts and humanities, philosophy, cognitive science,
computer science and the social sciences. Pragmatic perspectives have been employed in information technology and in the social sciences, particularly in economics, politics and education.
In the pragmatic perspective, language use and language users in interaction
are primary, as opposed to language as a system of signs or a set of rules. The pragmatic perspective scrutinizes neither just individual words nor sentences nor even
isolated texts, but rather whole speech events or language games in real social contexts, considering both the present state of affairs and its connectedness with prior
and succeeding actions. It rejects a localization of language in a limited segment of
the acts of speaking, understanding and responding or within the conscious of the
individual. It supplants a view of language as an abstraction without variation by
speaker, region or time, of language as a non-cultural, non-social, static, depersonalized fact independent of context and discourse. Pragmatics goes beyond the perspective of written texts with their carefully marshalled grammatical sentences to
embrace everyday talk and the messiness of language in real embodied human
contexts, where participants with personalities, feelings and goals interact in complex ways with physical objects and other participants within institutions and communities.
That pragmatics overcomes any narrow focus on language as a system of signs

or a set of rules does not mean, of course, that pragmatics (in contradistinction to
other scientific disciplines) attempts to describe language without recourse to the
systematic level of theoretical abstraction. Ultimately and in order to create and secure a sound scientific foundation, pragmatics, like any other scientific theory, has
to accomplish the transfer from the level of token to the level of type. For the pragmatic scholar, specific concrete linguistic events (and their contexts) are only relevant in as far as their properties and constraints can be integrated under analogous
conditions into a general concept of language and communication. While pragmatics is on a par with other scientific theories in this respect, it transcends them in
various ways, in particular by adopting a constructivist point of view which allows
for the inclusion of new contextual, situational and cognitive variables. Describing
language pragmatically thus means going beyond the description of language as an
autonomous, type-restricted principle and taking into account extra-linguistic phenomena and conditions emanating from the context and concrete situation of language use. Turning contextual and related conditions into prototypical conditions
(and thus theorizing them) constitutes the pragmatic surplus.
4.
The purview of pragmatics
If pragmatics is more a perspective or orientation toward language use than a particular theory or specific research object, then the data practitioners choose to analyze assume great importance. In linguistic research, as Saussure realized, it is the
viewpoint that determines the object rather than the other way around. In the tradition growing out of Natural Language Philosophy the intuitions and casual observations of the individual writer provide the stuff for analysis. Introspection counts
as data. All the early work on speech acts, presupposition and inference was based
on introspection and invented examples in imagined contexts (Austin, Searle,
Strawson and Grice).
Other traditional areas of pragmatic research were based on authentic data from
the start. Indexicality/deixis, discourse markers, particles and the like are ubiquitous in texts, and examples for study are thus easy to collect. As corpora of spoken
language became increasingly available and reliable, the problem of evidence versus introspection began to take care of itself. Scholars no longer needed to guess
about distributions and frequencies of linguistic items. One might begin with a
small pre-selected corpus representing a specific set of items or types of interaction
to identify a range of functions for a particular linguistic feature before embarking
on a general search in a large corpus to determine distributions and frequencies in
the corpus at large only to return to a small corpus representing specific contexts
for careful qualitative analysis. By contrast, investigations of large corpora might
provide the impetus for research, illustrating a range of items or contexts not found
in a small corpus and suggesting hypotheses to be tested. For statistical purposes,

larger was presumably always better, because scholars tended to correlate frequency with typicality (the more frequent the more typical). On the other hand, increasing interest in multi-modal data and description in research on spoken interaction brought researchers back to small, carefully annotated corpora. By its very
nature, research on prosody, gaze, gesture and the details of interaction with objects in a specific physical setting must proceed from a narrow recording and transcript of a particular event: Large corpora are extremely difficult to manage in
these circumstances.
Many pragmaticians maintain that language data must be as authentic as possible, recorded in real life contexts where something is at stake for the participants.
But the whole matter of recording, often with camcorders as well as microphones,
raises the spectre of the Observers Paradox. How natural is the language interaction, when we require technical apparatus to record it? Some pragmatic scholars
ascribe to the notion that language use must be understood in terms of embedded
practices, as part of the social world in which listening and speaking are modalities
of action, in line with the treatment of speech production and reception as embodied social facts (Hanks 1996), and practices as shared ways of doing things, ways
of talking, beliefs and values as a function of engagement in an activity (Wenger
1998; Eckert 2000). Conceptualizing speaking and listening in terms of practices
affords a natural purchase on nonverbal features of participant behavior, including
gaze, posture and gesture in face-to-face interaction (Goodwin 2003). This has
consequences for what counts as data, how much context must be considered, what
must be transcribed and in how much detail.
At the same time, certain questions will always go unanswered on the basis of
corpus data, no matter how careful the transcription and no matter how large the
corpus, since every individual exchange is participant designed and context dependent. With some kinds of structures like prosody, discourse markers and interjections, there is the problem of how to get enough examples to make sensible generalizations. Experimental procedures could complement corpus-based research in
order to generate parallel structures and contexts to test for recurrent speaker and
listener practices. Indeed, Experimental Pragmatics represents a new area of research at the intersection of traditional linguistic pragmatics and psycholinguistics. It seeks to apply experimental techniques from psycholinguistics to solve
problems defined in linguistic pragmatics (see Breheny this volume). So far such
experiments have focused on issues surrounding the conversational maxims and
implicatures of Grice (1975) and his followers (both Neo-Griceans like Horn 1984
and Levinson 2000, and Post-Griceans like Sperber and Wilson 1986 and Carston
2002), but creative experiment design could extend this new research paradigm to
all sorts of questions about spoken language. All the same, investigations of large
corpora provide a body of natural data against which experimental results can be
evaluated.
The discussions between scholars of different persuasions about what constitutes appropriate data are ongoing. The historical development of different data
sources and areas of inquiry as well as the continuing debates about their usability
and validity resonate in this handbook. Indeed, the authors contributing to this
handbook were chosen to represent various positions in these debates.
5.
Organization of this Handbook
To distinguish pragmatics as a linguistic discipline from pragmatics as an attitude or

a mode of behavior (a practical rather than an idealistic approach to problems and
dealings) as used in everyday, in political, in economic and in other kinds of discourse, the present handbook Foundations of Pragmatics opens with a section on
Conceptual Foundations consisting of the description of pragmatics as a linguistic
concept. Fundamental concepts covered are: micropragmatics, i.e. the pragmatics of
utterance based concepts such as speech acts etc. (which are covered in more detail
in vol. 2 of the series) as opposed to macropragmatics, i.e. the pragmatics of discourse or text based concepts such as topics, discourse markers etc. (cf. a more detailed overview in vol. 3 of the series), pragmalinguistics (dealing with forms) as opposed to sociopragmatics (dealing with pragmatic strategies), and metapragmatics.
The second major section is dedicated to the Theoretical Foundations of pragmatics. It begins with a historiographic overview of how pragmatics was established as a linguistic field of study, how it developed and spread, and how it eventually became an international success story. After all, since the so-called
pragmatic turn, linguistic pragmatics has, arguably, developed more rapidly and
diversely than any other linguistic field. The scientific disciplines that nourished
pragmatics in the first place range from semiotics (Peirce, Morris, Carnap) and
(Natural Language) philosophy (Wittgenstein, Strawson, Austin, Searle, Grice) to
anthropology (Malinowski, Wegener, Gardiner), contextualism (Firth) and functionalism (Bhler, Jakobson, Hymes), from ethnomethodology (Garfinkel, Goffman, Sacks) to European sociology (Habermas).
The third major section, Key Topics in Pragmatic Description, investigates
those central concepts of pragmatic description which were the first to make their
way from grammar into pragmatics (indexicality and deixis, followed by reference
and anaphora), on the one hand, and those notions that rapidly became the focal descriptive construct in mainstream pragmatics (speech acts and types of inference
such as entailment, presupposition, implicature), on the other.
In the fourth major section, the Place of Pragmatics in the Description of Discourse, the scope of pragmatics within linguistic description is delimited vis--vis
grammar, semantics, prosody. A final chapter on pragmatics and literature underscores the contention that the social aspect of language use is as prominent in literary contexts as in any non-literary context.
Methods and Tools constitutes the fifth and final major section. It contains an
initial chapter comparing various methodological approaches to pragmatics such
as intuition, introspection, ethnographic field work, elicitation and corpus analysis,
before it then proceeds with articles on experimental pragmatics, corpus-based
pragmatics and the transcription of face-to-face interaction.
No handbook can include every individual perspective and specific topic even
regarding the foundations of an area of study. Our decisions regarding which fields
and issues to include and how they should be addressed reflect both our own views
of pragmatics and the expertise and interests of our authors. As authors were recruited for the handbook project, joined us, and began to think and write about their
areas, topics were redefined, expanded and narrowed, titles were adjusted, some
boundaries were sharpened, others redrawn. Recent developments in pragmatics,
the general availability of new kinds of multi-media data and analytic procedures
along with dynamically evolving attitudes toward topics and research methods
helped shape the contents of this volume. Inevitably, the exigencies of scheduling
and the special interests of the contributing authors conspire to yield a final set of
articles representing individual views at a specific point in time. We embrace this
particular character of the snapshot offered by the articles in this handbook.
6.
In the lead article, Pragmatics as a linguistic concept, Anita Fetzer seeks to limit
the concept of linguistic pragmatics. The chapter compares pragmatics as a linguistic concept with pragmatics as used in everyday, in political, in economic and
other kinds of discourse. It discusses what is special about linguistic pragmatics,
covering General Pragmatics as well. Pragmatics as a common-sense notion is
functionally synonymous with practical or just right at that stage, describing a
language-bound and/or action-based phenomenon within a particular situation, as
is reflected in its usage in everyday, in political, in economic and other kinds of discourse. In linguistics, semiotics or philosophy, pragmatics refers to a theoretical
concept, comprising contextual, generally usage-related constraints and requirements, for instance presuppositions, indexical expressions, felicity conditions,
common ground and background assumptions. These are necessary conditions for
communication to be felicitous in context. Thus, the perspective is shifted from the
analysis of the language system and its constitutive parts to that of its rule-governed instantiation and embeddedness in context. In General Pragmatics the focus
of investigation is on universals of communicative action, such as directness versus indirectness. These are refined in Linguistic Pragmatics, where they undergo
language- and culture-specific modification. That change in perspective has important consequences for methodology. To account for the parts-whole connectedness, a relational frame of reference is required which accommodates the relation
between the language system and language use, and interlocutors and what they do
with language in context, comprising linguistic context (or co-text), social context
and cognitive context.
In the second article, Micropragmatics and macropragmatics, Piotr Cap addresses the central distinction between micropragmatics and macropragmatics.
This chapter distinguishes micropragmatics as referring to speech acts/actions etc.
from macropragmatics as directed at the discourse/text level. Defining pragmatics
by its general perspective, rather than by its specific object of investigation, entails
looking at the analytic scope of pragmatics in terms of micro- and macro-level concepts. These have been traditionally (cf. Mey 1993) associated with the opposition
between the analysis of speech act force at an utterance level and the analysis of
global intentionality at the level of a discourse/text. Following this division, as
well as its later refinements, the article, first, reviews the locutionary, illocutionary
and perlocutionary constituents of force of an utterance, relating them to pragmatic
concepts which characterize the process of encoding and decoding its message
(deixis, presupposition, implicature etc.). Second, it demonstrates how individual
utterances comprising different sets of constituents and markers of force, including
syntactic markers of explicitness and inferable carriers of implicitness, can form
sequences to shape the global intentionality of a discourse/text. Invoking such notions as speech event and macro speech act, the article shows which pragmatic concepts utilized in (micro-) analysis of individual utterances are essentially complementary, i.e. which markers of force denoted by these concepts can collectively
generate complex macro illocutions responsible for accomplishing global
discourse/text goals.
The important distinction between Pragmalinguistics and sociopragmatics
forms the topic of chapter 3 by Sophia Marmaridou. In Principles of Pragmatics
(1983), Leech introduced a distinction between pragmalinguistics (roughly, what
form to use to achieve an intended pragmatic effect) and sociopragmatics (roughly,
when to use a particular pragmatic strategy). This chapter describes past and current research on this distinction in pragmatics. Within a broad definition of pragmatics as the scientific study of all aspects of linguistic behavior, the distinction
between pragmalinguistics and sociopragmatics is intended to focus on two methodological approaches. Pragmalinguistics concerns the study of the particular resources that a given language provides for conveying pragmatic meaning (illocutionary and interpersonal), whereas sociopragmatics, as the sociological interface
of pragmatics (Leech 1983: 10), relates pragmatic meaning to an assessment of
participants social distance, the language communitys social rules and appropriateness norms, discourse practices, and accepted behaviors (Thomas 1995). The
above methodological distinction to some extent relates to earlier work in Marxist
pragmatics (Mey 1979; Prucha 1983), foreshadows work in Critical Discourse
Analysis (Fairclough 1995), and intersects with studies in sociolinguistics. A critical assessment of the two concepts and the respective methodologies emerges from
10
current research practices. Even though the borderline between the two concepts
often seems fuzzy, their relevance to language learning, teaching and testing has
been enhanced (McNamara and Roever 2006) and so has its significance in historical corpus linguistics (Archer and Culpeper 2003).
In the fourth chapter, Axel Hbler addresses the subject of Metapragmatics.
He focuses on metapragmatics concerning the actual management of discourse.
Key tasks in metapragmatics are the investigation of that area of speakers competence which reflects the judgments of appropriateness of their own and other
peoples communicative behavior and the exploration of the know how regarding the control and planning of, as well as feedback on, the ongoing interaction.
The human capacity for reflecting about language and language use/communication is a feature of language-related competence that no animal is endowed with.
Part of this capacity extends to areas which linguists have become accustomed to
referring to as metapragmatics. The term is fairly new, and its usage not yet fully
regulated; it accommodates various systematic perspectives and pertains to a wide
range of language-reflexive phenomena. The article provides a comprehensive survey along two lines: (1) what linguists (and experts from other disciplines such as
philosophy or social-psychology) have observed about how people act and interact
and what they express about their practice; and (2) what these experts have observed about their own descriptive, analytical and theorizing practice. Line (1) includes theories of pragmatics as prefiguring observation and resulting from it; and
it takes account of what could be categorized as (implicit or explicit) meta-practice
by members of a speech community (e.g., monitoring, remedying, evaluating by,
e.g., quoting, topicalizing, metaphorizing). Line (2) comprises meta-theoretical
considerations, which (critically) reflect upon and compare varying theoretical
stances or elaborate on the difference to be drawn between a scientific view on
pragmatics and folk theories.
7.
The opening chapter 5 in this section, The rise of pragmatics: a historiographic

overview, by Wataru Koyama traces the history of the still young field of linguistic pragmatics. This chapter stresses the success story as well as the newness of
the field of pragmatics. Especially after the artificial impoverishment of linguistics
due to structuralism and generative grammar, there was clearly a need for pragmatics as a field to amalgamate insights from rhetoric, anthropology, functionalism
etc. When and by whom was linguistic pragmatics invented? Who were its
(extra-linguistic) predecessors? What was the nature of the scientific and, particularly, the linguistic environment, in which pragmatics could arise (the 1950s; logical positivism etc.). Then the author considers further developments: how and why
pragmatics developed, slowly at first, abruptly later in the 1980s (Levinson 1983,
11
International Pragmatics Association, Journal of Pragmatics etc.). Finally, he

looks at how todays pragmatics differs from earlier stages of pragmatics.
In the sixth article, Semiotic foundations of pragmatics, Winfried Nth relates pragmatics to its foundations in semiotics. The chapter shows how semiotics
influenced the development of the theoretical foundations of pragmatics, focusing
on the contributions of Peirce, Morris and Carnap. Semiotics, the general study of
signs and sign processes, is fundamental to pragmatics, the study of language use,
speech acts, and communicative processes, and it extends the framework of language use to include nonverbal, visual, and other signs in the context of verbal behavior. According to the general framework outlined by Carnap and Morris, pragmatics is one of the three branches of semiotics besides syntax and semantics, but
twentieth century structuralist semiotics tended to neglect the pragmatic dimension
in its study of language and discourse. Central issues of pragmatics, such as the
questions concerning intention and purpose, meaning and reference, word and object, interpretation and communication, or the relationship between language and
action, find fundamental answers in the writings of the founder of general semiotics and philosophical pragmatism (or pragmaticism), Charles S. Peirce.
Chapter 7, Pragmatics in modern philosophy of language by Nikola Kompa
and Georg Meggle, traces the origins of linguistic pragmatics in the Philosophy of
Language, particularly in Natural Language Philosophy in the work of Wittgenstein, Strawson, Austin, Searle, Grice. There are two paradigms of Pragmatics in
modern studies in the Philosophy of Language. In the first, semantics is taken as
given and the aim is to incorporate as many context-sensitive aspects of language
as possible into that semantics; in the second, research is trying to spell out the semantics on a pragmatical (action-theoretical) basis itself. Accordingly, the difference between the two paradigms could be considered that between semantic pragmatics on the one side and pragmatic semantics on the other side. Characteristic
contributions to the first approach are those by Frege, Carnap, Kaplan and Stalnaker; characteristic contributions to the second are by Wittgenstein, Austin,
Grice, Lewis and Bennett. This article considers the basic problems as well as the
present state of the art in both paradigms of pragmatics and attempts to achieve
some clarity about their deeper (dis-) connection in particular
In chapter 8, Foundations of pragmatics in functional linguistics, Saskia
Daalder and Andreas Musolff trace the roots of pragmatics in functional approaches to linguistics. Pragmatics addresses the functions of language as found in
the work of Bhler, Jakobson, Hymes. This chapter traces the historical connections from these sources to contemporary thinking in pragmatics. The study of linguistic pragmatics received a major impetus from the systematic analysis of language functions as started by Karl Bhler, who distinguished three fundamental
dimensions of the linguistic sign: it functions at the same time as symbol, symptom
and signal. The model has proved its enduring value in the development of functionalist language theories, via its modification and augmentation into a six-func-
12
tions-model by Roman Jakobson. His model in turn informed subsequent functionally oriented theories of language and communication, e.g. Dell Hymes
Ethnography of Speaking (1962) and Michael A.K. Hallidays concept of Language as Social Semiotic (1978). In retrospect, Bhlers model (and, to a lesser extent, Jakobsons modified model) can be judged to suffer from its forced integration with the structuralist concept of the linguistic sign, which initially
hampered a fully fledged pragmatic interpretation.
In their chapter 9, Foundations: Ethnomethodology and Erving Goffman,
Christine Domke and Werner Holly discuss two related questions: What is Ethnomethodology or rather what are the crucial questions and assumptions of this
qualitative research approach? And how did Ethnomethodology impact linguistic
pragmatics? At first they overview the studies of Harold Garfinkel which contain
the groundbreaking thoughts about the most significant question: How is social
reality constructed and organized by activities of the societys members? Accordingly, it is shown how conversation analysts (e.g. Sacks, Schegloff) are applying
Garfinkels fundamental assumption of the social reality accomplished through
peoples everyday life activities in analyzing talk-in-interaction. Subsequently, the
main focus lies on the interface between social order, which Ethnomethodology
considers as produced by people interacting in their daily life, and the fundamental
pragmatic question what people are doing using speech and talking. How the idea
of (re-)produced social order is in line with and can be linked to the pragmatic
focus on the role of language/verbal interaction in different forms of daily interaction.
The work of Habermas has been central in the development of pragmatics in
Europe. In her contribution in chapter 10, Pragmatics in Habermas Critical Social Theory, Maeve Cooke considers the influence of Habermas on linguistic
pragmatics. She deals with the pragmatic turn that Jrgen Habermas introduced to
critical social theory with his program of universal pragmatics in the 1970s. With
this program, subsequently referred to as formal pragmatics, Habermas investigated everyday linguistic interaction from the point of view of its orientation towards validity. His investigations were guided by the critical intentions of his social theory. His principal intentions were to provide a normative basis for a conception of practical rationality without reference to otherworldly authorities or
entities and to develop a corresponding normative account of social action. However, his formal pragmatics has profound implications for social theory in general
and, indeed, for anyone concerned with questions of meaning and validity. This article sets out the key elements of Habermas formal pragmatics, shows its implications for social theory and beyond, and discusses some important objections that
can be raised against him.
8.
13
Pragmatics first made inroads into linguistics through the pioneering work of Yehoshua Bar-Hillel. In this chapter 11, William Hanks describes past and current research on Deixis and indexicality from a pragmatic perspective. He starts from a
historical sketch of indexicality in the semiotics of Peirce and Morris, and traces
the development of the concept by philosophers, linguists, linguistic anthropologists, sociolinguists (via style, register, contextualization) and conversation
analysts. Called shifters by Jespersen and Jakobson, deictics encode a relational
structure projecting from an indexical ground (a context of utterance) to a referential figure (the object denoted) via a relation (proximity, perceptibility, anaphora,
etc.). Deictic systems differ cross-linguistically in ways just beginning to emerge.
Pronouns are a related type of shifter and are closely tied into participation frameworks as treated by Goffman, conversation analysts and linguistic anthropologists. Recent debate is summarized regarding the role of spatial proximity vs.
cognitive-social accessibility as a basis for deixis. The final section outlines empirical approaches to pragmatic research on deixis.
Reference and anaphora are inherently pragmatic in nature and their treatment
in linguistic description helped bring pragmatic notions into prominence.
Chapter 12 on Reference and anaphora by Monika Schwarz-Friesel and Manfred
Consten concerns reference, the area in scientific research which deals with the
ancient problem of how language relates to the world. In the tradition of philosophical logics and formal semantics, reference has been defined as a static relationship between expressions and the things they denote. In the course of pragmatic change in linguistics, models of reference have arisen that (to different
extents) take cognitive properties of language users into account. In the framework
of the latter notion, anaphora is a specific kind of reference that has to be resolved
mainly through the activation of discourse based knowledge established by the
preceding text. Thus, anaphora resolution is analysed as a process of reactivating
pre-mentioned referents. The traditional view of anaphora as a means of referential
continuity and the availability of given information has to be revised. From a procedural perspective, anaphora may add new information to the mental files of the
referents in the text-world model.
Elena Collavin addresses the key topic of Speech acts in chapter 13. The
chapter provides an overview of the notions of act/action (versus behavior) and
speech act: how they developed, how they made their way into linguistics, how
they became the core notions of various (speech act) theories, how the latter developed etc. Collavin presents an account of the notion of speech act, which includes
an illustration of the main formulations of the theory, its limitations, and its current
relevance and application in the broad field of pragmatic research. She traces the
origins of this notion and describes the main formulations of Speech Act Theory,
focusing on conventionality, intentionality, and action as the key aspects of speech
14
acts around which similarities and incompatibilities in different versions of the theory are established. The discussion then turns to the criticisms moved to various
notions of speech act. The relevance of speech acts for the empirical analysis of
language use in interaction has been strongly debated. Collavin discusses how
some notion of speech act can relate to conceptualizations nowadays fundamental
in the study of language usage such as frame, voicing, indexicality, multimodality,
shared cognition and embodiment.
Chapter 14 provides an overview of the varieties of inference recognized in
pragmatics, how they developed, how they made their way into linguistics, how
they became core notions of various theories, how the latter developed etc. In this
chapter, Types of inference: entailment, presupposition, and implicature, Yan
Huang provides a critical survey of three fundamental types of inference recognized in semantics and pragmatics, namely entailment, presupposition and implicature, covering their origin, development, and the roles they play in various semantic/pragmatic theories in linguistics and the philosophy of language. Of the three
categories of inference, entailment is essentially semantic in nature, though the notion of pragmatic entailment is discussed. Presupposition straddles the semanticspragmatics boundary, resulting in semantic and pragmatic presupposition. Finally,
with regard to the two subtypes of implicature, while conversational implicature is
pragmatic in nature, conventional implicature can be considered either as semantic
or pragmatic, depending on how the distinction between semantics and pragmatics
is drawn. Notions parallel to conversational implicature (such as explicature) and
the argument that conversational implicature should not be treated as inference are
also considered. Finally, comparisons between entailment and presupposition on
the one hand, and between presupposition and implicature on the other, are presented.
9.
Part IV opens with chapter 15 Pragmatics and grammar by Arnulf Deppermann.

This chapter details the delineation of pragmatics and grammar and the integration
of pragmatics and grammar into linguistic theory in relation to each other, focusing
on issues, methods and areas covered. Deppermann discusses theoretical approaches, linguistic phenomena and empirical methods concerning the relationship
between grammar and pragmatics. He sketches functional approaches to grammar,
which claim that grammar is shaped and constrained by cognition, text and/or social interaction (functional grammar, systemic-functional linguistics, construction
grammar, cognitive grammar, interactional linguistics). Historical studies on the
paths of grammaticalization, psycholinguistic insights from usage-based approaches, and conversation analysis are introduced as methods for the study of
the relationship between grammar and pragmatics. The relevance of grammar for
15
pragmatics is discussed concerning a broad range of issues, such as cognitive processes (implicature, inference, information structure), the expression of cognitive
states (epistemic stance), conversational organization (repair, turn-construction
and turn-taking, collaborative action, intersubjective grounding), action structure
(speech acts, genres), bodily action (gesture) and socio-structural concerns (politeness, power, social styles).
In chapter 16 on Pragmatics and semantics, John Saeed discusses some contrasting views of the relationship between semantics and pragmatics. His chapter
details the delineation of pragmatics and semantics in linguistic theory, focusing
on issues, methods, areas covered. Within the philosophical and formal traditions,
he considers the growth of contextualism, that is the increasing emphasis on the
underspecification of meaning in linguistic form, illustrated by phenomena beyond
traditional deixis (indexicals) such as gradable and comparative adjectives, quantifiers, possessives etc. The article also looks at the defense against this from semantic minimalism (e.g. Cappelen and Lepore 2005). As an example of contextualism, Saeed discusses aspects of Relevance Theory (Sperber and Wilson 1986,
Carston 2002), with its identification of pragmatic principles reaching deep into
what earlier accounts viewed as semantic content. In particular, he considers the
theorys notions of explicatures and higher level explicatures, and looks at its view
of lexical pragmatics. Finally, he brings into the discussion the quite different view
from cognitive linguistics, where the very distinction between semantics and pragmatics is questioned. So in theories like Fauconniers, mechanisms of mental
spaces and conceptual blending seek to model phenomena from across the traditional divide, including e.g. tense, modality, presupposition and metaphor. An interesting question is whether a pragmatic theory like Relevance Theory and the
cognitive semantics of writers like Lakoff, Fauconnier and Langacker are in fact
compatible.
Elisabeth Couper-Kuhlen describes the role of prosody in pragmatics, and the
integration of prosody into linguistic theory. In chapter 17 Pragmatics and prosody, she contemplates what has been achieved so far in the field of prosody and
interaction and what still remains to be done. Instead of terms such as suprasegmentals, sentence types and illocutions, she speaks of prosodic resources for
single- and multi-turn construction, action formation and floor management. Comparing the use of prosodic resources across turns-at-talk, she acknowledges the
contribution of prosody to marking the continuation of same-speaker turns, to organizing sequences of turns, to cueing problems, negotiating agreement and signaling affiliation in talk. But recent developments also present challenges to researchers in the field of prosody and interaction: (i) What counts as prosodic? and
(ii) What is prosodys place in the analysis of multimodal interaction? The chapter
discusses these issues and ways of dealing with them in the future.
In chapter 18, Jacob Mey discusses Pragmatics and literature. According to
Mey (2000), pragmatics is concerned with the way humans use language in a social
16
context: When people use language in a literary context, the social aspect is no less
prominent. Literary pragmatics, therefore, concerns itself with the relation between the language users in the context of literature, that is, the producers (the
authors) and the consumers (the readers). In the context of literary production,
the processes of producing and consuming are paramount. The product (the written
text) can only be evaluated properly in terms of its social conditions and the concomitant roles of (authorial) production and (readerly) consumption. Here, the key
concept of voice (originally due to Bakhtin) plays a major role. The author assumes a narrative voice, while retaining a distinct personal voice. Conversely, the
readers have a voice, inasmuch as they identify with, and co-construct, the narrative personae; in this way, they take part in the building of the literary universe
and making it come alive. Some implications of this pragmatic perspective are followed up through analysis of snippets of literary texts.
10.
The initial chapter 19 in this final part by Monika Bednarek, Approaching the data
of pragmatics, provides an overview of how various aspects of language use
are studied using different methodological approaches: intuition, introspection,
ethnographic field work, elicitation (and experimentation), and corpus analysis.
It compares critically the various methodologies that are employed in areas such
as conversation analysis, politeness theory, and speech act theory as well as in related functional approaches. The chapter covers both quantitative and qualitative
methodologies ranging from intuition and introspection to interviews, discourse
completion tasks, ethnographic field work, elicited data and corpus analysis. It
takes a broad approach to the data of pragmatics in that it not only includes research in key pragmatic areas but also disciplines related to pragmatics such as ethnography and sociolinguistics. It describes different methodologies and compares
the insights gained from them as well as the different perspectives they provide on
linguistic data.
In chapter 20, Experimental pragmatics, Richard Breheny surveys recent research in this emerging field. Experimental pragmatics aims to use the methods of
experimental psychology to test claims found in the more theoretical linguistics literature. The strategy of testing theoretical ideas experimentally has been around
for a while and is notably exemplified in the work of Clark, Sanford, Gibbs and
Glucksberg. More recently, there has been an upturn in the interest in experimental
methods, coinciding with renewed interest in pragmatic questions such as the nature of generalized implicatures, presupposition, metaphor and the like. Breheny
reviews studies involving children, adolescents and adults from clinical populations as well as typically developing groups using the full range of methods to be
found in psycholinguistics and developmental psychology (including both beha-
17
vioral and neuro-scientific methodologies). The contribution also contains a discussion of the difficulty of drawing empirically testable predictions from claims
made in the theoretical literature.
The final three chapters in the volume all address the role and forms of corpus
data in linguistic pragmatics. The first two consider corpus-based research in
pragmatics, first qualitative then quantitative studies, and the third discusses
methods of transcription. In chapter 21, Corpus-based pragmatics I: qualitative
studies, Gisle Andersen describes the practice and potentials of corpus-based research in pragmatics, focusing on qualitative approaches to language data. He
considers the development of spoken and written corpora like the Brown Corpus,
the London-Lund Corpus, the Helsinki Corpus, the British National Corpus, the
Santa Barbara Corpus of Spoken American English etc., and their growing influence on research in pragmatics over the last thirty years, paying attention to
special corpora, large general corpora, continually growing monitor corpora and
web-based corpora. Corpora allow for in-depth studies of patterns of frequency,
distribution and collocation, thus facilitating statistically based observations of
innovative language use and variation. Dynamic and ever-growing corpora allow
for the observation of emerging repetitive patterns such as the routinization of
forms with discourse marking functions, ongoing grammaticalization and the development of new abstract and interactional meanings from old forms. In contrast
to quantitative, statistical approaches to corpus data, qualitative studies see corpora primarily as a source of natural data and a way of overcoming the fallible introspection of armchair linguistics. In the area of qualitative pragmatic studies,
corpora have provided a constantly growing pool of data, allowing researchers to
test their hypotheses about individual items and constructions, as well as practices
and usage within particular groups, but also suggesting new avenues of research.
Diachronically oriented corpora are responsible for the rapid progress made
in historical pragmatics. At the same time, corpora representing specific groups,
genres, registers or stages of a language have led to research of various kinds, including the language of adolescents, foreign language learners and other user
groups.
In the companion chapter 22, Corpus-based pragmatics II: quantitative
studies, Christoph Rhlemann takes up quantitative corpus-based research. Due
to the massive dependence of pragmatic phenomena on context, corpora, as a
relatively decontextualized medium, have long been seen by some researchers as
unfit for use in pragmatic research. Nonetheless, corpus linguistic analyses, both
qualitative and quantitative in orientation, have produced a wealth of new insights
into key pragmatic phenomena. The aim of this article is to illustrate key quantitative corpus studies into phenomena of pragmatic interest. The article is divided
into four sections. Following a brief introduction that addresses general issues in
corpus linguistic research on pragmatic phenomena, such as the question of how
much or little context-sensitive corpora are, the second section introduces quanti-
18
tative studies carried out on the small number of corpora that are pragmatically annotated. The third, and largest, section discusses key corpus linguistic research into
pragmatic phenomena such as speech acts, deixis, evaluation, discourse marking,
backchannelling and vagueness. Further, as a borderline case between pragmatics
and semantics, research into semantic prosody will be reviewed. The concluding
section looks to the future, outlining inter alia recent attempts at building multimodal corpora that will potentially shed light on the interplay between utterance
and kinesics.
The final contribution 23 deals with various methods of transcribing face-toface interaction and their significance for pragmatic theorizing. In this chapter,
The transcription of face-to-face interaction, Roger J. Kreuz and Monica A.
Riordan consider issues in the representation of linguistic data. Researchers in
pragmatics frequently transcribe audio- or videotaped recordings of conversational
participants. A comprehensive transcription coding system would help to unify diverse lines of research and make collaboration and interpretation of results easier.
However, a comprehensive system currently does not exist. Kreuz and Riordan
review coding schemes that have been previously proposed and assess their
strengths and weaknesses. Most systems have focused solely on the acoustic (i.e.,
lexical and intonational) aspects of discourse between adult speakers, but they also
review systems that have been developed to transcribe the speech of other populations (e.g., children, aphasics and cognitively impaired individuals). In addition,
they make recommendations about how non-acoustic information can be coded.
Specifically, the challenges and utility of coding for facial and gestural information
are addressed. It is hoped that a more comprehensive coding system will make it
easier for researchers to describe and study the multimodal aspects of conversational interaction.
Taken together, these chapters constitute an outline of the fundamental issues
and trends in the study of pragmatics at the beginning of the twenty-first century.
They can be taken as a kind of extended definition of the term pragmatics as
practiced today, along with critical attempts to set down the history of the discipline and identify the sorts of questions taken to be important and the sorts of
answers being proposed. We hope in this way to contribute to the definition and coherence of the field, and thereby to foster research in the generation to come.
References
Archer, Dawn and Jonathan Culpeper
2003
Sociopragmatic annotation: New directions and possibilities in historical corpus linguistics. In: Andrew Wilson, Paul Rayson, and Tony McEnery (eds.),
Corpus Linguistics by the Lune: A Festschrift for Geoffrey Leech, 3758.
Frankfurt am Main: Peter Lang.
19
Cappelen, Herman and Ernie Lepore

2005
Insensitive Semantics. Oxford: Blackwell.
Carston, Robyn
2002
Thoughts and Utterances: The Pragmatics of Explicit Communication. Oxford: Blackwell.
Eckert, Penelope
2000
Linguistic Variation as Social Practice. Malden, MA: Blackwell.
Fairclough, Norman
1995
Critical Discourse Analysis. London: Longman.
Goodwin, Charles
2003
Pointing as situated practice. In: Sotaro Kita (ed.), Pointing: Where Language,
Culture, and Cognition Meet, 217242. Mahwah, NJ: Erlbaum.
Grice, Herbert Paul
1975
Logic and Conversation. In: Peter Cole and Jerry L. Morgan (eds.). Syntax and
Semantics. Vol. 3: Speech Acts, 4158. New York: Academic Press.
Halliday, Michael A.K.
1978
Language as Social Semiotic. London: Arnold.
Hanks, William
1996
Language and Communicative Practices. Boulder, CO: Westview Press.
Horn, Lawrence R.
1984
Toward a new taxonomy for pragmatic inference: Q-based and R-based implicature. In: Deborah Schiffrin (ed.), Meaning, Form, and Use in Context: Linguistic Applications, 1142. Washington DC.: Georgetown University Press.
Huang, Yan
2007
Pragmatics. Oxford: University Press.
Hymes, Dell
1962
The ethnography of speaking. In: Thomas Gladwin and William C. Sturtevant
(eds.), Anthropology and Human Behavior, 1353. Washington: Anthropological Society.
Leech, Geoffrey
1983
Principles of Pragmatics. London: Longman.
Levinson, Stephen
1983
Pragmatics. Cambridge: University Press.
Levinson, Stephen
2000
Presumptive Meanings. The Theory of Generalized Conversational Implicature. Cambridge, MA: MIT Press.
McNamara, Tim and Carsten Roever
2006
The social dimension of proficiency: How testable is it? Language Learning
56/2: 4379.
Mey, Jacob L.
1979
Introduction. In: Jacob L. Mey (ed.), Pragmalinguistics. Theory and Practice,
1017. The Hague: Mouton.
Mey, Jacob L.
1993
Pragmatics. An Introduction. Oxford: Blackwell.
Mey, Jacob L.
2000
When Voices Clash: A Study in Literary Pragmatics. Berlin: Mouton de
Gruyter.
20
Morris, Charles
1938
Foundations of the Theory of Signs. In: Otto Neurath, Rudolf Carnap and
Charles Morris (eds.). International Encyclopaedia of Unified Science,
77138. Chicago: University of Chicago Press.
Prucha, Jan
1983
Pragmalinguistics: East European Approaches. Amsterdam: Benjamins.
Sperber, Dan and Deirdre Wilson
1986. Relevance: Communication and Cognition. Oxford: Blackwell.
Thomas, Jenny
1995
Meaning in Interaction: An Introduction to Pragmatics. London: Longman.
Wenger, Etienne
1998
Communities of Practice: Learning, Meaning, and Identity. Cambridge: Cambridge University Press.
Part I
Conceptual foundations
1.
Pragmatics as a linguistic concept1

Anita Fetzer
1.
Introduction
Pragmatics is fundamentally concerned with communicative action and its felicity

in context, investigating action with respect to the questions of what action is, what
may count as action, what action is composed of, what conditions need to be satisfied for action to be felicitous, and how action is related to context. These research questions and the object of research require action in general and communicative action in particular to be conceived of as relational concepts, relating action
and context, relating action and communicative action, relating communicative action and interlocutors, and relating interlocutors with the things they do with words
in context. The inherent parts-whole perspective requires the explicit accommodation of a theory of context, capturing the dynamic processes from the connectedness between parts and wholes thus transcending clearly delimited frames of investigation.
The multifaceted research paradigm of pragmatics has provided new directions
and perspectives in arts and humanities, philosophy, cognitive science, computer
science and social science. Pragmatic perspectives have been employed in information technology and in the social sciences, particularly in economics, politics
and education. In line with Wittgensteins conception of language games and Levinsons conception of activity types (Levinson 1979; Wittgenstein 1958), game
theory (Lewis 1979) has been adapted to their particular contextual constraints and
requirements. The change of perspective from objects to interlocutors (or users),
and the accommodation of their wants and needs is reflected in the paradigm-anchored collocations of pragmatic thinking and learning, pragmatic software, pragmatic design, pragmatic modelling and pragmatic technology on the one hand, and
pragmatic foreign policy, pragmatic politicians, pragmatic sanction and pragmatic
nationalism on the other. Hence, it is no longer solely individual actions and their
perlocutionary effects, which are at the heart of analysis, but rather whole games,
considering both the present state of affairs and its connectedness with prior and
succeeding actions. In philosophical contexts, pragmatics is related closely to William James, who pointed out its practical nature, and to Charles Peirce, who
referred back to Immanuel Kants conception of pragmatics as experimental, empirical and purposive thought (e.g., The New Encyclopaedia Britannica 1976; cf.
also Koyama this volume). Pragmatics is also frequently connected with Greek
pragma, meaning action and instructive or useful (e.g., Bublitz 2009; Cummings 2005).
24
Anita Fetzer
The connectedness between action and purpose is also reflected in its ordinary-language meaning learning by doing (cf. Longman DCE 2003; Collins
Cobuild 1990). As a non-scholarly, ordinary-language notion, pragmatics is functionally synonymous with practical or just right at that stage, describing a surface-bound and action-based phenomenon within a particular situation. It occurs
primarily in its adjectival form, foregrounding practical and unexpected, rather
than theoretical and dogmatic considerations. For instance, politicians and economists may argue for a case on wholly pragmatic grounds, brushing aside traditionally established borders, which tend to result from presupposed and taken-forgranted ideology. Whenever an action, decision or solution is assigned the status
of a pragmatic action, a pragmatic decision or a pragmatic solution, it is contrasted implicitly with its theory-based and abstract-principles-driven counterparts. Against this background, the adjective pragmatic in its function as a modifier anchors the head of a noun phrase to the here and now of discourse. To employ
Gricean terminology from his cooperative principle, it signals that the ongoing discourse is such as is required, at the current stage at which it occurs, by the accepted purpose or direction of the talk exchange (Grice 1975: 45). Pragmatics in
its non-scholarly ordinary-language meaning thus meets with one of the most fundamental premises of scholarly pragmatics, the Gricean cooperative principle.
There is a further bridging point between non-scholarly pragmatics and scholarly
pragmatics: the differentiation between brute facts and institutional facts. According to Searle (1995), facts, that is brute facts, mental facts, social facts and institutional facts, constitute and regulate the world. To have a good knowledge of the
present situation, interlocutors need to have a good knowledge of what regulates
the world and what constitutes the world. Only then may they construct alternative
visions of a present situation or alternative social realities.
2.
Pragmatics in context
Pragmatics is frequently conceptualized as the science of language use, the study

of context-dependent meaning and the study of speaker-intended meaning, presupposing the existence of language, language user and context on the one hand, and
context-independent meaning on the other. To capture its multifaceted nature, definitions tend not to concentrate on the questions of what pragmatics is and what it
does, but rather on what it is not and what it does not do. The majority of conceptualizations pay tribute to Charles Morriss definition as the study of the relation
of signs to interpreters (Morris 1971: 6) and to Austins differentiation between
constative and performative (Austin 1971), foregrounding the performance of
communicative action and the effects the uttering of words may have. Against this
background, pragmatics is considered to be the study of invisible meaning (Yule
1996: 127), the science of the unsaid (Mey 2001: 194), the study of meaning as it
25
emerges in language use (Marmaridou 2000: 1), and the study of linguistic acts
and communicative action and their appropriateness (Bublitz 2009; van Dijk 2008;
Fetzer 2004, 2007).
There have been various attempts to systematize the multifaceted and heterogeneous field of pragmatics, or to bring some order into the pragmatic wastebasket
(Bar-Hillel 1971), as is reflected in the differentiation between the broad fields of
scholarly pragmatics: general (or non-linguistic) pragmatics and linguistic pragmatics. Implicit in the philosophy-anchored general-action and linguistic-actionbased paradigms is the premise that pragmatics is conceptualized best as a perspective, which comprises a general pragmatic perspective, a social perspective, a compositional perspective and a relational perspective. These are introduced below.
2.1.
The pragmatic perspective
The pragmatic-perspective paradigm does not represent a clearly delimited field of

research but rather offers a general perspective towards the object to be examined.
It provides a point of view for analysis, in which not the object in isolation is examined but rather the nature of its connectedness with users and other objects, or with
the theoretical framework to which it is anchored. The perspective is thus shifted
from an analysis of the language system and its constitutive parts to that of its rulegoverned instantiation and embeddedness in context, considering its generalized
and particularized conditions of use. That change in perspective has important consequences for methodology. To account for the parts-whole connectedness, a relational frame of reference is adopted, which accommodates the relation between
the language system, language use and language users, and what they do with
words in context, comprising linguistic context (or co-text), social context and
cognitive context.
The pragmatic-perspective paradigm provides a general cognitive, social and
cultural perspective on linguistic phenomena in relation to their usage in forms of
behaviour, accounting for the dynamics of language and language use, as is reflected in the premises that meaning is not a product and given but rather dynamic,
multifaceted and negotiated in context. Language use is dependent on variability,
and language users adapt to contextual constraints accommodating them not
only in the formulation of utterances but also in their interpretation (e.g. Mey 2001;
Verschueren 1999). As in functional grammar and systemic functional grammar
(Givn 1993; Halliday 1994), the dynamic construct of meaning is a necessary
condition for making linguistic choices. Following Verschueren (1999: 5558), the
choices are made at every possible level of structure as regards form and strategy.
In line with the sociolinguistic premise of linguistic variation, which claims that
speaker-intended meaning can be formulated with different linguistic surfaces
(Brown 1995), these choices are not equivalent in status. They carry along their alternatives and are always seen and interpreted against the background of other,
26
Anita Fetzer
more or less marked possibilities. While unmarked variants are seen as more frequent in distribution, more conventional in formulation and easier to process,
marked variants are less frequent in distribution, less conventional in formulation
and harder to process (Fetzer 2002, 2007).
The pragmatic perspective is not only reflected in the more general approach
examined above, considering linguistic communication as well as linguistic behaviour. It is also found in the particularized, hyphenated fields of pragmatics, as in
cognitive pragmatics (cf. vol. 4 Cognitive pragmatics of this series Handbooks of
Pragmatics), empirical pragmatics (cf. Bednarek this volume, Andersen this volume, Rhlemann this volume), experimental pragmatics (cf. Breheny this volume), formal pragmatics, historical pragmatics (cf. vol. 8 Historical pragmatics of
this series Handbooks of Pragmatics), intercultural pragmatics (cf. vol. 7 Pragmatics across languages and cultures of this series Handbooks of Pragmatics), interlanguage pragmatics, sociopragmatics (cf. vol. 5 Pragmatics of society of this
series Handbooks of Pragmatics) and theoretical pragmatics.
2.2.
The social perspective
An analysis of the rule-governed and strategic use of language in context anchored

in a social-perspective goes beyond traditional linguistic analyses. It also considers
social and sociocultural aspects of communication in an explicit manner, assigning
them the status of a constitutive part of communicative action, thus overlapping
with the research paradigms of anthropological linguistics (Duranti 1997; Hanks
1996), interactional sociolinguistics (Auer and Di Luzio 1992; Gumperz 1992,
1996, 2003), (critical) discourse analysis (Fairclough 1992; van Dijk 2008) and
ethnomethodological conversation analysis (Garfinkel 1994; Heritage 1984).
Societal pragmatics examines social parameters which affect the production
and interpretation of utterances, placing language use in an external relation to language users. It is based on the premise that language use and social structure are
connected dialectically, (re)constructing social and sociocultural context by confirming (or disconfirming) social values in interaction, for instance gender, power
and social status. Consequently, the language user and her/his social roles and
identities are at the heart of societal pragmatics, analysing social and sociocultural
context on the micro level of face-to-face interaction as well as on the macro level
of institution. In interaction, language users interactionally organize their social
roles and identities while at the same time constituting social context. To use Heritages (1984: 242) terminology, the production of talk is doubly contextual. An utterance relies upon the existing context for its production and interpretation, and it
is, in its own right, an event that shapes a new context for the action that will follow.
The social perspective to pragmatics has shifted the focus of investigating context-dependent meaning from semantics-based methodologies and their premise of
27
semantic well-formedness and propositional truth to social context-based approaches and their premise of negotiated communicative meaning and sociocultural appropriateness (Fetzer 2007), thus adding a further layer of meaning: social
and sociocultural meaning. In this perspective, meaning is not given per se and carried as such into the utterance, metaphorically speaking. Rather, meaning, viz.
lexical meaning and utterance meaning, is brought into the utterance and its diverse layers of meaning are brought out in the utterance, foregrounding salient
meanings while backgrounding non-salient ones (to employ interactional-sociolinguistic terminology, Gumperz 1992, 1996). Language is assigned the status of a
sociocultural construct which is used strategically by rational language users in
context, considering possible perlocutionary effects their utterances may trigger as
regards negative and positive politeness (Brown and Levinson 1987), for instance.
The social-meaning perspective also prevails in ethnographic studies, sociolinguistics and interactional sociolinguistics, as well as in ordinary language philosophy (Brandom 1994; Recanati 2004; Searle 1969, 1975, 1995). It is in disaccordance with frameworks conceptualizing language and language use as mutually
exclusive concepts, as is the case with the dichotomies of linguistic competence
and communicative performance, I(nternalized)-language and E(xternalized)-language, or langue and parole.
In the Anglo-American research tradition, the social perspective and the pragmatic perspective are referred to as European-Continental pragmatics. This field
defines pragmatics in a far broader way, encompassing much that goes under the
rubric of sociolinguistics, psycholinguistics, and discourse analysis (Huang 2007:
4), thus not falling strictly within formal linguistic theory (Horn and Ward 2004:
xi). The Anglo-American tradition tends to favour the compositional perspective.
2.3.
The compositional perspective
In the pragmatics-as-component perspective pragmatics is assigned the status of

a necessary component (or module) of grammar, thus anchoring pragmatics firmly
to linguistics (Birner and Ward 2006; Carston 2002; Huang 2007; Horn 1984;
Jaszczolt 2005; Levinson 2000; Sperber and Wilson 1995). Pragmatics is conceived of as an individual component of grammar, interacting with the components
of semantics, morphology, syntax and phonology, as is the case in its most prominent representatives: theoretical pragmatics and formal pragmatics. Radical pragmatics departs from the premise of strict modularity or discrete compositionality
by merging semantics and pragmatics. It is based on the premise of sense generality and contextualism, presuming that the output of semantic meaning is underdetermined and assigned determinate meaning in context through inference and pragmatic enrichment.
The compositional perspective is based on the premise that language is a theoretical construct focussing on language-internal rules and regularities, and on in-
28
Anita Fetzer
ferencing processes, thus considering language use and context in an explicit

manner. In contrast to societal pragmatics, however, its conception of context tends
to be restricted to linguistic context. In the case of deixis and its realization as deictic expressions, e.g., temporal and local deixis, participant deixis and discourse
deixis (Hanks 1996 and this volume; Levinson 1983), the compositional perspective needs to accommodate physical components anchored in the social context,
e.g., speaker and hearer, and time and place. Participants tend to be conceptualized
as the prototypical constructs of Model Speaker, Model Addressee or Model Person, viz. ideally competent interlocutors in the domains of linguistic competence,
communicative competence and rationality (Brown and Levinson 1987; Jaszczolt
2005), abstracting from their social, discursive and interactional roles. The same
holds for the conceptualization of time and place.
At the heart of compositional-perspective investigations are pragmatic universals, viz. deixis, reference, non-literal meaning, indefiniteness, presupposition, information structure, discourse marker, speech act and implicature. To account for
their function, pragmatics differentiates between an abstract construct and its linguistic realization (or representation). For instance, the pragmatic universal of deixis is examined with respect to deictic expression, and the universal of speech
act is analysed with respect to its constitutive acts and their direct, indirect, and
conventionalized indirect realizations (Sbis 2002; Searle 1995; Brown and Levinson 1987). Closely connected with indirect speech acts are speaker- and heareroriented cognitive operations: the former are conventional implicature, generalized and particularized conversational implicature (Levinson 2000; Horn 1989;
Huang this volume), and impliciture (Bach 2006), and the latter are inference and
abduction (Cohen, Morgan and Pollack 1992; Givn 1989, 2005). Inferencing is
also indispensable for the definition of intentionality of communicative action and
non-literal meaning, for indefiniteness and discourse markers, and for presupposition and information structure. Another pragmatic universal is face (Brown and
Levinson 1987; Goffman 1981, 1986), which is also of key importance to the social perspective.
More recently, the compositional perspective has undergone a shift in focus
and methodology, calling for the accommodation of methodological compositionality to bridge the gap between two competing paradigms (Jaszczolt 2005).
Only then is it possible to account for speech act and uptake, speech act type and
cognitive reality, and speech act type and situation (mis)match. The change in direction from monological, speaker-centred methodologies to dialogical, co-participant-centred interaction requires the explicit consideration of the connectedness
amongst world knowledge, linguistic knowledge, discourse knowledge, the properties of the human inferential system and stereotypes and presumptions about society and culture, as is accommodated in the relational perspective.
2.4
29
The relational perspective
The three perspectives adopted towards the heterogeneous field of pragmatics all
share a particular perspective towards their object of investigation, accounting for
its connectedness with context as regards utterance meaning and conditions of use.
For this reason, they adopt a more or less explicit relational frame of reference.
They go separate ways, however, in their accommodation of social context. While
the social perspective considers social context to be a key concept, the compositional perspective tends to disregard it, concentrating on linguistic context and
cognitive context. All of them are based on the premise that language is as a rulegoverned system, and all of them intend to account for form, function and meaning. The types of meaning under examination and their conceptualizations are different, though.
In a relational frame of reference, language may be examined from a parts perspective, analysing the form, function and meaning of individual parts, viz.
sounds, morphemes and constituents. It may be investigated from a holistic viewpoint, viz. phonological words and lexical words, and propositions and sentences,
and it may be analysed from a parts-whole perspective considering the nature of
their connectedness. The latter perspective is intrinsically dynamic, focussing not
on the object as such, but rather on its embeddedness in context, its conditions of
use and its appropriateness. The examination of linguistic form does not generally
cause delimitation problems, while the analyses of linguistic meaning and function
have become increasingly messy, necessitating the explicit accommodation of a
pragmatic wastebasket for non-fits and other-fits.
Pragmatics is generally delimited from semantics by shifting the focus of investigation from what is said, viz. semantic, literal or truth-conditional meaning, to
what is meant, viz. what is communicated implicitly, what is implicated, utterance
meaning or meaningnn (e.g., Grice 1975; Jaszczolt 1999, 2005; Levinson 1983,
2000; Recanati 2004; Saeed this volume). While the object of research is meaning
in semantics as well as in pragmatics, accounting for world-to-language, languageto-world and language-to-world-to-mind relations, the delimitation of the object,
the focus of investigation and the methodologies employed differ. In semantics the
meaning of a lexeme, proposition or discourse unit is defined from a unit-of-investigation-internal perspective, and in pragmatics, the meaning of an utterance,
speech act or conversational contribution is defined from unit-of-investigation-external and unit-of-investigation-internal perspectives. Metaphorically speaking,
semantics examines what is in the container, and pragmatics examines what is
brought out of the container (Reddy 1979).
The following figure attempts to bring some order in the pragmatic wastebasket
by systematising the two closely related but yet distinct fields of research, which
intersect and overlap in their analysis of meaning in context.2 While meaning is
conceptualised as conventional, truth-conditional meaning with discrete bound-
30
Anita Fetzer
Figure 1. Meaning in context
31
aries in semantics, pragmatic meaning is seen as speaker-intended, non-truthconditional meaning with fuzzy boundaries, as is indicated by the dotted lines. Furthermore, meaning is autonomous and represented by well-formed context-independent sentences in semantics, and it is inferred and represented by appropriate
context-dependent utterances in pragmatics. Semantics and pragmatics meet at the
interpretation interface, where the calculation of utterance meaning is informed by
the linguistic code and at the same time by pragmatic inferencing. While the
former is a unidirectional operation as is signified by single arrows, the latter is dialectical and interactive in nature as is indicated by double arrows, capturing the dynamics of pragmatic meaning:
The performative and its constitutive parts, viz. meaningnn, non-truth-conditional meaning and presupposition, are among the most important pillars of pragmatics. They are also relevant to the definition of speech acts and their realization
in context as utterance meaning, which is a linguistic-pragmatics phenomenon par
excellence. It is at the interpretation interface where context-dependent implicatures are calculated and where pragmatics interacts with semantics, syntax, morphology and phonology.
In the following, the research domain of general pragmatics is introduced, concentrating on its bridging points with linguistic pragmatics.
3.
General pragmatics
General pragmatics is set firmly in the research paradigm of philosophy, considering cooperation, action theory, intentionality, rationality and context, thus providing relevant bridging points for language-anchored linguistic pragmatics. Both research paradigms analyse pragmatic universals and pragmatic mechanisms, such
as speech act and implicature (e.g. Austin 1971; Bach 1992, 2006; Grice 1975;
Sbis 2002; Searle 1975; Vanderveken and Kubo 2002), indexical and deixis (Recanati 2004; Levinson 1983), common ground and presupposition (Akman et al.
2001; Bouquet et al. 1999), focussing on the connectedness between parts and
wholes, and between content and context (Stalnaker 1999). While general pragmatics adopts an action-based frame of reference, linguistic pragmatics examines
the form and function of speech acts in various languages.
In general pragmatics, action and cooperation are analysed in the framework of
an ideal pragmatic situation in which communication is neither distorted by social
factors, e.g. institutional constraints, unequal social hierarchies and other forms of
oppression (e.g., Habermas 1998, cf. Cooke this volume), nor by interpersonal
considerations, such as face-wants and face needs (Brown and Levinson 1987),
politeness or etiquette (Watts, Ide and Ehlich 1992). The high degree of idealization inherent in this research paradigm is also reflected in its conception of language and language use, which does not refer to natural language as spoken by or-
32
Anita Fetzer
dinary language users in context but rather an ideal language built in accordance
with the language-system specific constraints and requirements. In an ideal pragmatic situation, all necessary felicity conditions obtain, such as normal input and
output conditions, propositional content and preparatory conditions, sincerity and
essential conditions (Searle 1969). Consequently, rational agents have ideal linguistic, communicative and cognitive competence as well as an ideal world knowledge (cf. Cohen, Morgan and Pollack 1992).
Pragmatic universals can be assigned the status of basic building blocks of
communication. Among the most prominent ones are the speech act and its constitutive acts, viz. propositional act composed of further constitutive acts and illocutionary act (Searle 1969; Vanderveken and Kubo 2002), indexicals, for instance
pure and impure indexicals, intentionals and perspectivals (Recanati 2004), and
presuppositions and presupposition triggers (Huang 2007; Levinson 1983; Stalnaker 1999). Another basic building block is the rational agent or model user,
who employs building blocks in a strategic manner to speaker-intend communicative meaning, thus achieving particular goals in context. The model user can be a
natural user, that is speaker and hearer of a natural language, or an artificial user,
such as a robot or a dialogue system. In accordance with the fundamental pragmatic premises of rationality and intentionality, the model user is equipped with a
reasoning device, accommodating pragmatic inferencing and default inferencing,
top-down reasoning and bottom-up reasoning, and abduction. This is a necessary
condition for a model user to calculate and formulate speaker-intended implicatures and to perform the hearer-intended logical operations of inference and abduction (Cohen, Morgan and Pollack 1992; Givn 1989, 2005; Huang this volume;
Sperber and Wilson 1995).
3.1.
Presupposition and common ground
Presupposition refers to a proposition or an inference whose validity is taken for

granted for a sentence to be true or for a speech act to be felicitous. Philosophy and
linguistics distinguish between pragmatic presupposition and semantic presupposition. The satisfaction of semantic presupposition is a necessary condition for
the truth value of a sentence, and the satisfaction of pragmatic presupposition is
necessary for a speech act to be appropriate in context. Pragmatic presuppositions
are accommodated in speech act theorys felicity conditions, which are considered
as linguistic and social context categories and their satisfaction is assigned the
status of a default configuration (Sbis 2002). Presuppositions are generated by
presupposition triggers, for instance definite descriptions, factive verbs, aspectual
verbs, change of state verbs, iteratives and clefts (cf. Huang 2007, this volume; Levinson 1983).
Some contexts allow presupposition neutralisation, and more recently presupposition triggers have been classified as hard presupposition triggers and soft pre-
33
supposition triggers. Presuppositions are connected intrinsically with propositions

and assumptions which are taken for granted and therefore do not need to be made
explicit. They are organized and administered in the framework of context sets
(Stalnaker 1999), which serve as common ground in communication. Consequently, pragmatic presupposition may be seen as a restriction on common ground.
Common ground is indispensable for philosophical and cognitive conceptions
of knowledge where it serves as background for reasoning and for retrieving
speaker-intended meaning and other types of implicit meaning, such as indexical
expression or implicature. In the field of computer science, common ground and
world knowledge are frequently conceptualized as a database. It is seen as comprising a set of propositions, which serve as a resource for the understanding of utterances. Common ground is implicit but can be made explicit via propositions,
and model users presuppose its validity and fall back on it when they retrieve implicatures (cf. Bublitz 2006).
Common ground in the sense of background also plays a fundamental role in
dialogue system modelling. According to Vanderveken and Kubo (2002), model
users negotiate the compatibility of background with utterances and their felicity
and satisfaction conditions in and through the process of communication. However, background and context are not identical because possible contexts of utterance can have different backgrounds. As a consequence of this, background contains not only mutual knowledge of facts about the conversational background but
also knowledge about the world and of the world, such as ethical norms and sociocultural values, transcending the common sense notion of context. Searle (1995)
considers background to be a necessary condition both for literal and non-literal
meaning, thus assigning it the status of a basic premise for felicitous communication. He defines background as an open-ended set of skills, pre-intentional assumptions and practices, which are not representational but rather enable intentional acts and states to be made manifest.
In Relevance Theory (Sperber and Wilson 1995), common ground is conceived
of as a common set of premises for inference rules. The relevance-theoretic conception differs from the traditional notion of mutual knowledge by its attempt to
avoid the logical consequence of infinite regress, which follows from the code
model of communication. To avoid infinite regress, Sperber and Wilson base their
theoretical framework on an approximation of mutual knowledge, namely cognitive environments and mutual manifestness.
Common ground is a presupposed common knowledge base, which is a necessary condition for felicitous communication. To that base, model users anchor their
communicative action and they fall back on it, should they require further information, which may not be encoded explicitly. The overall encompassing concept of common ground has been given a context-dependent interpretation as conversational record (Thomason 1992) based on Lewiss concept of accommodation
(Lewis 1979). Conversational record contains public information, presumptions,
34
Anita Fetzer
and an update operator. Other context-dependent subsets of common ground are

Clarks notions of personal common ground, which contains a model users subjective experience, and communal common ground, which stores social experience
(Clark 1996), and Fetzers dialogue common ground, which is differentiated in individual dialogue common ground, containing a model users representation of
her/his dialogue common ground, and collective dialogue common ground, viz. the
model users representation of the set of model users representation of dialogue
common ground (Fetzer 2004). All of the subsets need to be interconnected, constantly updated and, if necessary, revised.
3.2.
Context
The analysis of context-dependent meaning is at the heart of pragmatics, and for

this reason context is one of its key objects of investigation. The theory, practice
and implementation of context are also of relevance to diverse fields of investigation, ranging from philosophy and computer-mediated communication to cognitive science, in particular dialogue management, artificial speech production,
artificial intelligence, distributed knowledge representation, robotics and information technology.
The heterogeneous nature of context and the context-dependence of the concept itself have made it almost impossible for the scientific community to agree
upon one commonly shared definition or theoretical perspective, and frequently,
only a minute aspect of context is described, modelled or formalized (cf. Akman et
al. 2001; Blackburn et al. 2003; Bouquet et al. 1999). Because of its multifaceted
nature and inherent complexity, context is no longer considered an analytic prime
but rather seen from a parts-whole perspective as an entity containing sub-entities
(or sub-contexts).
The multilayered outlook on context contains a number of different perspectives. First, context is conceived as a frame of reference whose job it is to frame
content by delimiting the content while at the same time being framed and delimited by less immediate adjacent frames. The nature of the connectedness between the different frames is a structured whole composed of interconnected
frames (Goffman 1986). The gestalt-psychological figure-ground scenario prevails
in psychological and psycholinguistic perspectives on context. It is also adopted in
cognitive pragmatics as is reflected in the relevance-theoretic conception of context as an onion, metaphorically speaking. The individual layers are interconnected
and their order of inclusion corresponds to their order of accessibility (Sperber and
Wilson 1995), as is reflected in inferencing and other kinds of reasoning.
Second, context is seen as a dynamic construct, which is interactionally organized in and through the process of communication. This view prevails in ethnomethodology (Garfinkel 1994; Goodwin and Duranti 1992; Heritage 1984), interactional sociolinguistics (Gumperz 1996, 2003) and sociopragmatics (Bublitz 2009;
35
Fetzer 2007, Schmid 2003), where context is assigned the dual status of process
and product. The dynamic outlook is based on the premise of indexicality of social
action, and the joint construction of context. In the primarily qualitatively oriented
paradigms, context is connected intrinsically with adjacency pair, conditional relevance and the turn-taking system on the micro level, and with institution on the
macro level, whose order is captured through context-independent and contextsensitive constraints and requirements. Closely related to the conception of context
as a dynamic construct is its relational conception, which conceives it as relating
communicative actions and their surroundings, relating communicative actions,
relating individual participants and their individual surroundings, and relating the
set of individual participants and their communicative actions to their surroundings (Fetzer and Akman 2002).
Third, context is seen as given as is reflected in the presuppositional approach
to context, which is also referred to as common ground or background information
(Stalnaker 1999). Here, context is seen as a set of propositions, which participants
take for granted in interaction. This allows for two different conceptions of context: a static conception in which context is external to the utterance, and an interactive one, in which context is imported into the utterance while at the same time
invoking and reconstructing context.
The context-dependence of context is reflected in its status as given and external to the utterance, reconstructed and negotiated in communication, indexical, and
never saturated. A further classification of context is anchored to a holistic conception of context embedding its constitutive parts of model user, conversational contribution, surroundings and their presuppositions, viz. cognitive context, linguistic
context and social context (Fetzer 2002, 2004).
Linguistic context comprises language use and is delimited by the constraints
and requirements of genre. Language is composed of linguistic constructions (or
parts) embedded in adjacent linguistic constructions (further parts), composing a
whole clause, sentence, utterance, turn or text. Linguistic context is functionally
synonymous to co-text (de Beaugrande and Dressler 1981; Widdowson 2004), denoting a relational construct composed of local and global adjacency relations.
Analogously to speech-act-theoretic constitutive rules and regulative rules (Searle
1969), linguistic constructions constitute text. The production and interpretation of
linguistic constructions is constrained by the rules of grammar, and the production
and interpretation of speech acts are constrained by felicity conditions.
Cognitive context is the foundation on which inference and other forms of reasoning are based. Constitutive elements of cognitive context are mental representations, propositions, contextual assumptions and factual assumptions. Since cognitive contexts are anchored to an individual but are also required for a cognitively
based outlook on communication, they need to contain assumptions about mutual
cognitive environments. Thus, cognitive context is not only defined by representations but also by meta-representations. In the social-psychological paradigm,
36
Anita Fetzer
cognitive context is conceptualized along the lines of the gestalt-psychological distinction between figure and ground and the related metacommunicative concepts
of frame and framing (Bateson 1972; Goffman 1986). Frame is seen as a delimiting
device, which is (or delimits) a class or set of messages (or meaningful actions)
(Bateson 1972: 187). Because of its delimiting function, psychological frames are
exclusive, i.e. by including certain messages (or meaningful actions) within a
frame, certain other messages are excluded and they are inclusive, i.e. by excluding certain messages certain others are included (ibid.). This also holds for context which, analogously to frame is also structured and metacommunicative, or to
use Batesons words: the hypothesis depends upon the idea that this structured
context also occurs within a wider context a metacontext if you will and that
this sequence of contexts is an open, and conceivably infinite, series (Bateson
1972: 245).
Social context comprises the context of a communicative exchange and is defined by deducting linguistic context and cognitive context from a holistic conception of context. Constituents of social context are, for instance, model users, the
immediate concrete, physical surroundings including time and location, and the
macro contextual institutional and non-institutional domains. The connectedness
between language and language use on the one hand, and linguistic context, social
context and cognitive context on the other is reflected in deixis, viz. temporal deixis, local deixis, participant deixis, discourse deixis and social deixis (Hanks
1996). Furthermore, the category of model user (as speaker and hearer) can no
longer be conceived of as an analytic prime but needs to be refined by the accommodation of footing (Goffman 1981; Levinson 1988). The importance of social
context to communication is spelled out by Hanks as follows: Hence it is not that
people must share a grammar, but that they must share, to a degree, ways of orienting themselves in social context. This kind of sharing partial, orientational and
socially distributed may be attributed to the habitus, or relatively stable schemes
of perception to which actors are inculcated (Hanks 1996: 235).
3.3.
Cooperation
The pragmatic principle of cooperation is connected intrinsically with coordination and collaboration on the one hand (Clark 1996; Cummings 2005; Grosz and
Sidner 1992), and with the pragmatic principles of rationality and intentionality
(Cohen, Morgan and Pollack 1992; Searle 1983) on the other. The principles are
examined in philosophy as regards their contents, function and status (Brandom
1994; Penco 1999), and they are analysed in psychology as mental representations,
mental states and beliefs (Bouquet et al. 1999; Sperber and Wilson 1995). Against
this background, language and language use are seen as fundamentally cooperative, and subsequently as intrinsically discursive and social (e.g., Grice 1975;
Lewis 1979; Clark 1996).
37
The connectedness between cooperation and intentionality, and the differentiation between I-intentions and we-intentions are at the heart of Searles conception
of speech act theory. To use his own words: Collective intentionality presupposes
a Background sense of the other as a candidate for cooperative agency; that is, it
presupposes a sense of the others as more than just conscious agents, indeed as actual or potential members of a cooperative activity (Searle 1991: 414). In his
work, Searle is very explicit about the nature of the parts-whole connectedness of
individual speech acts and I-intentions on the one hand, and of a conversation or
discourse and we-intentions on the other: The reason that we-intentions cannot be
reduced to I-intentions, even I-intentions supplemented with beliefs and beliefs
about mutual beliefs, can be stated quite generally. The notion of a we-intention of
collective intentionality implies the notion of cooperation (Searle 1991: 406).
In the Gricean approach to natural-language communication, in which language is seen as both purposive and social, and rational and cognitive, cooperation
counts as the fundamental premise of communication. As a philosophical concept,
cooperation is anchored to the premise of rationality and intentionality. Consequently, conversations are rational endeavours and cooperative efforts, in which
model users realize and recognize a common purpose or set of purposes, or, at
least, some mutually accepted direction. This is reflected in the Gricean cooperative principle which reads as follows: Make your conversational contribution
such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged (Grice 1975: 45). The cooperative principle is a necessary requirement for the calculation of pragmatic
meaning or meaningnn, and is for this reason not open for negotiation. Model users
may, however, employ the pragmatic strategy of opting out, which is generally accompanied by explicit references to some higher-order moral principle, such as
confidentiality or trust. When they opt out, they generally spell out the reasons for
their non-compliance, thus acting in accordance with the cooperative principle on a
deeper level. The cooperative principle subsumes four conversational maxims, the
maxim of quality, the maxim of quantity, the maxim of relation, and the maxim of
manner (Grice 1975: 4546). Further constitutive parts of the cooperative principle are the conventional and conversational implicatures, which are dealt with
below in the section 4.2 on linguistic pragmatics.
A cooperation-anchored outlook on language and language use not only is
manifest in a conversation as a whole, but also in its building blocks of speech act
and in the act of referring and predicating. Reference operates on the phrasal level
and involves a speakers use of a linguistic expression, which tends to be a noun
phrase, to induce a hearer to access or create some entity in her/his mental model of
the discourse. Reference may be performed directly with a noun phrase and indirectly as deferred reference in the contexts of pars-pro-toto configurations, meaning transfer and figurative meaning. Referring seen from a cooperation viewpoint
is not restricted to the semantic domain of a proposition but also applies to the in-
38
Anita Fetzer
terpersonal and interactional domains, viz. negotiation of meaning and common

ground.
The status of cooperation as a universal pragmatic principle has been examined
extensively in general pragmatics, and there seems to be a general agreement on its
validity. However, the claim that the Gricean cooperative principle is universal has
been under severe attack. There have been modifications as regards the adoption of
a further constitutive pragmatic principle on the same level, the politeness principle (Leech 1983), and there have been various modifications of the number of
conversational maxims required for the calculation of pragmatic meaning as well
as their speaker- and hearer-oriented conceptualizations and interpretations (Horn
1984; Lakoff 1973; Levinson 2000).
Pragmatics is a field of research which intersects with various paradigms. As
regards meaning, it cannot but consider semantics. As regards rationality, it cannot
but consider cognitive science, and as regards cooperation, it cannot but consider
philosophy. Thus, pragmatics needs linguistics and linguistics needs pragmatics.
For a thorough investigation of meaning, we need to examine how things are done
with words, and how conversational contributions are structured to achieve particularized goals in context. This is done in linguistic pragmatics, where generalpragmatic principles and pragmatic universals undergo language- and culture-specific modification.
4.
Linguistic pragmatics
Linguistic pragmatics is defined as the science of language use, and [i]n the same
way as human actions change existing reality, linguistic actions also change the
world (Marmaridou 2000: 22). Linguistic pragmatics and general pragmatics
share almost identical goals: general pragmatics examines pragmatic principles,
mechanisms and universals in the context of action theory, rationality and intentionality, while linguistic pragmatics focuses on their instantiation in language and
language use. Hence linguistic pragmatics overlaps with general pragmatics, sharing its generalized principles, mechanisms and universals, and departs from the
generalized framework by concentrating on language as a general construct and on
languages as particularized instantiations.
In linguistics, language tends to be examined in its own right, accounting for
language use, but not for the model user. The language system comprises the constitutive subsystems of phonology, morphology, syntax and semantics. These may
be assigned the status of autonomous modules, as is the case in the formal paradigms, they may be conceived of as connected subsystems, as is the case in systemic functional grammar (Halliday 1994), or as dialectically connected subsystems with fuzzy boundaries, as for instance the cognitive paradigm (Givn 1993,
2005), where language use is connected dialectically with the language system and
39
the model user. Language is seen as a dynamic construct changing through patterned use in context, in which grammatical and pragmatic aspects of interpretation
interface (Ariel 2008: 257259).
The dialectical connectedness between languages constitutive subsystems
requires the explicit accommodation of pragmatic interfaces. The pragmatic perspective adopted towards the object to be examined does not consider the object in
isolation but rather its connectedness with its conditions of use and with the user.
This is reflected in grammaticalization or grammaticization (Traugott 1988) and in
pragmaticalization (Aijmer 1997). Pragmatic principles shape the lexicon, as is reflected in Horns division of pragmatic labour, viz. his hearer-oriented Q-principle
and speaker-oriented R-principle (Horn 1984), and their impact on the structure
and use of lexical items, e.g. pragmatic strengthening, scalar implicature or polarity items. It is also manifest in deictic expressions and tense and aspect markers,
whose meanings are constant but whose referents vary with the speaker and the
hearer, and with time and place of the utterance, and on the style, register or purpose of a speech act (Ariel 2008; Huang 2007). Furthermore, pragmatic principles
have a strong influence on the information status of constituents within and across
sentences regarding focus, topic (or theme) and comment (or rheme). Deviations
from the default generate implicatures, such as emphasis implicating a set of alternatives (Birner and Ward 1998; Lambrecht 1994).
4.1.
Speech act
Speech act theory is connected intrinsically with J.L. Austins groundbreaking

lectures on ordinary language philosophy, edited posthumously (Austin 1971).
Austin scrutinized the then prevailing view that language describes the world in
terms of true and false, pointing out its shortcomings if applied to natural (or ordinary) language. He compared and contrasted the constative view, according to
which sentences are true or false, with the performative approach, according to
which sentences are used to perform speech acts, and came to the conclusion that
model users do not just describe the world, but rather do things with words, even
when describing the world. This is further refined by J.R. Searles modifications
and refinement in his research on speech acts (Searle 1969, 1991), systematizing
the cognitive, linguistic and social contexts, in which speech acts are performed
by differentiating between constitutive rules, which are a necessary part of the
language game and make speech acts count as particularized speech acts, e.g.
promise, threat or rejection, and regulative rules, which are normative in nature.
Considering language use from an action-theoretic frame of reference has paved
the way for investigating linguistic meaning as well as speaker-intended meaning. It has shifted the focus of investigation from the rigid framework of formal
semantics to the action-theoretic premises of rationality, intentionality and communication.
40
Anita Fetzer
Speech act theorys basic unit of investigation is the speech act, and depending
on the frameworks employed, speech acts are divided into locutionary, illocutionary and perlocutionary acts (Austin 1971), or propositional and illocutionary acts
(Searle 1969), and their respective sub-acts, viz. reference and predication in the
Searlean paradigm, and the phonetic act, phatic act and rhetic act in Austins frame
of reference. Speech acts and their constitutive acts are performed simultaneously,
and a speech act may be realized as a direct speech act, as an indirect speech act
and as a conventionally indirect speech act. To account for a speech acts perlocutionary effects as regards the social and interpersonal planes of discourse, speech
acts have been further refined and redefined as face-threatening acts, in which the
impact of a speech acts perlocutionary effects on the interpersonal plane is calculated with respect to the model persons, to employ Brown and Levinsons terminology, negative and positive face wants (Brown and Levinson 1987).
Speech act theory is anchored in action theory on the one hand, and in language
and language use on the other. Speech acts are described in various ways: they are
classified as explicit performative utterances, viz. utterances in which the speech
act contains a performative formula realized by a first-person pronoun and a
speech-act verb, e.g. I/we hereby close the meeting or we hereby request you to pay
your dues; they are classified as first-order speech act and second-order speech act,
accounting for the differentiation between direct and indirect speech acts; they are
classified as micro speech act and macro speech act, accounting for the larger unit
of genre (van Dijk 1981; Fetzer 2002), and as nth order speech acts, capturing the
multi-layered nature of performing speech acts in context. Furthermore, speech
acts are categorized with respect to possible directions of fit between language and
the world, that is the words-to-world direction of fit for assertion, the world-towords direction of fit for directive and commissive, the double direction of fit for
declaration and the empty direction of fit for expressive (Searle 1979). In addition
to the illocutionary-point anchored taxonomies, speech act verbs are classified and
systematized according to the way they are used in a particular language (cf. Collavin this volume).
Speech act theory combines language-internal and language-external factors
by systematizing them in the framework of felicity conditions, which are classified
with regard to propositional content conditions, preparatory conditions, essential
conditions and sincerity conditions. While propositional content conditions generally specify restrictions on the content of the proposition, preparatory, essential
and sincerity conditions explicate language-external conditions concerning the
language game as well as speaker- and hearer-specific requirements.
Speech acts are neither true nor false but rather felicitous or infelicitous. To be
felicitous, they need to be produced (and interpreted) in accordance with generalized felicity conditions, viz. normal input and output conditions, the essential and
sincerity conditions, and with particularized felicity conditions, viz. proposition
content condition.
41
Speaking a language is seen as engaging in a (highly complex) rule-governed form of behavior (Searle 1969: 12). The rule-governed behaviour is specified in speech act theorys felicity conditions and in the requirement that the model
user in her/his role as speaker intends that her/his utterance will produce in a model
user in her/his role as hearer a belief that the sincerity and essential conditions obtain by means of the recognition of the intention to produce this belief, and that the
speaker intends that recognition to be achieved by means of the recognition of the
utterance as one conventionally used to produce such beliefs. Secondly, that the
semantic rules of the language spoken by the model users are such that the utterance is uttered sincerely and such as is required just in case all of the previous conditions obtain.
4.2.
Implicature
The cooperative principle counts as a universal principle in pragmatics, where it

represents the solid base to which communication in general and the formulation
and interpretation of communicative action in particular are anchored. This holds
especially for the calculation of context-dependent communicative meaning and
the necessary processes of inferencing required for the contextualization and enrichment of underspecified conversional contributions. To use Ariels words,
[u]nderdeterminacy is an inherent characteristic of human language, since no
natural language sentence can encode interlocutors intended statements fully
(Ariel 2008: 265). In a similar vein, but more explicitly, Levinson (1995) argues
that intentionality is a fundamental premise of natural-language communication.
In his words, human interaction, and thus communication depends on intentionascription. Achieving this is a computational miracle: inferences must be made
way beyond the available data. It is an abductive process of hypothesis formation,
yet it appears subjectively as fast and certain the inferences seem determinate,
though we are happy to revise them when forced to do so (Levinson 1995: 241).
Underdeterminacy and the necessary processes of inferencing are connected
with the Gricean paradigm and its differentiation between what is said and what is
meant, which both refer to utterances produced in context (Grice 1975); and they
are also connected with Relevance Theory (Sperber and Wilson 1995). In both
frames of reference, communication is seen as a context-dependent endeavour, in
which communicative meaning may go beyond the level of what has been said.
Hence, what is said cannot be equated with pure linguistic meaning but rather is
closely related to the conventional meaning of the words (the sentence) [] uttered (Grice 1975: 44). Unlike the rather controversial status of what is said in
semantics and pragmatics, what is meant has always been equated with what is
implicated.
Relevance Theory differentiates between a code model of communication and
an inference model. While the former is conventional, the latter is inferential and
42
Anita Fetzer
hence pragmatic by definition. Sperber and Wilson define inference as follows:

Inference is the process by which an assumption is accepted as true or probably
true on the strength of the truth or probable truth or other assumptions. It is thus a
form of fixation of belief (Sperber and Wilson 1995: 68). Their theory uses demonstrative inference as is employed in various types of logic, and non-demonstrative inference, which is employed in the process of inferential comprehension.
Unlike local deductive reasoning, inferential comprehension is global and based
on inductive rules.
Pragmatic inferencing occurs at various levels of comprehension. It is connected with the differentiation between explicature, which is a combination of linguistically encoded and contextually inferred conceptual features (Sperber and
Wilson 1995: 182), and different kinds of implication, viz. trivial implication, analytic implication, synthetic implication and contextual implication. Against this
background, implicature is a contextual assumption or implication which a
speaker, intending her utterance to be manifestly relevant, manifestly intended to
make manifest to the hearer. We will distinguish two kinds of implicatures: implicated premises and implicated conclusions (Sperber and Wilson 1995: 194195).
Implicated premises must be supplied by the hearer, who must either retrieve them
from memory or construct them by developing assumption schemas retrieved from
memory, and implicated conclusions are deduced from the explicatures of the
utterance and the context. What makes it possible to identify such conclusions as
implicatures is that the speaker must have expected the hearer to derive them, or
some of them, given that she intended her utterance to be manifestly relevant to the
hearer (1995: 195). With these tools, Relevance Theory explicates the relevant inferencing processes involved in communication. They delimit themselves from the
Gricean paradigm where, from their perspective, the successes of human non-demonstrative inference must be explained by appealing not to logical processes of
assumption confirmation, but to constraints on the formation and exploitation of assumptions (1995: 81), viz. the cooperative principle and the maxims.
Grice (1975: 4344) differentiates between implicate and the related nouns implicature (cf. implying) and implicatum (cf. what is implied). He distinguishes between two basic types of implicature: conventional implicature and conversational
implicature. The latter is subdivided into generalized conversational implicature
and particularized conversational implicature. Generalized conversational implicature is also referred to as default implicature or pragmatic regularities (Bach
2006). This is in line with Levinsons claim that utterance-types carry generalized
implicatures []: rational speakers meannn both what they say (except in non-literal uses of language) and what that saying implicates; different layers of meaning
all come under the umbrella of meaningnn (Levinson 2000: 373). While conventional implicature is connected closely with linguistic form, for instance with connectives (e.g., but), implicative verbs (e.g., manage, forget to), honorifics or nonrestrictive relative clauses, conversational implicatures are essentially connected
43
with certain general features of discourse, viz. the Gricean conversational maxims.
Both conversational and conventional implicatures can be suspended: conventional implicatures are detachable, but not cancellable and conversational implicatures are cancellable but not detachable (for a thorough analysis of implicature, cf.
Huang this volume).
Conversational implicatures are generated or triggered by model users exploiting a maxim. That is to say, a model user gets in a conversational implicature if s/he
flouts a maxim, blatantly failing to fulfil it. Grice (1975: 4950) characterizes the
notion of conversational implicature as follows:
A man [or person, A.F.] who, by (in, when) saying (or making as if to say) that p has implicated that q, may be said to have conversationally implicated that q, PROVIDED THAT
(1) he is to be presumed to be observing the conversational maxims, or at least the cooperative principle; (2) the supposition that he is aware that, or thinks that, q is required
in order to make his saying or making as if to say p (or doing so in THOSE terms) consistent with this presumption; and (3) the speaker thinks (and would expect the hearer to
think that the speaker thinks) that it is within the competence of the hearer to work out,
or grasp intuitively, that the supposition in (2) IS required.
The conversational implicature is characterized by some degree of indeterminateness because it is, by definition, defeasible, non-detachable, calculable and nonconventional (cf. Bach 2006; e.g., Levinson 1979, 1983). Moreover, conversational implicatures are universal and reinforceable (Ariel 2008: 13). That is to say,
they can be asserted explicitly without bringing about a redundancy effect. It needs
to be pointed out, however, that the universality claim only holds for the inferential
mechanism and not for the relevant background assumptions.
Implicature is a cognitive mechanism anchored to model users in their roles as
speakers or producers of conversational contributions, and inference is a cognitive
mechanism anchored to model users in their roles as hearers. Cognitive processes
can be local, they can be global or they can be both. Not only natural model users,
but also non-natural intelligent agents may perform the cognitive operations of implicating and inferring, thus drawing invited inferences. For the operations to be
felicitous, compatible contexts for reasoning and drawing analogies are required
(cf. e.g., Bouquet et al. 1999; Cohen, Morgan and Pollack 1992).
In Gricean pragmatics, the conversational maxims are seen as specifications of
some unmarked communicative context representing the ground on which the conversational implicature is calculated. For this reason, deviations, however common, are seen as special or marked and signify that the speaker intends to communicate conversationally implicated meaning. The signalling is only possible
against the background of the Gricean premise of a reflexive intention, which is
succinctly explicated by Levinson: [w]hat distinguishes a Gricean reflexive intention from other kinds of complex reflexive intention is that the communicators
goal or intention is achieved simply by being perceived: recognition exhausts or
realizes the intention (Levinson 1995: 228).
44
Anita Fetzer
5.
Outlook
Neither general pragmatics nor linguistic pragmatics examine its objects of investigation in isolation but rather focus on their conditions of use, the connectedness
with their surroundings, and the necessary and sufficient conditions which assign
the object, e.g. intentionality, rationality, model user or action, the status of a particular object and make it count as that object. While general pragmatics concentrates on the analysis of these fundamental premises of practical action, identifying
their necessary and sufficient conditions, linguistic pragmatics establishes the explicit connection between those foundations and their language-specific and language-use specific constraints and requirements.
Pragmatics is more of a perspective towards an object under investigation than
the examination of the object as such. For this reason, it needs to touch on and interface with neighbouring disciplines, in particular philosophy, cognitive science,
neuroscience, linguistics and the social sciences. In spite of the diversity of the
field, the key research question of pragmatics is concerned with communicative
action, particularly with the expression and interpretation of meaning in context.
To tackle that question felicitously, pragmatics needs to accommodate extra-linguistic world knowledge, cultural and social stereotypes, and situation of discourse on the one hand, and word meaning, sentence structure and the cognitive
system, especially inference and abduction on the other. The generalized knowledge is stored as typed world knowledge and typed discursive knowledge. These
defaults represent pillars, against the background of which the argument from ignorance holds:
The argument from ignorance, it thus emerges, is an integral part of the inferential
mechanism by means of which implicatures are recovered in conversation. This argument permits us to conclude that conversational principles and maxims are being observed on the grounds that there is no reason to believe that they are not being observed.
Such a conclusion is at all possible because these principles and maxims have the status
of presumptions in communication principles that stand in the absence of counterindications. (Cummings 2005:109)
Presumptions or defaults concern both linguistic and extra-linguistic knowledge,

which need to be stored as social and cultural world-knowledge defaults on the one
hand, and as linguistic and discursive defaults on the other. Only then is it possible
to account for a pragmatic perspective which accommodates the inherent dialogic
and pragmatic principles of speech act and uptake, speech act type and cognitive
reality, and speech act type and situation (mis)match, which is calculated against
the background of the argument of ignorance. If a conversational contribution is in
accordance with the presumptions, no counterindications are found. Consequently,
the contribution is assigned the status of a match and interpreted accordingly. If a
conversational contribution is in disaccordance, there are counterindications and a
particularized process of inferencing is generated to calculate nonce meaning.
45
The dialogue-oriented pragmatic mechanisms of inferencing and abduction

require not only a dynamic framework accounting for the parts-whole connectedness, but also a conscious model user who is accountable for her/his actions (Fetzer
2004). At the same time, s/he makes use of automatic processing and default interpretations which figure as salient and strong interpretative probabilities, unless
counterindications in the context signify a non-default match. Only an integrated
framework, making allowances for methodological compositionality and synthesizing relevant findings (Fetzer 2004; Jaszcolt 2005; Marmaridou 2000) may capture the multifaceted nature of pragmatics.
Notes
1. I am grateful to the anonymous reviewer and the editors of this volume for helpful comments on the first version of this article, and I would like to thank Varol Akman and Teun
van Dijk for discussing previous versions of this chapter.
2. The seemingly clear-cut distinction between semantics and pragmatics, though useful for
analytic reasons, is highly idealized and has been strongly contested (cf. Bublitz 2009).
References
Aijmer, Karin
1997
I think an English modal particle. In: Toril Swan and Olaf Jansen (eds.), Modality in Germanic Languages. Historical and Comparative Perspectives,
147. Berlin: de Gruyter.
Akman, Varol, Paolo Bouquet, Richmond Thomason and Roger A. Young (eds.)
2001
Modeling and Using Context. Third International and Interdisciplinary Conference, CONTEXT 2001, Dundee, UK, July 2001, Proceedings. Springer: Heidelberg.
Ariel, Mira
2008
Pragmatics and Grammar. Cambridge: Cambridge University Press.
Auer, Peter and Aldo di Luzio (eds.)
1992
The Contextualization of Language. Amsterdam: Benjamins.
Austin, John L.
1971
How to Do Things with Words. Cambridge: Cambridge University Press.
Bach, Kent
1992
Communicative intentions, plan recognition, and pragmatics: comments on
Thomason and Littman and Allen. In: Cohen, Philip R., Jerry Morgan and
Martha E. Pollack (eds.), Intentions in Communication, 389400. Cambridge,
MA: MIT Press.
Bach, Kent
2006
The top 10 misconceptions about implicature. In: Betty Birner and Gregory
Ward (eds.), Drawing the Boundaries of Meaning. Neo-Gricean Studies in
Pragmatics and Semantics in Honour of Laurence R. Horn, 2130. Amsterdam: Benjamins.
46
Anita Fetzer
Bar-Hillel, Yehoshua
1971
Out of the pragmatic wastebasket. Linguistic Inquiry 2: 401407.
Bateson, Gregory
1972
Steps to an Ecology of Mind. Toronto: Chandler Publishing Company.
Beaugrande, Robert de and Wolfgang Dressler
1981
Einfhrung in die Textlinguistik. Tbingen: Niemeyer.
Birner, Betty and Gregory Ward
1998
Information Status and Noncanonical Word Order in English. Amsterdam:
Benjamins.
Birner, Betty and Gregory Ward (eds.)
2006
Drawing the Boundaries of Meaning. Neo-Gricean Studies in Pragmatics and
Semantics in Honour of Laurence R. Horn. Amsterdam: Benjamins.
Blackburn, Patrick, Chiara Ghidini, Roy M. Turner, and Fausto Giunchiglia (eds.)
2003
Modeling and Using Context. 4th International and Interdisciplinary Conference, Context 2003, Proceedings. Heidelberg: Springer.
Bouquet, Paolo, Patrick Brezillon, Massimo Benerecetti, Francesca Castellani and Luciano
Serafini (eds.)
1999
Modeling and Using Context. Second International and Interdisciplinary Conference on Modeling and Using Context, Context99, Trento, Italy, September
1999, Proceedings. Heidelberg: Springer.
Brandom, Robert B.
1994
Making it Explicit: Reasoning, Representing, and Discursive Commitment.
Cambridge, MA: Harvard University Press.
Brown, Penelope
1995
Politeness strategies and the attribution of intentions: the case of Tzeltal irony.
In: Esther Goody (ed.), Social Intelligence and Interaction, 153174. Cambridge: Cambridge University Press.
Brown, Penelope and Stephen C. Levinson
1987
Politeness. Some Universals in Language Usage. Cambridge: Cambridge University Press.
Bublitz, Wolfram
2006
It utterly boggles the mind: knowledge, common ground and coherence. In:
Hanna Pishwa (ed.), Language and Memory: Aspects of Knowledge Representations, 359386. Berlin: de Gruyter.
Bublitz, Wolfram
2009
Englische Pragmatik. (2nd rev. ed.). Berlin: Erich Schmidt.
Carston, Robyn
2002
Thoughts and Utterances. The Pragmatics of Explicit Communication. Oxford: Blackwell.
Clark, Herbert H.
1996
Using Language. Cambridge: Cambridge University Press.
Cohen, Philip R., Jerry Morgan and Martha E. Pollack (eds.)
1992
Intentions in Communication. Cambridge, MA: MIT Press
Collins Cobuild Essential English Dictionary
1990
Glasgow and London: Collins.
Cummings, Louise
2005
Pragmatics: A Multidisciplinary Perspective. Edinburgh: Edinburgh University Press.
47
Duranti, Alessandro
1997
Linguistic Anthropology. Cambridge: Cambridge University Press.
Fairclough, Norman
1992
Discourse and Social Change. Cambridge: Polity Press.
Fetzer, Anita and Varol Akman
2002
Contexts of social action: guest editors introduction. Language and Communication 22(4): 391402.
Fetzer, Anita
2002
Communicative intentions in context. In: Anita Fetzer and Christiane Meierkord (eds.), Rethinking Sequentiality: Linguistics Meets Conversational Interaction, 3769. Amsterdam: Benjamins.
Fetzer, Anita
2004
Recontextualizing Context: Grammaticality Meets Appropriateness. Amsterdam: Benjamins
Fetzer, Anita
2007
Non-acceptances in context. Intercultural Pragmatics 4(4): 493520.
Garfinkel, Harold
1994
Studies in Ethnomethodology. Cambridge: Polity Press.
Givn, Talmy
1989
Mind, Code and Context. Hillsdale: Laurence Erlbaum.
Givn, Talmy
1993
English Grammar: A Function-Based Introduction. Amsterdam: Benjamins.
Givn, Talmy
2005
Context as Other Minds. Amsterdam: Benjamins.
Goffman, Erving
1981
Forms of Talk. Oxford: Blackwell.
Goffman, Erving
1986
Frame Analysis. Boston: Northeastern University Press.
Goodwin, Charles and Alessandro Duranti
1992
Rethinking context: an introduction. In: Alessandro Duranti and Charles
Goodwin (eds.), Rethinking Context. Language as an Interactive Phenomenon,
142. Cambridge: Cambridge University Press.
Grice, Herbert Paul
1975
Logic and conversation. In: Peter Cole and Jerry L. Morgan (eds.), Syntax and
Semantics, 4158. New York: Academic Press.
Grosz, Barbara J. and Candace L. Sidner
1992
Plans for discourse. In: Philip R. Cohen, Jerry Morgan and Martha E. Pollack
(eds.), Intentions in Communication, 417444. Cambridge, MA: MIT Press.
Gumperz, John J.
1992
Contextualization and understanding. In: Alessandro Duranti and Charles
Goodwin (eds.), Rethinking Context: Language as an Interactive Phenomenon, 229252. Cambridge: Cambridge University Press.
Gumperz, John J.
1996
The linguistic and cultural relativity of inference. In: John J. Gumperz and
Stephen C. Levinson (eds.), Rethinking Linguistic Relativity, 374406. Cambridge: Cambridge University Press.
48
Anita Fetzer
Gumperz, John J.
2003
Response essay. In: Susan L. Eerdmans, Carlo L. Prevignano and Paul J. Thibault (eds.), Language and Interaction. Discussions with John J. Gumperz,
105126. Amsterdam: Benjamins.
Habermas, Jrgen
1998
On the Pragmatics of Communication. Cambridge, MA: MIT Press.
1994
Introduction to Functional Grammar. London: Arnold.
Hanks, William F.
1996
Language form and communicative practices. In: John J. Gumperz and
Heritage, John
1984
Garfinkel and Ethnomethodology. Cambridge: Polity Press.
Horn, Laurence R.
1984
Toward a new taxonomy for pragmatic inference: Q-based and R-based Implicature. In: Deborah Schiffrin (ed.), Meaning, Form and Use in Context: Linguistic Applications, 1142. Washington: Washington University Press.
Horn, Laurence R.
1989
A Natural History of Negation. Chicago: Chicago University Press.
Horn, Laurence R. and Gregory Ward (eds.)
2004
The Handbook of Pragmatics. Oxford: Blackwell.
Huang, Yan
2007
Pragmatics. Oxford: Oxford University Press.
Jaszczolt, Kasia
1999
Discourse, Beliefs, and Intentions: Semantic Defaults and Propositional
Attitude Ascription. Oxford: Elsevier Science.
Jaszczolt, Kasia
2005
Default Semantics: Foundations of a Compositional Theory of Acts of Communication. Oxford: Oxford University Press.
Lakoff, Robin
1973
The logic of politeness: or minding you ps and qs. Papers from the 9th Regional Meeting of the Linguistic Society of Chicago: 292305.
Lambrecht, Knud
1994
Information Structure and Sentence Form. Topic, Focus and the Mental Representation of Discourse Referents. Cambridge: Cambridge University Press.
Leech, Geoffrey
1983
Levinson, Stephen C.
1979
Activity types and language. Linguistics 17: 365399.
1983
Pragmatics. Cambridge: Cambridge University Press.
1988
Putting linguistics on a proper footing: explorations in Goffmans concepts of
participation. In: Paul Drew and Anthony Wootton (eds.), Erving Goffman. Exploring the Interaction Order, 161227. Cambridge: Cambridge University
Press.
49
1995
Interactional bias in human thinking. In: Esther Goody (ed.), Social Intelligence and Interaction, 221260. Cambridge: Cambridge University Press.
2000
Presumptive Meanings: The Theory of Generalized Conversational Implicature. Cambridge, MA: MIT Press.
Lewis, David
1979
Scorekeeping in a language game. Journal of Philosophical Logic 8: 339359.
Longman Dictionary of Contemporary English
2003
Harlow: Pearson Education Limited.
Marmaridou, Sophia
2000
Pragmatic Meaning and Cognition. Amsterdam: Benjamins.
Mey, Jacob
2001
Morris, Charles
1971
Writings on the General Theory of Signs. The Hague: Mouton.
Penco, Carlo
1999
Objective and cognitive context. In: Paolo Bouquet et al. (eds.), 2nd International and Interdisciplinary Conference on Modeling and Using Context
(Context99), 270283. Heidelberg: Springer.
Recanati, Franois
2004
Literal Meaning. Cambridge: Cambridge University Press.
Reddy, Michael
1979
The conduit metaphor. In: Andrew Ortony (ed.), Metaphor and Thought,
Sbis, Marina
2002
Speech acts in context. Language and Communication 22(4): 421436.
Schmid, Hans-Jrg
2003
Introduction: context and cognition a concern shared by cognitive sciences.
In: Ewald Mengel, Hans-Jrg Schmid and Michael Steppat (eds.), Anglistentag
2002 Bayreuth. Proceedings, 363367. Trier: Wissenschaftlicher Verlag Trier.
Searle, John R.
1969
Speech Acts. Cambridge: Cambridge University Press.
Searle, John R.
1975
Indirect speech acts. In: Peter Cole and Jerry L. Morgan (eds.), Syntax and
Searle, John R.
1979
Expression and Meaning: Studies in the Theory of Speech Acts. Cambridge:
Cambridge University Press.
Searle, John R.
1983
Intentionality. Cambridge: Cambridge University Press.
Searle, John R.
1991
Collective Intentions and actions. In: John R. Searle, Herman Parret and Jef
Verschueren (eds.), (On) Searle on Conversation, 401415. Amsterdam: Benjamins.
Searle, John R.
1995
The Construction of Social Reality. New York: The Free Press.
50
Anita Fetzer
Sperber, Dan and Deidre Wilson

1995
Relevance: Communication and Cognition. Oxford: Blackwell.
Stalnaker, Robert
1999
Context and Content. Oxford: Oxford University Press.
The New Encyclopaedia Britannica.
1976
The University of Chicago, Vol 14. Chicago: Helen Hemmingway Benton.
Thomason, Richmond H.
1992
Accommodation, meaning, and implicature: interdisciplinary foundations for
pragmatics. In: Philip R. Cohen, Jerry Morgan and Martha E. Pollack (eds.),
Intentions in Communication, 325363. Cambridge: MIT Press.
Traugott, Elizabeth Closs
1988
Approaches to Grammaticalization. Amsterdam: Benjamins.
Vanderveken, Daniel and Susumo Kubo (eds.)
2002
Essays in Speech Act Theory. Amsterdam: Benjamins.
Vanderveken, Daniel and Susumo Kubo
2002
Introduction. In: Daniel Vanderveken and Susumo Kubo (eds.), Essays in
Speech Act Theory, 121. Amsterdam: Benjamins.
Van Dijk, Teun
1981
Studies in the Pragmatics of Discourse. The Hague: Mouton.
Van Dijk, Teun
2008
Discourse and Context. A Sociocognitive Approach. Cambridge: Cambridge
University Press.
Verschueren, Jef
1999
Understanding Pragmatics. London: Arnold.
Watts, Richard, Sachiko Ide and Konrad Ehlich (eds.)
1992
Politeness in Language. Berlin: de Gruyter.
Widdowson, Henry
2004
Text, Context, and Pretext. Critical Issues in Discourse Analysis. Oxford:
Blackwell.
Wittgenstein, Ludwig
1958
Philosophical Investigations. Oxford: Blackwell.
Yule, George
1996
The Study of Language. Cambridge: Cambridge University Press.
2.

Piotr Cap
1.
Introduction
The present chapter discusses micropragmatics (the pragmatics of utterance based

concepts such as deixis, anaphora, presupposition, etc.) and macropragmatics (the
pragmatics of discourse based concepts such as speech events, global intentionality or macro speech acts). Its aim is to show how micropragmatics and macropragmatics (have come to) complement each other, both conceptually and in terms of
the division of labor across the field of contemporary linguistic pragmatics (henceforth: pragmatics).1
After a brief overview of the perspectivist approach to pragmatics in section 2.,
I turn to the discussion of context (section 3.) and discuss how its potential and
actual manifestations affect pragmatic inferences at different levels of the utterance and discourse structures. In 4., the perspectivist approach and the considerations of context are bridged to constitute two crucial factors defining micro- and
macropragmatics, in addition to several other factors and developments that have
led to such a distinction. In section 5., I deal with the four concepts central to the
methodology of pragmatics in general: deixis, presupposition, implicature and
speech acts. I discuss them from a micropragmatic perspective, showing their contribution to the proposition of an utterance, its illocutionary force, and the range
and kinds of effects the utterance is able to bring about. At the same time, however,
I introduce the macropragmatic approach, by raising three observations. The first is
the different potential each of the four concepts possesses to contribute to the description of function and effects of an utterance. The second is the fact that, in the
process of describing an utterance, some of the concepts go together and form
more sophisticated methodological tools, better than others. The third is the different range of utterance contexts each of the concepts is able to account for, which
opens up prospects for their hierarchical application in analysis. In 6., I take these
observations as prompts for an integrated study of intentionality at the macro level
of discourse. Addressing such notions as speech event and macro speech act, I
revisit the potential of individual speech acts for combining interacting discourse
topics into larger functional units. I discuss the conceptual infiniteness pertaining
to speech events in terms of their composition and hierarchy, and show the resulting relativity of the concept of the macro speech act.
52
2.
Piotr Cap
Pragmatics offers a unique, function-based perspective on all aspects of human

(linguistic) communication (cf. Verschueren 1999; Fetzer 2002). It is concerned
with all facets of a communicative act or a series of acts, such as the speaker, his/
her background knowledge and contextual assumptions, the lexical and grammatical constituents of an utterance, the hearers interpretations and patterns of inferencing, etc. All these are studied against a network of social factors, preconditions,
norms and expectations that govern communication, both within a culture and
across cultures. Since communicative acts involve linguistic units, whose choice is
dictated by language-internal rules, as well as their interpersonal, social and cultural embedding, pragmatic studies bridge the system and the use side of language.
They examine what is lexically and grammatically available for a speaker to accomplish a communicative goal, and at the same time explore the ways in which
the linguistic potential is realized in a specific social context.
Since the arsenal of language forms and the catalogue of social embeddings in
which these forms are used in communication are virtually infinite and heterogeneous, pragmatic scholarship naturally draws upon a number of diverse disciplines
in linguistics (and beyond). It is generated collectively by phonologists, morphologists, syntacticians, discourse analysts but also psychologists, sociologists and
other scientists in the Humanities as long as their work has a functional-communicative bent. Pragmatics is thus defined by its perspective more than by a set
number of its objects of investigation.2 It offers explanations which apply to different lexical and structural levels of language word, phrase, sentence, discourse.
3.
Context
The conception of pragmatics as a (functional) perspective foregrounds and endorses its preoccupation with context. Different types of context are naturally salient at different levels of communication (utterance, series of utterances, discourse, genre) but there are also manifestations of different aspects of context at
any specific level. Hence, context is a fundamental concept to account for in describing the micro-macro dialogue in pragmatic studies.
In a nutshell (cf. e.g. Bazzanella 2002), one can speak of a static and a dynamic conception of context. The static view (cf. e.g. Halliday and Hasan 2000;
Cummings 2005) sees context as a set of variables (linguistic, cognitive, social)
that surround strips of text. They constitute an a priori, static infrastructure of resources which are there for interlocutors to determine the meaning of utterances at
hand. On the dynamic view (cf. e.g. Duranti and Goodwin 1992; Bublitz 2003,
2006), context is never given and interlocutors keep creating current contexts for
current utterances the moment a sentence is uttered it becomes part of the envi-
53
ronmental resources on which the contextual interpretation of the following as well

as the preceding utterances has to draw (Bublitz 2003: 383).
As the next section will show in detail, the dynamic view of context is especially
important in elucidating the relationship between micropragmatics and macropragmatics the prerequisite being, of course, the move from a discourse-participant to
a discourse-analytic perspective. The complementarity of micropragmatics (pragmatics of the utterance) and macropragmatics (pragmatics of the discourse) consists in the micro-level analysis getting verified from the macro standpoint, a procedure often leading to the redefinition of the original analytic track. Text and context are, not only in interlocution but also in analysis, constantly matched with and
against each other, both prospectively and retrospectively. Any micropragmatic
analysis of an utterance, reshaped as a result of macropragmatic considerations of
the discourse to which the utterance contributes, is thus an example of reinterpretation from hindsight endorsed by the dynamic view of context. Such a reinterpretation is often indicative of the change in the number and kind of textual cues
which are used before and after the reshaping takes place. It is likely that in the
original micro-pragmatic analysis the reliance on the most prominent textual cues
will be significant, since the macro factors and constraints are, as yet, disregarded.
However, the reshaped micro-analysis may well see some of the initial cues
brushed aside, often for the benefit of the cues originally neglected (see section 4.).
All in all, the dynamic notion of context is a vital concept for micro- and macropragmatics since it allows speculations over, a) the extent to which the interpretation of meaning is guided by text; b) what factors may cause such an interpretation
to be insufficient; c) at which level of analysis such a deficit becomes visible.
The dynamic conception of context recognizes one more important premise for
the existence of the micro-macro dialogue, i.e. the different degrees of accessibility of contextual cues, at different stages of interpretation.3 The linguistic cues
(often referred to as co-text) are immediately accessible, but some of the cognitive and social cues might not be. This, altogether, makes discourse participants abstract from the holistic context (the context which is often defined at the macro
level, according to the characteristics of an attitudinal frame, genre, speech event,
discourse type) the cues that seem actually relevant to a given communicative situation (micro-level). The above process has been extensively described in terms of a
number of methodologies, which account for the relation between context and subcontexts (cf. Goffman 1986), figure and ground (cf. Langacker 1987, 2001), frame
and framing (cf. Bateson 1972; Fetzer 2007), generalized (default, unmarked) context and particularized (marked) context, global context and local context, etc.
Their common assumption is that the concept of context is best investigated from a
parts-whole perspective, which elucidates its function of delimiting content. Such
a perspective is naturally in line with the perspectivist view of pragmatics in general and, at the same time, congruent with the scope of micropragmatics and macropragmatics.
54
4.
Piotr Cap
Micropragmatics and macropragmatics: conceptual origins

and scope
As has been indicated before, micropragmatics can be defined as the study of illocutionary force at the utterance level. In contrast, the focus of macropragmatics is
not on the utterance, but on a series or sequence of utterances which form discourses/texts, seen as bearers of global intentionality of the speaker (i.e. the intentionality resulting from different speech act configurations, often referred to as
speech events) and as initiators of complex effects.
The distinction between micro- and macropragmatics follows from several factors as well as research developments. First, it has to do with flexible definitional
boundaries of pragmatics. If pragmatics involves the study of meaning in context,
then context is not limited in formal ways, such as the length of sentence/utterance
or the amount of its discourse history. Contextual considerations which are used for
determining the function and effect of an utterance may or may not stop within the
boundaries of the language form used. Thus, we need a conceptual handle on the interface between smaller and larger functional units of discourse (e.g. speech act
versus speech event), coded in smaller or larger forms (e.g. utterance versus text).
Second, since context is complex, dynamic and multilayered (cf. section 3.),
the accessibility of contextual cues is a matter of degree (cf. Marmaridou 2000).
The more access to contextual (including co-textual) embedding of an utterance or
a series of utterances, the better profiled the meaning. Hence a tendency on the part
of the analyst to pile up contextual cues and to add to the picture as much of the
prehistory of the discourse situation (cf. Mey 1993) as possible (and feasible).
But since access to contextual cues is each time different and invariably limited by
a dynamic network of constraints, the analyst can hardly ever claim to have exhausted all context to determine the meaning. In this sense, the micro-macro dichotomy involves a relation between the minimal and the maximal contextual
input into analysis, irrespective of its scope, i.e. the length of the language form
under scrutiny.
The domains of micro- and macropragmatics are complementary in terms of
their contribution to analytic labor. There is no micropragmatic analysis that would
not provoke a macropragmatic extension of scope; similarly, there is no macropragmatic study that would not question, retrospectively, its micropragmatic components, thus prompting revision or modification of the original analytic track. Altogether, we arrive at a bottom-top-bottom cycle of upgrades on the explanatory
power of both micropragmatic and macropragmatic concepts. This cycle involves
looking at different aspects of context (linguistic, cognitive, social) from two different perspectives, a more linguistic bottom-up perspective and a more social topdown perspective.
By way of illustration, consider a sample analytic procedure whose objective is
to account for the pragmatics of speeches of the American cold war presidents (say,
55
Harry S. Truman or Dwight D. Eisenhower). Imagine that a micro-level analysis of

utterances making up these speeches points to a large number of deictic, referential
and anaphoric markers, embedded in a multitude of direct speech acts. In contrast,
some other units or phenomena explored in such an analysis, for instance implicatures and presuppositions, turn out to be underrepresented. The next step is to classify these findings under controlling categories, such as speech events or macro
speech acts, to establish the performative denominators and thus postulate global
function(s) of the discourse genre (i.e. the presidential cold war speeches) as a
whole. This macro-analytic task requires inspecting the extralinguistic context,
which informs and complements the linguistic observations. The likely outcome of
the extralinguistic considerations is that the cold war presidents address a world
that is ideologically divided and respond to a need to clearly define to their audience the essence of this divide, which eventually leads them to construct the wellknown us and them opposition. Hence the rhetorical urge for directness and
clarity, rationalizing the opulent use of indexicals or anaphoric expressions and
banning most forms of implicit communication such as implicatures or presuppositions. Reaching this conclusion means that the analyst was able to use micropragmatic considerations as building blocks of a functional hypothesis, which he or she
then tested against contextual factors (including his or her expert knowledge), in
order to generate a macro-function of the genre. This is, roughly, how the analytic
curiosity about regularities governing the salience of some phenomena observed at
an utterance level can provoke the macropragmatic extension of scope. However,
the upgrade cycle does not stop here, there is still room for a reverse procedure,
which has the macropragmatic findings relate to the original component premises.
Accordingly, the analyst may want to return to the bottom level of an utterance to
look for more data, to further strengthen the macro conclusion(s). This may mean
leaving out some of the now useless forms (e.g. markers of implicitness) and
delving more deeply into those which have passed the extralinguistic context
verification. As a result, further micropragmatic activity can follow, with a view to,
either narrowing down the study of the original parameters, e.g. deixis or direct
speech acts, to only those aspects which are in line with the macropragmatic conclusion, or, possibly, identifying more bottom-level forms which are relevant, for
instance more cohesive devices (cf. Halliday and Hasan 1976), relational propositions (cf. Mann and Thompson 1983), etc. Which is where the first round of the
bottom-top-bottom cycle ends, but, of course, more rounds can still take place,
with the micro-macro conclusions refining each other virtually ad infinitum. The
existence of such a dialogue is probably among the most significant methodological arguments for having the two concepts, micropragmatics and macropragmatics,
in the theoretical framework of the field.
The distinction between micro- and macropragmatics has been strongly influenced by work in disciplines which pragmatics addressed while defining its
scope, and upon which it imposed its functional perspective. The ethnography of
56
Piotr Cap
communication (cf. Hymes 1974; Saville-Troike 1989; Gumperz 1992) has provided insights into communicative acts seen as socially situated cultural forms, and
micro-specifications of the more general social and linguistic contexts. Interactional socioliguistics (cf. Gumperz 2003) distinguishes between two levels of inference in analyses of interpretation processes. The higher level accommodates
global inferences of what the exchange is about and what mutual rights and obligations apply, what topics can be brought up and what reply is desired by the cultural macro-convention, and the lower level contains local inferences concerning
what is intended with the speakers own move and what is required by way of a response. Research in corpus linguistics (cf. Biber 1988) has adapted social and institutional factors to the examination of variation in language, both spoken and
written. It is argued (cf. Biber 1988) that conversational and discourse goals comprise both conventional social and institutional goals, as well as personal microgoals. Furthermore, the micro-macro interface has been at the heart of studies in
activity types (cf. Levinson 1979) and communicative projects (cf. Linell 1998).
These studies have paved the way for bridging the micro with the macro on the
plane of the explicit or implicit accommodation of context in interpretative processes. As it seems, the only way in which a distinction between the macro and the
micro domains can be attempted is indeed by looking into the different degrees of
lexical and grammatical coding of context in an utterance or a discourse vis a vis its
extralinguistic coding, i.e. the import of mostly non-indexical cues. Such analytic
endeavors are central to the most recent approaches to the analysis of communicative genres (cf. Bauman 1992; Luckmann 1995; Thibault 2003; Fetzer 2007).
Genre (recall our sample analysis above) is a classificatory concept; it specifies
typical ways in which lexicogrammatical resources are deployed to enact a communicative goal, by matching them against its own delimiting frame. Thus, micro
analysis is an analysis of individualistic phenomena (tokens), which macro analysis verifies against the generic requirements (type). The token contributions are not
merely cumulative, they can get encapsulated in one another or form a hierarchy
as will be shown in the next section.
5.
A micropragmatic perspective on deixis, presupposition, implicature

and speech acts with implications for macropragmatics
While the basic scope of micropragmatics is an utterance, any micropragmatic

analysis is, ultimately, a journey into the macro domain. This follows from the conceptual and methodological characteristics of the most common parameters utilized in description: deixis, presupposition, implicature, speech act. They are never
static components of the analytic enterprise; instead, they prompt an analysis
which is essentially cumulative, interactional and hierarchical. At the bottom of the
hierarchy is the input obtained from the interpretation of the lexical and grammati-
57
cal coding of the utterance (immediate) context. This input (from deixis, presupposition lexical triggers, etc.) is usually insufficient for the utterance interpretation
since it does not include non-textual cues. Accordingly, the analysis proceeds to
the study of inferencing (pragmatic presuppositions, implicatures), which bridges
the linguistic cues with the extralinguistic ones. Finally, the speech act level sees
the function of the utterance emerge from the premises and interpretations (speaker-, hearer-, content-oriented) that have accumulated. Such a function remains,
however, tentative, until a wider discourse context has been addressed for verification (viz. the sample analysis in 4.). Thus, the category of the speech act sits at
the methodological borderline between the micro and the macro domain, which the
latter involves generic, institutional, etc., characteristics and constraints. The
speech act is at the same time an umbrella category for deixis, presupposition, implicature, and, potentially, a component category in the macro considerations. Altogether, micropragmatics feeds sequentially into macropragmatics because the
consecutive-collective use of micropragmatic tools in analysis entails a continual
extension of scope, which eventually straddles the original utterance boundaries.
The four subsections below unfold this sequence, showing, ultimately, the inherent
relativity of the micro-macro distinction.
5.1.
Deixis
The outset of micropragmatic analysis is marked by establishing a tangible, lexicogrammatical architecture of the utterance, and especially the explicit manifestations of the relation between the words used and the context. The analysts task is
to identify the lexicalized pointers which indicate who utters the words to whom
to accomplish what, when and where. This task endorses the status of deixis as the
initiator of the (micro)pragmatic analysis since deixis is the concept that captures
the relationship between the language form and the context in the most evident and
direct manner. It can be viewed as the main phenomenon whereby features of context are encoded in utterances by primarily lexical (e.g. demonstratives) but also
grammatical (e.g. tense) means.4
While the analysis of deixis sets the stage for a more complex scrutiny of the
utterance, its potential to interact with the other parameters of description (presupposition, implicature) is limited. As a concept involving lexicalized or grammaticalized forms of expression, it makes a relatively small contribution to the illocutionary force of the utterance, especially in implicit communication. Naturally,
there are exceptions. For instance, deictic projections (the cues are processed not
from the speakers own perspective, but from the perspective projected on the
hearer, cf. Levinson 1983; Fillmore 1997; Marmaridou 2000; Manning 2001) enable deictic expressions to generate implicatures. If, at 4:01 p.m. on Friday, I send
an e-mail to my publisher which says:
58
Piotr Cap
(1) I will submit the manuscript on Saturday.

I have good reason to expect the publisher (who works until 4 p.m. and only on
weekdays) to read the e-mail no earlier than Monday and, seeing no attachment
containing the ms. in the mailbox, to be puzzled did I mean the Saturday that has
just passed or the one to come? Of course, I retain the rhetorical comfort of denying
whichever interpretation. In fact the comfort starts the moment the Friday e-mail
is sent out I can feel free to submit the manuscript on whichever of the two Saturdays I prefer. There is no way I can be blamed for not submitting the ms. on the
first of the two Saturdays, as I can always respond that I was taking my addressees
perspective (i.e. receiving time), not mine.
Most types of deixis (person, time, place, for an overview cf. Hanks this volume) are analyzed at the utterance level. However, a notable exception is discourse
deixis, which crosses the utterance-discourse boundary, thus becoming a macropragmatic worktool. Discourse deixis involves the use of a lexical item within an
utterance to point to the preceding or following utterances in the same discourse
situation (speech event). The backward and forward reference can be illustrated by
(2) and (3), respectively, where already refers to the earlier stretch of discourse and
here anticipates the upcoming stretch:
(2) As already indicated, all languages possess deictics.
(3) Here goes my argument.
There is no rule to how much textual distance should hold between the deictic expression and its referent. While it is reasonable to expect that the referent of here
will be the immediately following utterance(s), the referent of already is surely not
the closest, preceding utterance, but rather an utterance made much earlier in the
unfolding discourse. This brings us to an important conclusion regarding the pragmatics of discourse deixis. By using a specific number of deictic expressions, the
speaker is able to control the overt connectedness of discourse and, in consequence, its comprehension by the hearer. The presence of deictic markers in utterances which make up a discourse situation where a specific topic is pursued
usually contributes to explicitness and clarity. On the other hand, the speaker may
choose to withhold the use of deixis to purposefully obscure the message.
These exceptions (deictic projections contributing to implicatures; discourse
deixis as controllers of discourse comprehension), however important, do not detract from the core function of deixis in pragmatic studies. Deixis is an essentially
micropragmatic phenomenon which sets analysis of the utterance in motion, but,
by itself, adds little to the characteristics of the speech act(s) residing in it. The
deictic framework of the utterance is sketched to establish the basic, lexically encoded relations between the referent(s) of the utterance and the common ground
of knowledge the utterance assumes to exist between the speaker and the hearer.
Yet, a full account of these relations (and of the utterance function(s) the relations
59
contribute to) needs complementation from concepts that go beyond lexically encoded relations (e.g. presupposition, implicature).
5.2.
Presupposition
Presupposition can be defined as a mechanism whereby the speaker addresses a

body of knowledge and experience, involving both linguistic and non-linguistic
contexts, which he or she assumes to be common to him-/herself and the hearer.
The assumption of the existence of the shared knowledge may cause the speaker
not to grammaticalize (or lexicalize) it in the utterance. This characterization takes
presupposition to be a phenomenon lying at several intersections: the encoded and
the assumed, the semantic and the pragmatic, (or even) the linguistic and the nonlinguistic.5
Presupposition comes in contact with deixis on the plane of its partial anchoring in lexical and structural forms. However, since many instances of presupposition can only be approached with reference to (non-linguistic) context, presupposition also reaches out in the direction of the implicit, constituting, in a sense, a
shared knowledge prerequisite for communicating messages whose final destination is their inference by the hearer. Hence its feasible combination with the apparatus of implicature and, altogether, its relevance to the hierarchy of micropragmatic analysis, which derives its output from both accumulation and interaction of
descriptions offered by the individual conceptual tools. As one of the latter, presupposition targets the communicative act at the stage where it develops upwards
from the lexicogrammatical coding of context to its further abstraction and elaboration by the speaker, with a view to producing a speech act. Throughout this
stage, the speaker decorates the deictic framework of the utterance with instantiations of knowledge shared by the speaker and the hearer with regard to all entities
indexed, referred to, or implied in the utterance.
Traditionally, the more a presupposition was linked with a lexical item or a linguistic construction generating it, the more it was treated as a semantic phenomenon, the other cases deemed pragmatic and worth less attention precisely because of the absence of fixed language forms responsible for enacting particular
presuppositions.6 This view has produced multiple typologies of presupposition,
based on its embedding in lexicogrammatical forms called presupposition triggers.
Furthermore, a number of properties have been assigned to presuppositions, including cancellability (a possibility of denying a presupposition, usually by adding
more content to the utterance where it occurs) and constancy under negation (negating the predication carrying a presupposition does not detach or change the presupposition) (cf. e.g. Beaver 2001).
From the perspective of micropragmatic analysis oriented toward the speech
act characterization of the function of the utterance, as well as the macropragmatic
perspective of the discourse, a rigid distinction between semantic and pragmatic
60
Piotr Cap
presupposition seems far from necessary. It offers little explanatory power compared to an integrated, global view of presupposition as a concept which should be
studied against the utterance and discourse goals it serves. Such a view is quite naturally pragmatic, because even though there are specific lexical items (in fact,
many) associated with specific assumptions (for instance, the assumptions of existence), their descriptive capacity does not expire within the structural boundaries of
the linguistic expression. On the contrary, their significance goes much beyond as
they are able to combine with primarily experiential premises and thus successfully contribute to a network of contextual, often non-linguistic beliefs making up
the entire load of knowledge shared by the speaker and the hearer (see fn. 5).
Most communicative goals served by presupposition (relevant to both the
micro- and macropragmatic analysis) have to do, in one way or another, with economy of expression, though the latter is rarely the only goal sought. If a speaker
could not rely on shared assumptions, the lexical and grammatical load of his/her
utterance would grow in size, potentially obstructing its comprehension. Although
economizing on the linguistic form for the benefit of unobstructed communication
could be an utterance goal in itself, it is often a contribution to a larger utterance or
discourse function. This becomes clear when analyzing, at the macro level, a series
of utterances containing consecutive acts of adjustment to shifting discourse expectations. If I am overweight and say I started jogging after visiting my doctor
and my confession meets with a blatant attack like So it took you going to the doctors to work out like you should, I can always follow a defense line in continuing
Well, to be honest, I tried to do some jogging a few times before, but now I do it
regularly. Since the short form of the initial utterance makes it underdetermined
in terms of meaning, the denial of the only then and never before presupposition
comes rather easy, contributing to the overall explanation and justification (see
Huang 2007 on cancellability of presuppositions).
5.3.
Implicature
The relevance of the analysis of implicature7 to micropragmatics (and, potentially,

macropragmatics) is twofold. First, implicature studies account for the further processing of information that has been encoded by the speaker based on his/her presuppositions. Second, in doing the latter, they eventually recognize the contribution implicature makes to the update of the utterance-discourse context.
From the analytic standpoint, the explanatory powers of presupposition and
implicature are inherently complementary, shedding light on both the speaker and
the hearer side of the speech act formation. They are suited to cover, in combination, the whole process of encoding messages by speakers and decoding them by
their hearers. This process is essentially a continuum, where making a presupposition paves the way for the utterance before it takes on a linguistic form, in which
the presupposition is lexically or non-lexically salient. From that point on, i.e. the
61
point of making the utterance by the speaker, the recovery of the implicature(s) by
the hearer may begin of course, if the hearer senses a prompt to search for it/
them. The inference of the implicature(s), whether in accordance with the speakers expectations or not, finalizes the entire process, thus updating the status of the
interaction and creating a new contextual basis on which to build presuppositions
for further utterances in the exchange. The cycle in question corroborates the dynamic view of context (see section 3.) and endorses the intrinsic relativity of the
micro-macro dichotomy. While the update takes place, technically speaking,
within the utterance, its effect is on the prospective discourse.
Delving deeper, implicatures created within the boundaries of the utterance
(the micro level) are often returned to or readdressed purposefully later on in
the unfolding discourse (the macro level). Since implicature is rarely encoded in
language form,8 it involves a virtually indeterminable number of more or less complex contextual inferences. As such, it constitutes a valuable rhetorical tool
whereby the speaker can control the flow of discourse, adopting his or her consecutive utterances to the current goals. This is due to a central property of implicature,
cancellability (cf. Sadock 1978; Levinson 2000, etc.), which makes it possible for
the speaker to deny, at any moment of speech situation, any implicature he or she
apparently created. We have seen this property as partly relevant to presuppositions, but it is implicature that permits its broadest manifestation. Indeed, a great
many implicatures are cancelled to re-establish adherence to the conversational
norms (see the Cooperative Principle and the Maxims of Conversation by Grice
(1975, 1989) as well as their numerous reformulations and supplements9) but also
to play with the addressee, pull a trick on him/her, or simply annoy him/her. Following on this note, many implicatures are cancelled for ironic or sarcastic effects,
which is well documented in humor studies (cf. Attardo 1990, 1993). Another area
where implicatures and their cancellations surface, is public (especially political)
discourse. Implicatures contribute to the rhetorical safety of public speakers, who,
on the one hand, wish to make statements that are universally acceptable, but on
the other, want to retain the possibility of refining or fine-tuning these statements for hearers who hold different expectations of the meanings conveyed in
them (cf. Cap 2008, 2010). Altogether, the phenomenon of the cancellability of implicature belongs to macropragmatics, since, first, the context that determines the
cancellations is made up of a heterogeneous number of social and institutional factors, second, the distance between implicature and its cancellation is a matter of
discourse, rather than utterance.
5.4.
Speech acts towards macropragmatics
The three brief subsections above have shown that deixis, presupposition and implicature make their distinctive micropragmatic contributions to understanding
how an utterance is built, what its referents are and how they are encoded, what as-
62
Piotr Cap
sumptions are made before the utterance is produced, what effects can be expected
after it has been produced and what inferential processes determine these effects.
They partake in the process of enacting goal(s) of the utterance, from the speakers
intention to realize its envisaged function via application of specific indicators of
force, to the hearers successful (or not) recognition of this function and its results.
A procedure this complex needs a controlling, umbrella parameter of description.
It needs a conceptual tool that is able to cover both speaker and hearer related aspects of the utterance function, and, while doing so, draw upon and thus systematize the particular contributions from deixis, presupposition, and implicature in
order to make them fit for macropragmatic work at the discourse level. The concept
of the speech act seems an excellent theoretical candidate to take up this task.10
The orientation of speech acts to both parties of a verbal exchange, as well as to
its linguistic matter, is visible at a glance from the traditional distinction between
the locutionary, the illocutionary, and the perlocutionary aspects of a speech act.11
While the locutionary aspect is the most objective since it concerns the stable language form of the utterance, the illocutionary and perlocutionary aspects involve a
dynamic negotiation of meaning between the speaker and the hearer. In saying Its
hot in here a speaker may be producing an (implicit) illocutionary act requesting
the hearer to open the window, and the perlocutionary act (effect) might be that the
hearer indeed opens it, but it might also be that he or she turns on the air-conditioning instead. Thus, the illocutionary-perlocutionary relation not only mirrors the
complex process of meaning evolution as sketched at the beginning of this subsection; it also inscribes in the distinction between explicit (direct) and implicit (indirect) ways of communicating a speech act. Consequently, it invokes the notions of
deixis, presupposition, and implicature, since they all situate themselves at some
specific yet different points of the conceptual axis which links what is said with
what is effected.
The classificatory, controlling power of the speech act is further reflected in its
network of felicity conditions, i.e. the conditions that underlie a successful, logical,
felicitous production of different acts.12 For example, a speaker cannot make a
successful order if he or she does not sincerely want the order to be followed, or if
he or she deems the hearer incapable of following it. These two felicity conditions
are excellent illustrations of the connection that holds between the concept of the
speech act and the other micropragmatic concepts a relation we have postulated
at the beginning of this subsection. The speakers awareness of cognitive and social context obtaining at the moment of producing a speech act gives rise to pragmatic presuppositions underlying the utterance that contains this act. Then, once
the act is accomplished, the speakers presuppositions can be assessed against the
effectiveness of implicatures they helped to create. While this part is naturally the
domain of implicature studies, we do get an extra insight from the cover framework of the speech act theory and, specifically, from its contribution to the research
in direct versus indirect acts.
63
Ironically, the greatest advantage of this research is that it almost removes one
of the two groups of acts from our scope of interest, at both the micro and the
macro level of description. As pointed out in several studies, most speech act
usages are indirect (cf. Bertolet 1994; Holdcroft 1994, etc.) and indeed, even the
apparently straightforward act of asking for salt in Can you pass the salt? is, formally speaking, indirect. This, however, should not be discouraging. The orientation of the speech act theory to indirect usages is only reflective of the visible emphasis on indirectness that is salient in the other (micro-)pragmatic concepts; after
all, our discussion of deixis, presupposition, and implicature has been mostly preoccupied with hidden meanings. Thus, speech act theory turns out to be nothing
but congruent with the other apparatuses. Moreover, it extends over all of their
fields of application, from the stage where knowledge is assumed to get encoded in
the utterance, to the stage where the utterance is interpreted. Conceivably, a single
speech act can be realized through recourse to an x number of presuppositions,
an x number of deictic markers in the utterance, which the latter could produce an
x number of implicatures. We have thus arrived at a micropragmatic hierarchy of
analysis, with the speech act constituting a category superordinate over the other
micropragmatic categories. At the same time, however, the speech act should not
be considered the top-most variable of description; in order to account for discourses, rather than individual utterances, we are in need of yet higher-rank concepts.
6.
Macropragmatics
Most speech acts residing in individual utterances tend to combine into larger functional units, thus paralleling the combination of utterances into texts and discourses. This process can be seen from the analysis of (4). Assume that instead of
reading out students names from the attendance list at the beginning of a seminar,
I simply ask:
(4) Hello, are we all here?
In uttering (4), I perform two direct speech acts (greeting and asking), as well as an
indirect act of requesting my students to reveal the names of the absentees. The
speech acts involve a deictic anchoring (e.g., here recognized as classroom), presuppositions (e.g., of some students still missing), and the indirect act involves a
process of inference. This is, with some simplification, where a micropragmatic
analysis of (4) can get us.
However, being myself the producer of (4), I can enjoy the privilege of stating
with absolute certainty that, most of the time, I do not open the class with this utterance for the mere sake of greeting, asking, and requesting, as described above.
That would be unlikely considering that (4) happens regularly and has thus become
64
Piotr Cap
(a part of) expectable routine, a pattern that arises, usually, from some kind of a
larger intention.13 Mark an analogy: If I follow the routine of buying each morning
a newspaper (an action which involves micro-actions analogical to the acts defined
in (4)), the larger intention can be described as a general aspiration to stay updated with the current news. The larger intention in (4) is to begin the seminar,
making sure all things are in place for a productive meeting. This finding does not
sound terribly original in itself, but we will take it as a starting point to show,
eventually, that the nature of intentionality is such that it cannot be a matter of the
individual utterance and thus the scope of inquiry must be extended accordingly.
More often than not, I may need to say more than just (4), to successfully begin
the seminar. I may choose from an apparently infinite catalogue of other utterances
which are able to serve the same intention. Some typical cases may be: Its getting
late, Would you close the doors please?, Now, Adam, listen up, will you?,
Right, shall we begin?, And the chalk is missing as usual!, I cant remember
when we last started on time, etc. All these seemingly diverse utterances contribute, in one way or the other, to realization of the principal intention. Interestingly, they do so while performing different speech acts (assertives, directives, expressives) whose force may also be different, direct or indirect. Consequently, each
of these acts may accommodate a different input from deixis, presupposition, and
implicature.
The routine of performing individual speech acts (which may in turn involve
the other micropragmatic categories) to realize a more general intention, can be defined as a speech event.14 Speech events may contain, as we have seen, a potentially
infinite number of utterances, depending on how many are actually necessary to
carry out the intention. If, for instance, my seminar group enters the classroom discussing a just-finished exam in which they participated, I may be in for a longer
stretch of talk to set up the stage for my own class. Otherwise, it may take a few
short remarks to cover the technicalities and initiate the topic proper. Whatever
happens in actuality is thus dictated, as anything in pragmatics, by context. A
speech event can comprise just as many utterances (and speech acts) as needed to
match the contextual preconditions. These utterances do not have to come in a
monologic pattern. I may keep performing the principal speech event by producing
a few utterances in a row, then pausing, then, possibly, reacting to a question that
comes in the meantime, then resuming the monologue, etc. Thus, my performance
is a genuine discourse performance, which exists in and responds to a social setting, though, of course, some parts of it can mirror textual units and their connectedness patterns.
The concept of infiniteness resurfaces once we go on to discuss speech events
as building blocks for realization of a still larger intention. To readdress the
case initiated by (4), the speech event which realizes the intention to successfully
begin the seminar is not only an umbrella category for a series of individual acts,
but may itself be subordinate to a larger discourse goal be it, for instance, con-
65
ducting a productive seminar meeting as a whole. This goal entails that an apparently infinite number of speech events are carried out on the way, from a speech
event of providing a complex, multi-act explanation to a problem that has arisen
during the session, to a speech event of assigning homework, which may again involve a number of component speech acts. The latter (event) may include: an expressive (e.g., reprimanding students for not completing the previous assignment),
a commissive (e.g., threatening to fail students at the end of the course, if they keep
neglecting their assignments), a directive (e.g., telling students to do a particular
task for the next meeting), an assertive (e.g., describing a rationale for the task),
etc. The variety and diversity of the acts involved is, here, no smaller than in the
case of the speech event performed to begin the seminar.
In this clearly bottom-up fashion, we have approached the problem of the uppermost or global category of intentionality enactment, one whose promise
would be to encompass all the subordinate intentions realized in speech events and
their attendant acts. Such a promise is partly fulfilled by the classical concept of the
macro speech act, which is, in van Dijks words, a global speech act performed by
the utterance of a whole discourse, and executed by a sequence of possibly different speech acts.15 Still, while doing some useful job in the way of systematizing
speech events, the macro speech act suffers from a problem of an inherent relativity as regards the range of its own operation. The definition of the macro act only
corroborates the problem; we do not get to know how much is a whole discourse.
Is it, to return to our example above, the whole body of discourse produced within
the duration of the seminar, controlled by the global intention to make it a productive meeting? Then the relationship between the macro speech act and the component speech events (and their individual acts) seems analytically elegant. But, does
this account exhaust the potential of the macro speech act to combine with further
macro speech acts, to serve a yet-more-global intention? Apparently not. The intention to carry out an academically rewarding seminar can be considered subordinate to the intention to conduct the entire course as planned, which in turn partakes in the intention to perform my (academic) duties properly as a whole, etc.,
etc., which, naturally, makes the consecutive macro acts accumulate accordingly.
To envisage the highest-rank intention, pursued in the highest-rank speech act,
turns, then, a philosophical undertaking.
One of the undisputed methodological values of the search for the clear-cut categories signposting the particular levels of intentionality and its enactment lies,
paradoxically, in the recognition of limits to which intentionality could be accounted for in larger stretches of discourse. Thus, endeavors like the macro speech
act theory should not be carelessly brushed off, as they eventually motivate research, however minimalist, in better demarcated and better empirically equipped
areas. Apart from the aforementioned research in genres, a prominent example of
such an area is Conversation Analysis.16 Conversation analysts have elaborated an
impressive arsenal of techniques for the description of speech act deployment,
66
Piotr Cap
though in necessarily limited contexts, often just co-texts. As a result, the apparatus of macropragmatics has been endowed with a number of relevant concepts,
such as floor, topic, turn, turn-taking, transition relevance place, etc., each
of which contributes its share of insight in how people manage their intentions and
goals in particularized (conversational) settings.
A related and very important significance is that a macro perspective on the
pragmatics of discourse makes the analyst approach intentionality as continually
re-shaped and updated by extralinguistic context and thus work out methodological tools to handle the variability of enactment of the speakers intention(s) over an
extensive period of time. Let us illustrate this phenomenon with an example from
political discourse, a domain rich in speakers attempts to continue with an enactment of a global intention, though often in consecutively modified patterns, responding to the changing extralinguistic reality. Consider an excerpt from one of
G.W. Bushs speeches legitimizing US involvement in the recent Iraq war:
(5) By advancing freedom in the greater Middle East, we help end a cycle of dictatorship and radicalism that brings millions of people to misery and brings
danger to our own people. By struggling for justice in Iraq, Burma, in Sudan,
and in Zimbabwe, we give hope to suffering people and improve the chances
for stability and progress. [] Had we failed to act, the dictators programs for
weapons of mass destruction would continue to this day. For all who love freedom and peace, the world without Saddam Husseins regime is a better and
safer place.17
This speech comes eight months into the war, which started on the assumption that
the Iraqi regime (and its alleged terrorist allies) had developed access to weapons
of mass destruction (WMD), thus becoming a world threat and unwilling to disarm
unless forced to. In his address, Bush attempts to maintain the aura of legitimization of the US intervention, against the increasingly evident collapse of the original premise no weapons of mass destruction were found in Iraq ever since the
US and the coalition troops entered it in March 2003. Since he can no longer employ the single WMD argument to execute his global intention (to keep the legitimization process intact), he deftly switches to a more universal, ideological
rationale. Instead of invoking a direct threat from a destructive impact of WMD, he
concentrates on building up an ideological framework for a potential growth of
such a threat in the future and presents the US strike as part of the necessary means
to ensure that the antagonistic ideologies (dictatorship, radicalism, regime)
do not materialize in the form of physical impact. The localization of these ideologies in more countries than Iraq itself (viz. Burma, Sudan, Zimbabwe) is
in fact a clever rhetorical ploy. By extending the spectrum of the ideological conflict and, thus, the spectrum of US foreign involvement, he encourages the construal of the WMD intelligence failure in terms of an isolated incident, fully justifiable given the range of the American mission as a whole. Altogether, Bush
67
keeps up with the global intention to continue with legitimization of the Iraq war,
now on ideological grounds, but the pragmatic patterns (and, consequently, lexical
choices) responsible for realizing this intention within a macro-temporal reality
need to be updated to meet the extralinguistic developments.
In section 4., I argued that macropragmatic analyses often lead to redefinitions
of the micropragmatic input that has been utilized to build up their tentative versions. They also tend to review the micropragmatic concepts and point to those
whose descriptive potential has not yet been fully exhausted. The analysis of
Bushs speech is a good case in point. At its macro-level, we consider the general
patterns of adaptation of the political speaker to contextual conditions. This leads
us to the more focused question which of the micro-level concepts are able to account for such processes in the possibly richest lexical way, i.e. which of them are
most frequently reflected in lexical items and structures that eventually make up
the matter of the macroanalysis. The analysis of (5) reveals, for instance, that much
explanatory power is yet to be drawn from a careful scrutiny of implicature forms
in the text. If we gather from macro-contextual considerations that Bush is forced
to switch to a new argumentative strategy, yet, as seems logical, without discrediting the previous argumentation, what better way to accomplish the goal than
through implied meanings, which are always subject to cancellation as the speaker
sees fit? Recall the phrase programs for weapons of mass destruction. It is flexible enough to concede that Saddam did not indeed possess WMD understood as a
product ready for use, at the outset of the war, but it does not completely detract
from the original assumption that he did. It is quite likely that a microanalysis of
this implicature (as well as of many other bottom-level forms in (5)) could be overlooked but for the prompt from the macropragmatic approach, involving a vast
range of contextual factors affecting the entire discourse of the Iraq war.
In this way we have returned to the question of the micro-macro dialogue,
which I defined in the fourth section of this chapter as one of the most significant
methodological arguments for having the [distinction between the] two concepts,
micropragmatics and macropragmatics, in the theoretical framework of the field.
It should be said in closing that prompts for such a dialogue need not come from
pragmatic analyses alone; in fact, the dialogic relationship between micropragmatics and macropragmatics is kept alive by insights from approaches which, at
least definitionally, go much beyond linguistics-based disciplines or perspectives.
An example of such an approach is Critical Discourse Analysis (CDA; cf. e.g. Fairclough 1995; Wodak and Chilton 2005), which extends onto domains traditionally
associated with various branches of sociology and social psychology. Rooted in the
conception that discourse is just as much reflective of the existing reality as it is potentially constitutive of a new reality, CDA offers a number of useful ways for the
analysis of the reality (which pragmatics would rather term extralinguistic aspects of context). In so doing, it provides (macro-)pragmatic considerations with
socio-cognitive grounding of discourse (cf. van Dijk 1995, 2002), thus signposting
68
Piotr Cap
the downward, top-down analytic procedure. Two examples of how such a procedure could operate have been, in this article, the suggested refinements and extensions of the analysis of deixis and direct speech acts in the genre of the cold war
presidents speeches, as well as the just discussed hint for a fully fledged study of
implicature in (5). Of course, CDA has much more to offer, as does any approach
or discipline that recognizes the paramount importance of (holistic, dynamic) context in human (linguistic) behavior. In whatever kind of acting, including pragmatic acting, we adapt ourselves to a context as is, but at the same time we
change it with a view to future effects. As Mey (1993: 263) writes, all this is done
through the intermediate use of language as the tool helping us to select the relevant features of any situation in the total context. Micropragmatics and macropragmatics are two perspectives which, on the one hand, statically account for the
different size manifestations of this total, or holistic, context. On the other
hand, they dynamically inform each other (and thus the pragmatic analysis as a
whole) of, first, which instances of the utterance-level use of language are worth a
broader contextual verification to determine their function, second, which parts of
the broader contextual check are applicable as determiners of further analytic activity at the utterance level.
Notes
1. See Fetzer (this volume) for the distinction between general (non-linguistic) pragmatics
and linguistic pragmatics.
2. See Fetzer (this volume) or Huang (2007) for a discussion of perspectivist versus componential views of pragmatics.
3. Cognitive pragmatics (e.g. Sperber and Wilson 1986) illustrates this by the onion metaphor. The pattern of inclusion of the individual layers, which are interconnected, reflects
their order of accessibility in the process of inferencing.
4. Foundations of research in deixis are presented in Bhler (1934/1982). Among the most
useful modern overviews are Nunberg (1993), Fillmore (1997), and Levinson (2004); cf.
also Hanks (this volume).
5. The standard view of presupposition (see Huang this volume) is that (i) it involves a proposition whose truth is taken for granted in the utterance of a sentence and remains in force
even if the sentence is negated, and that (ii) it is engendered by certain lexical and structural triggers. This definition goes back to philosophical and semantic views of Frege
(1892) and Strawson (1952), later enriched with semantic-pragmatic analyses (cf. Karttunen 1973; Stalnaker 1973; Gazdar 1979). Though these characterizations of presupposition are still generally accepted (cf. Saeed 2009), many linguists have pointed to the
existence of counter-examples deeming the mainstream approach (necessarily) imperfect (cf. Huang this volume). In the present chapter, presupposition is treated as a phenomenon which does not rely for its existence on linguistic form alone, but on a complex
relation between the linguistic form, the speakers and the hearers (common) understanding of the world (cf. Grundy 1995: 78), as well as the speakers fulfillment of
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
69
felicity conditions for performance of an act involving presupposition(s). This stance

integrates the standard view with several other research perspectives: the speech act approach to presupposition (cf. Grice 1981; Green 1989), the cognitive approach (cf. Carston 1996), the experiential approach (cf. Marmaridou 2000), or the common ground
approach (cf. Grundy 1995).
Examples of battles over the status of presupposition are Stalnaker (1973), Soames
(1989); while Horn (1996), Atlas (2004) and Huang (2007) offer well-balanced overviews.
In the present chapter the use of the term implicature is synonymous with conversational implicature. See Huang (this volume) for a comprehensive account of various
types of implicature.
We do not deal with conventional implicatures here (cf. e.g. Davis 1998; Bach 1999).
See reductionist attempts such as Horns (1984) Q-R model or Levinsons (2000) Q-I-M
principles; for supplements and general overviews see Levinson (2000), Horn (2007,
2009), Huang (2009a, b and this volume).
See Austin (1962/1975) and Searle (1975, 1979) for the foundations of the Speech Act
Theory. Tsohatzidis (1994), Thomas (1995), Marmaridou (2000), and Sadock (2004)
are excellent, objective overviews of both the classical ideas and later developments.
See also Collavin (this volume).
See Austin (1962/1975) and a review of Austins ideas in e.g. Alston (1994).
Cf. Austins (1962/1975) account of felicity conditions; see also Searle (1979).
Despite a temptation to the contrary, I save the term global for the discussion of macro
acts. I refer to this intention as larger not just to say that it is more indirect than the
intentions to greet, grasp students attention, or work out the number of absentees. My
point is that all of the latter can accumulate into a bigger intentional schema and that
there is no limit to the number of micro-intentional components making up such a
schema. Hence the use of the comparative larger, which allows for a downward
specification (i.e. the specification of what the larger intention is supposed to comprise), as well as an upward specification (i.e. the specification of what higher or highest level of intentionality it could contribute to). Interestingly, van Dijks (1980)
speech event approach to intentionality levels, as well as the activity type and genre
oriented approaches to intentions and context (cf. Levinson 1979; Bauman 1992; Luckmann 1995; Thibault 2003; discussed in section 4), are far from elucidating this bi-directional possibility, which is an important element of the micro-macro interface.
See van Dijk (1980) and Mey (1993). Levinson (1979) uses the term activity type. The
most frequent, however, is the term genre (cf. fn. 13). Recall (section 4) that [genre]
specifies [multiple] typical ways in which lexicogrammatical resources are deployed to
enact a communicative goal, so even such short one-utterance forms as (4) apply, since
the ways in which they could be supported by other forms (recall from the above the potentially infinite catalogue of expressions such as its getting late!) are controlled by
common, conventional, and thus generic constraints. Unlike speech events, which combine into macro speech acts (cf. van Dijk 1980), genres do not presuppose a methodologically superordinate concept. Thus, they constitute a more elegant framework in
terms of their downward specification, but fall short of describing the infiniteness of
consecutive layers in the whole intentionality structure.
See mainly van Dijks work, from classical contributions to text grammar (cf. van Dijk
1977, 1980), to recent research in critical discourse analysis (cf. e.g. van Dijk 2008).
70
Piotr Cap
See also his joint work with Walter Kintsch (e.g. van Dijk and Kintsch 1983), which in
my view represents the best step ever taken to integrate the concepts of discourse and
text in a pragmatic analysis.
16. Its protagonists include Emanuel Schegloff, Harvey Sacks, Gail Jefferson, Jonathan
Potter; cf. Ten Have (2007) for an overview.
17. The Whitehall Palace address, November 19, 2003. The analysis of (5) shows that macropragmatic studies often proceed in a top-down manner (cf. section 4); they draw on
social and cognitive context accessible to the analyst and employ the analysts expert
knowledge to an a priori formulation of the functional thesis. The source of the analysts expert knowledge is not only his/her cognitive experience, but also the data provided by disciplines outside the principal domain of linguistics. Thus, macropragmatics
fosters interdisciplinarity, a notable example of which is Critical Discourse Analysis.
References
Alston, William
1994
Illocutionary acts and linguistic meaning. In: Savas Tsohatzidis (ed.), Foundations of Speech Act Theory: Philosophical and Linguistic Perspectives,
2949. London: Routledge.
Atlas, Jay
2004
Presupposition. In: Laurence Horn and Gregory Ward (eds.), The Handbook of
Pragmatics, 2952. Oxford: Blackwell.
Attardo, Salvatore
1990
The violation of Grices maxims in jokes. In: Kira Hall, Michael Meacham and
Richard Shapiro (eds.), Proceedings of the 16th Annual Meeting of the Berkeley
Linguistics Society, 355362. Berkeley, CA: University of California Press.
Attardo, Salvatore
1993
Violation in conversational maxims and cooperation: The case of jokes. Journal of Pragmatics 19: 532558.
Austin, John
1962/1975 How to Do Things with Words. Oxford: Oxford University Press.
Bach, Kent
1999
The myth of conventional implicature. Linguistics and Philosophy 22:
327366.
Bateson, Gregory
1972
Steps of an Ecology of Mind. Toronto: Chandler Publishing Company.
Bauman, Richard
1992
Contextualization, tradition, and the dialogue of genres: Icelandic legends of
the kraftskald. In: Alessandro Duranti and Charles Goodwin (eds.), Rethinking
Context: Language as an Interactive Phenomenon, 125146. Cambridge:
Bazzanella, Carla
2002
The significance of context in interpretation: The we case. In: Bruce Edmonds
and Varol Akman (eds.), Context in Context. Special issue of Foundations of
Science 7, 239254.
71
Beaver, David
2001
Presupposition and Assertion in Dynamic Sentences. Stanford: CSLI.
Bertolet, Rod
1994
Are there indirect speech acts? In: Savas Tsohatzidis (ed.), Foundations of
Speech Act Theory: Philosophical and Linguistic Perspectives, 335349. London: Routledge.
Bezuidenhout, Anne
1997
Pragmatics, semantic underdetermination and the referential/attributive distinction. Mind 106: 375409.
Bezuidenhout, Anne
2002
Truth-conditional pragmatics. Philosophical Perspectives 16: 105134.
Biber, Douglas
1988
Variation across Speech and Writing. Cambridge: Cambridge University
Press.
Blakemore, Diane
1992
Understanding Utterances: An Introduction to Pragmatics. Oxford: Blackwell.
Bublitz, Wolfram
2003
Emotive prosody: how attitudinal frames help construct context. In: Ewald
Mengel, Hans-Jrg Schmid and Michael Steppat (eds.), Anglistentag 2002
Bayreuth, Proceedings, 381391. Trier: Wissenschaftlicher Verlag.
Bublitz, Wolfram
2006
Bhler, Karl
1934/1982 Sprachtheorie. Die Darstellungsfunktion der Sprache. Stuttgart/New York:
Fischer.
Cap, Piotr
2008
Towards the proximization model of the analysis of legitimization in political
discourse. Journal of Pragmatics 40: 1741.
Cap, Piotr
2010
Axiological aspects of proximization. Journal of Pragmatics 42: 392407.
Carston, Robyn
1996
Metalinguistic negation and echoic use. Journal of Pragmatics 25: 309330.
Carston, Robyn
2002
Cole, Peter (ed.)
1978
Syntax and Semantics 9: Pragmatics. London: Academic Press.
Cummings, Louise
2005
Davis, Wayne
1998
Implicature: Intention, Convention, and Principle in the Failure of Gricean
Theory. Cambridge: Cambridge University Press.
72
Piotr Cap
Duranti, Alessandro and Charles Goodwin (eds.)

1992
Rethinking Context. Language as an Interactive Phenomenon. Cambridge:
Elugardo, Reinaldo and Robert Stainton
2004
Shorthand, syntactic ellipsis, and the pragmatic determinants of what is said.
Mind and Language 19: 44271.
Fairclough, Norman
1995
Critical Discourse Analysis: The Critical Study of Language. London: Longman.
Fetzer, Anita
2002
Fetzer, Anita (ed.)
2007
Context and Appropriateness: Micro Meets Macro. Amsterdam: Benjamins.
Fillmore, Charles
1997
Lectures on Deixis. Stanford: CSLI.
Frege, Gottlob
1892
ber Sinn und Bedeutung. Zeitschrift fr Philosophie und Philosophische Kritik 100: 2550. Trans. as On sense and reference. In: Peter Geach and Max
Black (eds.), 1952 Translations from the Philosophical Writings of Gottlob
Frege. Oxford: Blackwell.
Gazdar, Gerald
1979
Pragmatics: Implicature, Presupposition and Logical Form. London: Academic Press.
Goffman, Erving
1986
Green, Georgia
1989
Pragmatics and Natural Language Understanding. Hillsdale, NJ: Lawrence
Erlbaum.
Grice, Paul
1975
Logic and conversation. In: Peter Cole and Jerry Morgan (eds.), Syntax and
Semantics 3: Speech Acts, 4158. New York: Academic Press.
Grice, Paul
1981
Presupposition and conversational implicature. In: Peter Cole (ed.), Radical
Pragmatics, 183198. New York: Academic Press.
Grice, Paul
1989
Studies in the Way of Words. Cambridge, MA: Harvard University Press.
Grundy, Peter
1995
Doing Pragmatics. London: Edward Arnold.
Gumperz, John
1992
Gumperz, John
2003
Response essay. In: Susan Eerdmans, Carlo Prevignano and Paul Thibault
(eds.), Language and Interaction. Discussions with John J. Gumperz, 105
73
Halliday, Michael and Ruqaiya Hasan

1976
Cohesion in English. London: Longman.
2000
System and text: making links. Text 20: 201210.
Holdcroft, David
1994
Indirect speech acts and propositional content. In: Savas Tsohatzidis (ed.),
Foundations of Speech Act Theory: Philosophical and Linguistic Perspectives,
Horn, Laurence
1984
Toward a new taxonomy for pragmatic inference: Q-based and R-based implicature. In: Deborah Schiffrin (ed.), Meaning, Form, and Use in Context: Linguistic Applications, 1142. Washington, DC: Georgetown University Press.
Horn, Laurence
1996
Presupposition and implicature. In: Shalom Lappin (ed.), The Handbook of
Contemporary Semantic Theory, 299320. Oxford: Blackwell.
Horn, Laurence and Gregory Ward (eds.)
2004
Horn, Laurence
2007
Toward a Fregean pragmatics: Voraussetzung, Nebengedanke, Andeutung. In:
Istvan Kecskes and Laurence Horn (eds.), Explorations in Pragmatics: Linguistic, Cognitive and Intercultural Aspects, 3969. Berlin: Mouton de
Gruyter.
Horn, Laurence
2009
WJ-40: Implicature, truth, and meaning. International Review of Pragmatics
1: 334.
Huang, Yan
2007
Huang, Yan
2009a
Neo-Gricean pragmatics and the lexicon. International Review of Pragmatics
1: 118153.
Huang, Yan
2009b
Neo-Gricean pragmatic theory of conversational implicature. In: Bernd Heine
and Heiko Narrog (eds.), The Oxford Handbook of Linguistic Analysis,
669696. Oxford: Oxford University Press.
Hymes, Dell
1974
Foundations in Sociolinguistics: An Ethnographic Approach. Philadelphia:
University of Philadelphia Press.
Karttunen, Lauri
1973
Presuppositions of compound sentences. Linguistic Inquiry 4: 169193.
Lakoff, George and Mark Johnson
1999
Philosophy in the Flesh: The Embodied Mind and its Challenge to Western
Thought. New York: Basic Books.
Langacker, Ronald
1987
Foundations of Cognitive Grammar. Vol. I: Theoretical Prerequisites. Stanford, CA: Stanford University Press.
Langacker, Ronald
2001
Discourse in Cognitive Grammar. Cognitive Linguistics 12: 143188.
74
Piotr Cap
Levinson, Stephen
1979
Levinson, Stephen
1983
Levinson, Stephen
2000
Linell, Per
1998
Approaching Dialogue. Amsterdam: Benjamins.
Luckmann, Thomas
1995
Interaction planning and intersubjective adjustment of perspectives by communicative genres. In: Esther Goody (ed.), Social Intelligence and Interaction:
Expressions and Implications of the Social Bias in Human Intelligence, 175
Lyons, John
1995
Linguistic Semantics. Cambridge: Cambridge University Press.
Mann, William and Sandra Thompson
1983
Relational propositions in discourse. Technical Report ISI/RR-83115. University of Southern California: Information Sciences Institute.
Manning, Paul
2001
On social deixis. Anthropological Linguistics 43: 54100.
Marmaridou, Sophia
2000
Mey, Jacob L.
1993
Pragmatics. Oxford: Blackwell.
Nunberg, Geoffrey
1993
Indexicality and deixis. Linguistics and Philosophy 16: 144.
Recanati, Franois
1993
Direct Reference: From Language to Thought. Oxford: Blackwell.
Recanati, Franois
2001
What is said. Synthese 128: 7591.
Recanati, Franois
2004
Saeed, John
2009
Semantics. 3rd edition. Oxford: Blackwell.
Sadock, Jerrold
1978
On testing for conversational implicature. In: Peter Cole (ed.), Syntax and
Semantics 9: Pragmatics, 281298. London: Academic Press.
Sadock, Jerrold
2004
Speech acts. In: Laurence Horn and Gregory Ward (eds.), The Handbook of
Saville-Troike, Muriel
1989
The Ethnography of Communication. Oxford: Blackwell.
Searle, John
1975
A taxonomy of speech acts. In: Keith Gunderson (ed.), Minnesota Studies in
the Philosophy of Science 9: 344369.
75
Searle, John
1979
Soames, Scott
1989
Presupposition. In: Dov Gabbay and Franz Guenthner (eds.), Handbook of
Philosophical Logic, 553616. Dordrecht: Reidel.
1986
Stalnaker, Robert
1973
Presuppositions. Journal of Philosophical Logic 2: 447457.
Ten Have, Paul
2007
Doing Conversation Analysis. A Practical Guide. Thousand Oaks: Sage.
Strawson, Peter
1952
Introduction to Logical Theory. London: Methuen.
Thibault, Paul
2003
Contextualization and social meaning-making practices. In: Susan Eerdmans,
Carlo Prevignano and Paul Thibault (eds.), Language and Interaction. Discussions with John J. Gumperz, 4162. Amsterdam: Benjamins.
Thomas, Jenny
1995
Tsohatzidis, Savas (ed.)
1994
Foundations of Speech Act Theory: Philosophical and Linguistic Perspectives.
London: Routledge.
Van Dijk, Teun
1977
Text and Context. London: Longman.
Van Dijk, Teun
1980
Macrostructures. An Interdisciplinary Study of Global Structures in Discourse,
Interaction, and Cognition. Hillsdale, NJ: Erlbaum.
Van Dijk, Teun
1995
Discourse, semantics and ideology. Discourse and Society 6: 243259.
Van Dijk, Teun
2002
Ideology: Political discourse and cognition. In: Paul Chilton and Christina
Schaeffner (eds.), Politics as Text and Talk, 3357. Amsterdam: Benjamins.
Van Dijk, Teun
2008
University Press.
Van Dijk, Teun and Walter Kintsch
1983
Strategies of Discourse Comprehension. New York: Academic Press,
Verschueren, Jef
1999
Understanding Pragmatics. London: Edward Arnold.
Wodak, Ruth and Paul Chilton (eds.)
2005
A New Agenda in (Critical) Discourse Analysis: Theory, Methodology and Interdisciplinarity. Amsterdam: Benjamins.
3.

Sophia Marmaridou
1.
Introduction
Within a broad definition of pragmatics as the scientific study of all aspects of linguistic behaviour, the distinction between pragmalinguistics and sociopragmatics
is intended to focus on two methodological approaches to pragmatic analysis.
Pragmalinguistics typically concerns the study of the particular resources that a
given language provides for conveying pragmatic meaning (illocutionary and interpersonal), whereas sociopragmatics relates pragmatic meaning to an assessment
of participants social distance, the language communitys social rules and appropriateness norms, discourse practices, and accepted behaviours. This methodological distinction, initially launched by Leech (1983) and Thomas (1983) and widely
adopted in subsequent work in pragmatics, is distinct from, though not totally unrelated to, earlier work in Marxist pragmatics, in which the term pragmalinguistics
essentially marks the pragmatic turn in linguistic analysis and is broadly associated, or even identified, with sociolinguistics, stylistics, or text linguistics (see
Mey 1979; Prucha 1983). Moreover, the term sociopragmatics has been used in
contrast to psychopragmatics (Dascal and Franozo 1989), also independently
from the previous distinction just mentioned.
In this paper I shall argue that, as defined above, the distinction between pragmalinguistics and sociopragmatics suggests a conceptual dichotomy in our understanding of language either as a system used for communication purposes, or as an
action domain reflecting, or co-constructing, social order in culturally sanctioned
ways. Regardless of whether this conceptual dichotomy is systematically reflected
in the methodological distinction it motivates, the latter has been particularly productive. In fact, it has both affected the internal organisation and theoretical refinement of other concepts, such as pragmatic failure and pragmatic transfer, and also
informed particular methodological approaches to second language teaching and
testing. Moreover, it has been of particular relevance to historical pragmatics and
historical corpus linguistics. The apparent usefulness of the distinction between
pragmalinguistics and sociopragmatics seems to be at odds with the fact that the
borderline between the two is often fuzzy, to the extent that they are often thought
of as the two end points of a continuum. Given the variety of studies straddling different points along this continuum, I propose to use the figure/ground schema afforded us by psychology to further argue that the two concepts under investigation
essentially reflect a methodological choice in current practices to focus on one area
of pragmatic research by conferring background status to another. The use of the
78
Sophia Marmaridou
figure/ground relationship in this context seeks to explain the bibliographically attested fluidity of the content of these concepts and to present them as distinct, but
interacting, instances of perspectivisation.
It has become obvious from the above that the discussion of the concepts
underlying the terms under investigation will adopt a historical approach to their
analysis. This is expected on the one hand to place pragmalinguistics and sociopragmatics in the wider framework of pragmatic study that was flourishing in the
two crucial decades of its theoretical formation, and, on the other, to show how
these concepts have gradually permeated more recent trends in pragmatic study.
In the next section reference will be made to the ways in which the terms pragmalinguistics and sociopragmatics have been used independently of each other and
prior to the methodological distinction they have been mainly associated with as
introduced by Leech (1983) and Thomas (1983). In particular, the term pragmalinguistics as used in Marxist linguistics in the late seventies and early eighties will
be shown to be broadly equivalent to pragmatic linguistics or societal pragmatics
(Mey 1993). In this framework, the term covers research which emphasises functional aspects of language use within various paradigms (Marxist, or otherwise).
Moreover, sociopragmatics, as a theory of the ways in which the non-linguistic environment affects language use, will be contrasted with psychopragmatics as a theory of how the linguistic environment affects thought (Dascal and Franozo 1989).
In the third section the distinction between pragmalinguistics and sociopragmatics
will be discussed from the perspective of Leech (1983) and Thomas (1983). More
specifically, the conceptual underpinnings of this distinction will be set off against
established research practices concerning issues of contrastive linguistics, pragmatic failure, pragmatic transfer, pragmatic development, and the teaching and
testing of pragmatics in second language learning settings. In the fourth section the
contribution of the pragmalinguistic and sociopragmatic perspective to historical
pragmatics and historical corpus linguistics will be examined with a view to highlighting more recent trends in pragmatic research. In the last section of this paper
an overall assessment of the concepts of pragmalinguistics and sociopragmatics
will be attempted and their distinction will be evaluated with respect to the fields of
study in which it has been adopted.
2.
The borders between pragmalinguistics and sociopragmatics
In order to appreciate the conceptual dichotomy encoded in the terms pragmalinguistics and sociopragmatics, one should be reminded of the two main trends that
have dominated the field of pragmatics ever since its inception as a distinct, established field of linguistic enquiry. In the Anglo-American tradition, pragmatics is
mainly concerned with speakers meaning as intentionally communicated to an addressee. As Jucker (2008: 894) observes, pragmatics in this tradition is a rescue
79
plan for syntax and semantics that deals with a systematic account of implicit
meanings and deals with intentional human action. Levinson (1983: 732), while
espousing this tradition, has identified, discussed, and amply demonstrated problems of defining pragmatics in this narrow sense, to conclude that the best way to
delimit this field of study is to actually observe what practitioners do, or provide a
list of the phenomena which a pragmatic theory must account for, namely deixis,
implicature, presupposition, speech acts, and aspects of discourse structure (Levinson 1983: 27). He also points out that the upper bound of pragmatics is provided
by the borders of semantics and the lower bound by sociolinguistics (and perhaps
psycholinguistics too), and admits that drawing a boundary between sociolinguistic and pragmatic phenomena is likely to be an exceedingly difficult enterprise
(Levinson 1983: 29).
Within the tradition of Levinsons topical definition of pragmatics, the term
pragmalinguistics has been used in the area of literary linguistics, or stylistics, to
refer to the study of Gricean maxims, speech acts and politeness phenomena in a
literary work. An example is Haverkates (1994) study of the dialogues of Don
Quixote de la Macha in Cervantess well known novel. Apparently, this pragmalinguistic frame of reference provides new insights into the interactional roles and
personality traits of Don Quixote and Sancho Panza, which traditional literary theory does not typically address.
The Anglo-American perspective on pragmatics has been criticised as leaving
out, or backgrounding, the larger social context of communication, which is of
central importance in the European tradition (Jucker 2008: 895). The latter derives more directly from Morriss (1938) definition of semiotics, distinguishing between syntactics as the formal relation of signs to one another, semantics as the relation of signs to the objects to which the signs are applicable, and pragmatics as
the relation of signs to interpreters. In Morriss account, the user/speaker is seen in
a wider perspective, since the human being, as an interpreter of signs, does not only
have a mental identity, but also a biological and social identity, which affects her
interpretation of signs (Marmaridou 2000: 18). In fact, Morris acknowledges the
dependence of semiosis on the human being as an entity that is simultaneously the
source and the subject of all psychological, biological and sociological phenomena
that have to do with the functioning of signs comprising language use:
It is a sufficiently accurate characterization of pragmatics to say that it deals with the
biotic aspects of semiosis, that is, with all the psychological, biological, and sociological phenomena which occur in the functioning of signs. (Morris 1938: 108)
Within the context of these two traditions in pragmatics, the terms pragmalinguistics and sociopragmatics have both been used to refer broadly to pragmatic phenomena that characterise the use of language for communication purposes. In what
follows, it will be shown that, while pragmalinguistics has been associated with a
societal and Marxist perspective on language use, sociopragmatics generally refers
80
Sophia Marmaridou
to the study of the external pragmatic factors affecting language use. In this sense,
sociopragmatics has also been juxtaposed to psychopragmatics, the latter focusing
on internal, cognitive operations such as reasoning and problem solving. It should
be pointed out that pragmalinguistics and sociopragmatics, as defined above, constitute two different and unrelated projects. Therefore they do not reflect the distinction and complementariness they have been associated with in the work of
other scholars, such as Leech (1983) and Thomas (1983) (see section 3). Rather,
the use of these two terms in the frameworks discussed below is indicative of the
various developments of pragmatics within the two decades of its initial expansion.
Focusing on sociological phenomena, Mey (1979) uses the term pragmalinguistics to redefine the object and goal of linguistic enquiry in a Marxist framework
as the study of language (the linguistic object) in a variety of pragmatic settings. In
his analysis he identifies pragmatics with the description of action in general and
pragmatic linguistics (or pragmalinguistics) with the study of linguistic action as
determined by the social conditions for speaking and understanding. As he claims,
from a pragmatic viewpoint, the conditions for speaking and understanding, for
production and consumption of language, cannot be divorced from the conditions of production and consumption in the society in general (Mey 1979: 11).
The totality of conditions that are active in the production and consumption of
texts, including co-text, constitute the context, which is dynamic, rather than static,
and creative, rather than passive.
Following Meys understanding of pragmalinguistics, Olsen (1979) raises the
issue of linguistic meaning as located in, and defined by, language use in its social
context. As he observes,
the meanings of words and utterances were not universals which merely were differently labelled in different languages. The meanings were dependent on, and part of, the
different social systems []. Utterances are produced and understood within the definition of the situation and it is the rules by which we define the situation that are shared
by the members of a social system (Olsen 1979: 247).1
That linguistic meaning is a function of language use and that it is being constructed all the time by the process of interaction are facts that, according to Olsen,
must be accounted for by pragmatic linguistics. It transpires that, in his analysis,
pragmalinguistics is to be broadly understood as the pragmatic turn, or a pragmatic
perspective on linguistics as informed by social theory.
As already mentioned, pragmalinguistics is also understood to generally cover
the study of functional aspects of language use as practised in Eastern Europe in
the late seventies. The relevant studies spread over the areas of functional stylistics, text linguistics, rhetorics, psycholinguistics, sociolinguistics, social communication theory, and semiotics (see Prucha 1983). Within such a broad framework, pragmalinguistics emerges as a theory of the pragmatic component of a
grammar and as a theory of verbal activity in social intercourse. These two per-
81
spectives are reflected in the analysis of the meaning of utterances as comprising

contextual and situational aspects respectively, as in Bang and Drs (1979) work.
Contextual meaning is determined by the lexicon and syntax, style and genre, and
textual implications, and it is to a high degree reproducible and general. Situational
meaning expresses an interpretation of the world, society and the human being, and
carries an evaluative orientation. To accept a situational meaning is to subscribe to
a certain set of values. In this understanding, contextual meaning belongs to the
pragmatic component of grammar, whereas situational meaning develops from
verbal activity in social intercourse. These two kinds of utterance meaning are said
to have a dialectal relation to one another and to define a continuum ranging from
the unique and unrepeatable (situational) to the general and reproducible (contextual) (Bang and Dr 1979: 46).
It transpires from the above that pragmalinguistics in the Eastern European and
Marxist traditions encompasses aspects of language use which range from grammatically and textually encoded elements to socially constructed discursive practices. The emphasis is typically placed on the latter, which figures prominently in
this tradition against the background of the grammatical and textual aspects of language. This theoretical turn is to be interpreted both as a reaction to explicitly formal accounts of language in the various structural approaches of the time (see, for
example, the work of Chomsky), and as an attempt to articulate an ideological
stance towards the epistemological foundations of linguistics. The realisation that
utterance meaning, as the domain proper of pragmalinguistic investigation, is a
function of linguistic means in interaction with, or in a dialectal relation to, the social conditions of language use creates a conceptual frame for what is later to be established as a distinction between pragmalinguistics and sociopragmatics, albeit
outside explicitly stated ideological agendas.
Apart from the societal perspective on language and the association of pragmatics with the social conditions of language use, the cognitive perspective on language contributes to its construal as an instrument of thought. The latter line of investigation has led to the use of the terms sociopragmatics and psychopragmatics
to carry the distinction between the use of language for communication purposes
and the use of language as an instrument of thought, respectively. More specifically, in their attempt to shed light on the pragmatic turn in psycholinguistics, Dascal and Franozo (1989) refer to pragmatics as the field of study concerned with
conversational maxims, politeness requirements, contextual circumstances of utterances, and other factors that impose constraints on the selection and understanding, in a given context, of a particular linguistic expression for a certain cognitive
content (Dascal and Franozo 1989: 5). In many cases, the pragmatic turn in psycholinguistics amounts to the study of the acquisition of pragmatics by young
children and, in particular, the study of how children acquire and perform speech
acts and how they establish interpersonal communicative relations (Bates 1976).
However, Dascal and Franozo (1989) observe that language plays an important
82
Sophia Marmaridou
role in our mental processes and higher level cognitive processes or states, such as
reasoning, problem solving, storing and retrieving information, believing, etc.
They further claim that the use of language for such purposes also belongs to pragmatics: we use language not only to communicate, but also to do our reasoning, to
solve problems, etc. They therefore use the term psychopragmatics to refer to the
study of such mental use of language and distinguish it from the study of its use for
communicative purposes, i.e. from sociopragmatics. By the former term they refer
to internal pragmatic factors that affect the performance of the cognitive operations
themselves which in some way involve language. By sociopragmatics they refer to
external pragmatic factors that concern the perception and the production of linguistic signs in a particular situation, such as indirectness in the performance of
speech acts. It follows that psychopragmatics is a theory of the way in which the
linguistic environment of thought influences thought, whereas sociopragmatics is a
theory of the ways in which the non-linguistic environment (which includes both
thought and situation) affects language interpretation and production (Dascal and
Franozo 1989: 13).
The proposed distinction between sociopragmatics and psychopragmatics
serves to compare and contrast language use in jokes and dreams, topics mainly addressed in the framework of Freuds writings (see Freud 1900; 1905). More specifically, Dascal (1985: 98) argues that jokes depend for their effectiveness on the
existence of the sociopragmatic phenomenon of indirectness; they are intended to
be understood as such by an audience and therefore perform a social function. As
opposed to jokes, dreams are mental objects, not meant to be communicated to an
audience, i.e. they have an asocial character. Therefore, they are not socially constrained, their only constraint being the existence of some kind of associative link,
perhaps arbitrary and fortuitous, that provides some path for the expression of
a problematic content. In this respect the psychopragmatic role of language
in dreams is [] similar to its use for another cognitive purpose namely, as a mnemonic device (Dascal 1985: 104). In this analysis, Dascals sociopragmatics
covers the broader area of general pragmatics in both the Anglo-American and the
European traditions with the aim of distinguishing it from psychopragmatics, the
latter illustrating a pragmatic turn in psycholinguistics.
3.
The pragmalinguistics/sociopragmatics distinction

in a cross-linguistic and cross-cultural perspective
The distinction between pragmalinguistics and sociopragmatics was essentially established within the framework of applied linguistics studies on which it was to
exert a strong influence. It was initially introduced in the work of Thomas (1981),
further explored in Thomas (1983), and explicitly spelt out as a methodological approach in pragmatics in Leech (1983). In the latters programme, General Prag-
83
matics is that aspect of human language which concerns the use of language in
communication. More specifically, the understanding of the nature of language
depends both on the study of grammar, as the abstract, formal system of language
(comprising phonology, syntax and semantics), and on the study of pragmatics, as
the interaction of the Cooperative and Politeness Principles partly characterizing
language use. Moreover, Leech (1983: 7) views semantics as a level of contact between grammar and pragmatics, so that linguistic meaning can be fruitfully studied
by a combination of semantic and pragmatic approaches, the former targeting the
sense of an utterance and the latter its force.
Within Leechs project, General Pragmatics is a fairly abstract study of the general conditions of the communicative use of language, excluding more specific
local conditions on language use: The latter may be said to belong to a less
abstract field of SOCIO-PRAGMATICS, for it is clear that the Cooperative Principle and the Politeness Principle operate variably in different cultures or language
communities, in different social situations, among different social classes, etc.
(Leech 1983: 10). Alongside sociopragmatics, Leech introduces PRAGMALINGUISTICS as the study of the more linguistic end of pragmatics where we consider the particular resources which a given language provides for conveying particular illocutions (Leech 1983: 11). It follows that sociopragmatic studies are
culture-specific, while pragmalinguistic studies are language-specific. Apparently,
Leechs observation that sociopragmatics is the sociological interface of pragmatics is implicitly an attempt to address an issue that Levinson (1983: 29) also
raises when he observes that the boundary between sociolinguistic and pragmatic
phenomena is hard to define.
Since Leech (1983) is concerned with establishing the Cooperative and Politeness Principles as the two cornerstones of General Pragmatics and with arguing
their universality, he does not pursue the pragmalinguistics/sociopragmatics distinction any further. Nevertheless, he views both principles as the natural constraints of universality. For example, while claiming the universality of the principles, he concedes that they do not apply in an identical manner in all societies:
Indeed, one of the main purposes of sociopragmatics [] is to find out how different societies operate maxims in different ways, for example by giving politeness a higher rating than cooperation in certain situations, or by giving precedence to one maxim of the
Politeness Principle rather than another (Leech 1983: 80).
Similarly, he observes that cultural stereotyping (e.g. the British are more tactful
that the Americans) is related to pragmalinguistic strategies such as strategies of
indirectness, and the norms observed in the performance of these strategies in different speech communities (Leech 1983: 231).
Leechs (1983) distinction between pragmalinguistics and sociopragmatics is
the natural outcome of his claim about the universality of the principles of pragmatics he proposes. By prioritizing conversational principles, he proposes a rhe-
84
Sophia Marmaridou
torical model of pragmatics referring to the study of the effective use of language
in communication and focusing on the goal-oriented speech situation, whereby the
speaker uses language in order to produce a particular effect in the mind of the
hearer. Leechs principles of general pragmatics are to be held constant and form
the basis, or the ground, against which all other language- and culture-specific
pragmatic variables can be studied as distinct figures. This position lends itself to
cross-linguistic and cross-cultural comparisons of communicative behaviour with
obvious applications for language teaching/learning and testing. Moreover, the distinction proposed within this framework seems compatible with the aims and
methods of the study of discourse and historical pragmatics in particular. In what
follows, these areas of investigation will be related to the conceptual and methodological distinction between pragmalinguistics and sociopragmatics as introduced in this section.
4.
Applications of the pragmalinguistics/sociopragmatics distinction
4.1.
Preliminary investigations in contrastive linguistics
The impact of the pragmalinguistics/sociopragmatics distinction has been particularly felt in the conceptual organisation of some theoretical constructs it has motivated, such as pragmatic failure, pragmatic transfer, and pragmatic development.
This development in the area of applied linguistics is to some extent foreshadowed
by Riley (1979) advocating a contrastive pragmalinguistics.
Taking the broad view of pragmalinguistics as pragmatic linguistics, Riley
(1979) defines its aim as the study of communicative acts. These may be realised
verbally, paralinguistically, or non-verbally, and rest on a theory of illocution; illocutionary force is extended in this system to all kinds of communicative acts. Focusing on linguistic acts, he notes that what is central to contrastive pragmalinguistics is language functions rather than linguistic structures, discourse, not
grammar, the communicative act in context, not the sentence in isolation (Riley
1979: 57). Drawing in this way a methodological line between form and function,
he highlights instances of same form and different function (e.g. the possibility of
using Youre not going out to express the illocutions of prohibiting, confirming,
threatening, expressing surprise, or stating), and instances whereby the same function is performed by different forms (e.g. the function of agreeing as expressed by
the forms Yes, of course, I agree, nodding, or repetition of the previous speakers utterance). In his model, form-function pairings constitute the ground against
which cross-linguistic contrasts are to be drawn.
In order to place this model in a wider framework encompassing all aspects of
pragmatic study, Riley (1979) further proposes cross-linguistic contrasts across
what he calls discourse structure, consisting of three sub-levels of structure:
85
Firstly, formal structure includes all aspects of message-bearing elements, e.g. verbal, paralinguistic, or non-verbal, in a particular situation. Secondly, illocutionary
structure consists of sequences of illocutionary acts, e.g. inviting accepting, confirming thanking, etc. Thirdly, interactive structure refers to linguistic organization in terms of interactional tactics such as turns (e.g. opening, reply, closing, address, exchange, transaction, etc.). Evidently, Rileys work is an attempt to include
in his model all topics of pragmatic interest and put them to the forefront of contrastive linguistics. However, his efforts to systematise the proposed model lead to
a schematic and idealised understanding of communicative acts. For example, the
sequences of illocutionary acts seem to refer to idealised prototypes. In reality, invitations are typically, but not always, followed by acceptances, or refusals, confirmations by thanks, etc.
This apparent theoretical weakness of Rileys model, though, is counterbalanced by its applicability in contrastive analysis and language teaching in particular. More specifically, the model allows for the identification and the teaching of interactional styles and illocutionary point in relation to sociolinguistic variables
such as roles, status, formality, and discursive control, as for example, in a typical
teacher student illocutionary sequence, or in a typical asking-for-street-instructions exchange between strangers. In this way, Rileys proposal addresses newly
created needs in the methodology of communicative language teaching research
that was being carried out at the time (see, e.g., Candlin, 1976; Brumfit and
Johnson, 1979; Widdowson, 1979). Perhaps the most important contribution of
Rileys work in introducing contrastive pragmalinguistics is that he creates an
awareness of the significance of pragmatic knowledge to language teaching and
emphasises the systematicity of such knowledge so that it can be actually taught.
His suggestions clearly encourage a methodological turn, to be further explored in
the context of the pragmalinguistics/sociopragmatics distinction introduced by
Thomas (1981; 1983) and Leech (1983).
4.2.
Pragmatic failure
Following Leechs (1983) division of linguistics into grammar and pragmatics and
adopting his goal-oriented understanding of language use (see above), Thomas
(1983) first construes linguistic competence as consisting of grammatical competence (abstract knowledge of phonology, syntax and semantics) and pragmatic
competence as the ability to use language effectively in order to achieve a specific
purpose and to understand language in context (Thomas 1983: 92). Within this
framework, pragmatic failure is defined as an inability to recognise the force of
the speakers utterance when the speaker intended that this particular hearer should
recognise it (Thomas 1983: 94). Notably, whereas grammatical error indicates the
speakers restricted linguistic proficiency, pragmatic failure reflects badly on her as
a person. It follows that, in the context of cross-cultural communication, pragmatic
86
Sophia Marmaridou
failure may lead to a misunderstanding of the foreign speakers intentions. Awareness-raising tasks aiming to develop foreign students metapragmatic ability to
consciously analyse language use are commendable in this case, but inadequate in
Thomass (1983: 99) view, unless informed by the distinction between pragmalinguistic and sociopragmatic failure.
The proposed categorisation of types of pragmatic failure, consistent with
Leechs (1983) dyad of language vs. culture specificity, is not intended to impose
rigid boundaries. Thomas (1983: 101) explicitly states that the two concepts form a
continuum from what is language-specific to what is culture-specific. Pragmalinguistic failure results from the speakers mapping of pragmatic force to an utterance
in a way that is systematically different from a native speakers. Sociopragmatic
failure results from the speakers miscalculation of the social conditions placed on
language in use, such as the size of the speakers imposition on the addressee in
performing a particular speech act, a cost/benefit scale of acts in the particular culture, the social distance between interlocutors, and relative, culture-specific rights
and obligations of interlocutors to each other in specific situations. As one moves
from the communicative intent of a speakers utterance to the interpersonal dynamics between speaker and addressee, one is also moving from the pragmalinguistic to the sociopragmatic end of the continuum.
In providing examples of pragmalinguistic failure, Thomas (1983: 101) concedes to the systematicity of formfunction pairings in a particular language and
culture in the sense that native speakers assign pragmatic force to certain utterances fairly predictably. Unlike Riley (1979), she seems to restrict the scope of this
systematicity to highly conventionalised utterances, such as the often-quoted can
you do X English utterance used to issue a request rather than ask for ones ability.
The fact that in other languages, such as French or Russian, the opposite is true
may lead the foreign language learner to an inappropriate transfer of a formfunction pair, and hence to pragmalinguistic failure. A further cause of pragmalinguistic failure may be the inappropriate transfer of speech act strategies from L1 to
L2, such as the use of direct vs. indirect forms of issuing requests politely. For
example, it is reported that polite, indirect requests for street instructions (e.g. excuse me, could you please tell me how to ) would be counterproductive in Russian, while they are pretty standard in English. Hence, the Russian learner of English using direct means to perform a request might be misunderstood as being
impolite (Thomas 1983: 102).
Politeness is indeed an area in which pragmalinguistic failure merges with sociopragmatic failure. Whereas the choice of polite forms and strategies seems to
belong to the pragmalinguistic end of the continuum, when and to whom to be polite concerns the sociopragmatics end. In Thomass view, the size of imposition of a
speech act, the awareness of taboos in a particular cultural context, and differing
assessments of relative power or social distance between interlocutors may lead to
sociopragmatic failure. One may add to these sources of sociopragmatic failure the
87
orientation of a culture to positive or negative politeness (see Sifianou 1992). However, it is not always clear whether an instance of pragmatic failure can be assigned
to the pragmalinguistic or to the sociopragmatic level. An example that Thomas provides has to do with the conception of what constitutes free goods in
the former Soviet Union as opposed to Britain. Asking for a cigarette in the Soviet
Union requires a minimal degree of politeness, as these items are free goods in
this cultural setting. A Russian asking for a cigarette in Britain and using a similar
strategy may have either encoded wrongly the correct amount of politeness she
intended, or misjudged the amount of imposition of her request. The former case
would represent an instance of pragmalinguistic, while the latter of sociopragmatic
failure. The effect of either of these two types of pragmatic failure may be the same
for the addressee and their causes difficult for the analyst to distinguish. However,
the distinction is essential in the context of language teaching, as it reflects two different types of decision-making. As Thomas observes, pragmalinguistic decisions
are language-specific and the teacher can correct them in a fairly straightforward
manner. Sociopragmatic ones are culture-specific and sensitive to the learners
own system of values; hence they should not be corrected, but only pointed out
and discussed (Thomas 1983: 109).
It is therefore clear that the discussion of pragmatic failure in the context of language teaching relies heavily on the distinction between pragmalinguistics and sociopragmatics. However, given that sociopragmatic decisions are social before
they are linguistic (Thomas 1983: 104), the importance of the distinction between
sociopragmatics and sociolinguistics becomes apparent. The identification of social roles and stereotypes, power relations between interlocutors relating to age,
social status, or dialect, socioculturally-based evaluations of situations, and linguistic value judgements have been typically considered social variables and have
traditionally concerned sociolinguistic research (see, e.g. Holmes 2008). How are
all these to be differently researched in sociopragmatics? Indeed, this question is
often raised, at least implicitly, in the relevant literature and is differently approached (see the discussion in Bou-Franch and Garcs-Conejos 2003; Roever
2006; Archer and Culpeper 2003, below). It transpires that the application of the
pragmalinguistics/sociopragmatics distinction to pragmatic failure on the one hand
enriches language teaching methodology, but on the other, raises the issue of the
theoretical distinction between sociopragmatics and sociolinguistics, to be addressed in the next section.
4.3.
Pragmatic transfer
The pragmalinguistic/sociopragmatics distinction is also reflected in the study of

pragmatic transfer within the wider context of second language acquisition research. The relevant literature strongly supports the position that non-native speakers understanding and production of linguistic action is influenced by their L1
88
Sophia Marmaridou
pragmatic knowledge (Kasper 1992; Kasper and Rose 2002). Prominent in this
kind of research is the concept of interlanguage as a stage on a continuum representing the second language learners development towards the L2 system. In Selinkers (1972) definition, interlanguage is a psycholinguistic phenomenon which
mainly refers to the learning process, rather than the outcome of this process. However, the dominant practice in interlanguage pragmatics concerns the outcome, i.e.
the collection and comparison of data from native speakers linguistic performance
in the target language, the learners use of her native language, and the learners interlanguage (i.e. her use of the target language at a particular stage of development).2 In this context, pragmatic transfer concerns the possible influences from
the learners L1 on her interlanguage performance.
Even though in practice pragmatic transfer has been typically associated with
transfer of speech act knowledge, indirectness, and politeness, it may also make
reference to sociolinguistic, or discourse parameters in the same instance of language use. For example, transfer of address terms and register concern social variation in language use and are therefore to be considered in the domain of sociolinguistics. Yet, their strategic exploitation in pursuing some illocutionary goal
places their investigation in the area of pragmatics proper. Moreover, some speech
acts are characterized by their position in discourse: greetings occur discourse-initially and refusals occur as seconds in a conversational exchange. Furthermore,
the speech act force of an utterance may become evident only after the utterance is
responded to by the interlocutor, which points to its negotiable character and conversational grounding.
As in the case of pragmatic failure, the pragmalinguistics/sociopragmatics distinction has enabled a more detailed discussion of pragmatic transfer and has added
clarity to this concept by maintaining the corresponding form/function dyad. In
Kaspers (1992) definition, pragmalinguistic transfer shall designate the process
whereby the illocutionary force or politeness value assigned to particular linguistic
material in L1 influences learners perception and production of form-function
mappings in L2 (Kasper 1992: 209). However, as shown in the discussion of pragmatic failure, politeness value does not only depend upon linguistic material, but
also upon social parameters of the speech event, such as the relative social distance
of interlocutors and the power relations between them. For example, L2 learners
may be familiar with address terms of varying degrees of deference or intimacy in
L2, but their use depends upon the learners assessment of social roles, settings,
etc., which may be based on the their own culture. Therefore, politeness is not only
a candidate for pragmalinguistic, but also for sociolinguistic transfer. As Kasper
(1992: 209) observes, the latter is operative when the social perceptions underlying language users interpretation and performance of linguistic action in L2 are
influenced by their assessment of subjectively equivalent L1 contexts. It follows
that the form-function pair does not guarantee appropriate constraints on the discussion of pragmalinguistic and sociopragmatic transfer. Whether politeness
89
transfer occurs at the pragmalinguistic or the sociopragmatic level is largely a

matter of perspectivisation. If construed as a transfer of (in)appropriate forms, the
latter are foregrounded against the background of the social conditions of politeness. If the social conditions of politeness are focused upon, the formal aspects
constitute the background to the analysis.
The issue of pragmalinguistic and sociopragmatic transfer is also associated
with the distinction between positive and negative transfer. Positive transfer of elements from L1 in L2 performance is methodologically difficult to identify, given
that acceptable L2 performance may have sources other than positive transfer of
L1 elements, for example the application of general pragmatic principles, generalisation of L2 specific knowledge, and teacher input, among others. As Kasper
(1992: 213) correctly points out performance data alone cannot tell us whether
[] learners did indeed consult their L1 knowledge or operated strictly on their interlanguage pragmatic competence. More specifically, positive transfer has been
associated with the possibility of pragmatic universals (as in Grice 1975; Leech
1983; Brown and Levinson 1987) that are assumed to function cross-culturally. Interlanguage research has shown that L2 learners have access to the same range of
strategies as native speakers in implementing linguistic actions, such as requests,
suggestions, invitations, refusals, complaints, compliments, etc. (Wolfson 1981;
Trosborg 1987; Beebe and Takahashi 1989; Blum-Kulka 1991; Olshtain and Weinback 1992). This has been taken as evidence of the existence of pragmalinguistic
universals. Similarly, L2 learners sensitivity to social factors such as interlocutors relative status, or degree of imposition, etc., possibly indicates the availability
of sociopragmatic universals. However, in spite of the fact that the evidence is
based on a variety of languages and cultures, the existence of pragmatic universals
at the pragmalinguistic or sociopragmatic level is by necessity an empirical matter,
especially since the universality of theoretical concepts in general pragmatics, such
as politeness, or the co-operative principle, have been seriously questioned (see,
e.g. Keenan 1976; Rosaldo 1982; Wierzbicka 1987; Eelen 2001). Negative pragmalinguistic transfer occurs when inappropriate forms or strategies are used in performing particular speech acts affecting the politeness value of the utterance, or
even its illocutionary force. For example, it has been reported that, in expressing
refusals, Japanese learners of English use strategies such as generalisations (I
never yield to temptation), statements of philosophy (to err is human), or suggestions for alternative action (why dont you ask someone else?) that are atypical of native English speakers (Beebe et al. 1990). At the sociopragmatic level, L2
learners perception of contextual factors may originate in their L1 and affect politeness style. It is reported that Japanese learners of English selected their refusal
strategies on the basis of whether the refusers status was higher or lower than the
interlocutors, whereas American native speakers choice was made on the basis of
status-equal or status-unequal speaker addressee relationships. In spite of the
great number of studies that address the issue of negative pragmatic transfer, the
90
Sophia Marmaridou
systematicity of the relevant experiments, and the careful assessment of their results, L2 learners inappropriate performance may also be a function of other factors, such as learning context, shifts in the available target variety of the L2, little
exposure to the target language and culture, varying L2 development stages, and
lack of contextualisation cues, among others. To the extent that the pragmalinguistics/sociopragmatics distinction shows the need for more precise experimental
techniques in the study of pragmatic transfer, it can contribute to the validation of
hypotheses and is therefore to be understood as a useful methodological tool in interlanguage pragmatics.
4.4.
Pragmatic development
A further contribution of the pragmalinguistics/sociopragmatics distinction to

pragmatic research concerns the issue of the development of grammar and pragmatics in interlanguage research. An important argument in interlanguage pragmatics evolves round the primacy of grammatical or pragmatic knowledge in
interlanguage development. While focus on the development of L2 learners
pragmatics as an autonomous component of communicative competence has had a
serious impact on second language acquisition research, its relation to other aspects of communicative ability has also attracted attention (see Bardovi-Harlig
1999 and references therein). In this context, the pragmalinguistics/sociopragmatics distinction serves to investigate what happens when existing grammatical
knowledge (such as the semantics of the progressive as a grammaticalised aspectual category) is not put to target-like pragmalinguistic use (e.g. to mitigate imposition), as in I was wondering if I could have a word (Kasper and Rose 2002: 180);
moreover, this distinction serves to investigate what happens when grammatical
knowledge leads to target-like pragmalinguistic use, but non-target-like sociopragmatic use. Robinson (1992) reports that, although a Japanese learner of English
had the appropriate knowledge to form and use the I would like to form to express preference, she nevertheless used I want to on the sociopragmatic assumption that the former form would be too polite when addressing her American
friend.
Several studies show that L2 learners may have knowledge of a particular
grammatical structure, which, however, they do not put to native-like pragmalinguistic use. For example, Bodman and Eisenstein (1988) report on L2 learners use
of sophisticated grammatical knowledge which they put to unconventional L2
pragmalinguistic use, as in expressions of thanking like May God grant you a long
life. It is hypothesised that such instances also testify to L1 pragmatic transfer
phenomena. Similarly, Takahashi (1996; 2001) has found that advanced Japanese
EFL students avoided bi-clausal I was wondering whether you could VP and favoured the mono-clausal Could you (please) VP? even though they knew the semantics of the bi-clausal structure. On this evidence Takahashi (2001: 173) con-
91
cludes that the Japanese EFL learners lack the L2 pragmalinguistic knowledge
that an English request can be mitigated to a greater extent by making it syntactically more complex. Sophisticated as the above studies may be, they cannot exclude the possibility that the particular adult L2 learners production is constrained
by sociopragmatic, rather than pragmalinguistic, considerations. For example, it
could be argued that Bodman and Eisensteins subjects may have chosen the particular non-native-like forms of thanking as a way of maintaining their cultural
identity. Besides, it is possible that the Japanese EFL students may not have assessed the particular request(s) as requiring a comparatively higher level of mitigation than others. It is therefore obvious that the lack of sharp boundaries between
pragmalinguistics and sociopragmatics is also testified in pragmatic development
studies.
Knowledge of a grammatical structure and its pragmalinguistic functions, accompanied by lack of familiarity with the sociopragmatic conditions of target-like
use, is also reported in the relevant literature. Bardovi-Harlig and Hartford (1991)
show that, in a student-advising session at an American University, a Japanese student wishing to reject the faculty advisors suggestion uses a series of questions as
a strategy of indirectly conveying negative intent. In the researchers analysis, the
student clearly had the grammatical knowledge to form questions and the pragmalinguistic knowledge to use questions in order to indirectly convey intent, but did
not have the sociopragmatic knowledge to assess when the particular strategy is
appropriate and effective. It is also reported that this question strategy was not observed in the rejections of American-English speaking students. The particular
data fall short of providing evidence for the appropriate use and effectiveness of
the question strategy in conveying intent by L2 learners or by native speakers of
the L2. For example, they do not show whether American-English speaking students employ the question strategy to indicate negative intent in other social settings, or to what extent this can happen. If it turned out that the particular strategy
to convey negative intent is not favoured by native speakers of the L2, then it is
possible that the Japanese student has erroneously associated a grammatical form
with a pragmatic strategy, which is evidence for insufficient development of her
pragmalinguistic rather than her sociopragmatic knowledge. Moreover, even if the
question strategy was associated with the same function by American speakers, it
is not clear whether the understanding of negative intent would result from the operation of sociopragmatic factors or from on-line inferencing heuristics, i.e. pragmalinguistic factors. As previously noted with respect to pragmatic failure and
pragmatic transfer, it is sometimes difficult to diagnose the source of a learners inappropriate use of a speech act strategy at any particular stage of L2 development.
92
4.5.
Sophia Marmaridou
Second language teaching and testing
The most widely stated aim of L2 teaching is to enable successful communication

in the foreign language. This involves not only the mastery of the phonology, syntax and semantics of the L2, but also its pragmatics. The question that is typically
addressed in this context is whether pragmatic knowledge can be taught in the
classroom, i.e. not necessarily by cultural immersion in the native environment,
and how it can be taught to learners of different ages, educational backgrounds, etc.
The applications of the pragmalinguistics/sociopragmatics distinction in second
language acquisition research and the consequent refinement of the concepts of
pragmatic failure and pragmatic transfer relate in an obvious way to second language teaching and testing. If this distinction is useful in, perhaps selectively, identifying causes of pragmatic failure and areas of pragmatic transfer, then L2 teaching and testing should be oriented towards the development of pragmalinguistic
and sociopragmatic appropriateness (Harlow 1990: 328).
Following Leech (1983), Kasper and Rose (2001) claim that communicative
action includes not only using speech acts (such as apologising, complaining,
complimenting, and requesting), but also engaging in different types of discourse
and participating in speech events of varying length and complexity. More importantly, speakers do not only want to perform speech acts in order to achieve
their goals, but also wish to attend to interpersonal relationships with the other
speech participants in the communicative event. Within this context, pragmalinguistics refers to the resources for conveying communicative acts and encoding interpersonal meanings. According to Kasper and Rose (2001: 2), these resources include pragmatic strategies such as directness and indirectness, speech act routines,
and the intensification or softening of communicative acts. The social perceptions
underlying participants interpretation and performance of communicative action
fall within the area of sociopragmatics and refer to speakers and hearers social
distance and social power, their respective rights and obligations, and the degree of
imposition involved in particular communicative acts.
It transpires from the above that in language teaching research, as in second
language acquisition research, pragmalinguistics and sociopragmatics focus on the
area of speech acts and, more specifically, on their linguistic realisation and their
strategic use in particular social circumstances. In this view, teaching L2 pragmatics primarily involves teaching the means and strategies of performing particular speech acts, i.e. the pragmalinguistics of L2 (see McNamara and Roever 2006),
and teaching politeness, i.e. teaching how to assess social distance between interlocutors, social power and degrees of imposition in the L2 sociocultural setting
(see Bou-Franch and Garcs-Conejos 2003). Clearly, to be pragmatically competent in L2, learners need knowledge both at the pragmalinguistic and the sociopragmatic levels. For example, learners need to know not only how to express a
particular speech act indirectly, but also the relation between indirectness and po-
93
liteness, as well as when and to whom to be polite, the degree of politeness that is
required in a particular context of interaction, etc.
Apart from the linguistic expressions themselves, strategies of communicative
actions may vary not only according to context within a particular language, but
also across languages and cultures. However, as Kasper and Rose (2001) observe,
adult learners get a considerable amount of L2 pragmatic knowledge for free, since
some pragmatic knowledge is universal, while other aspects of such knowledge
may be successfully transferred from the learners L1. At the pragmalinguistic
level, L2 teaching is facilitated by the cross-cultural availability of the major categories of speech acts (e.g. Searles (1979) directives, commissives, expressives,
etc.), and specific acts such as greetings, leave-takings, etc. Moreover, it has been
found that the major realisation strategies of some speech acts occur across several
ethnolinguistically distant speech communities (see e.g. House and Kasper 1987;
Faerch and Kasper 1989; Rose 2000, among others). It is shown that in early learning states, learners may not have acquired the linguistic means necessary to implement such strategies, but once linguistic obstacles are removed by teaching, they
will use the appropriate strategies without instruction. Similarly, corresponding
form-function pairings between L1 and L2 used in the same contexts facilitate
pragmalinguistic performance.
Positive transfer and awareness of universally existing social variables, such as
social distance, social power and degree of imposition, often affected by other social parameters such as sex, age, and familiarity between interlocutors (Harlow
1990), can also facilitate the learners task in acquiring sociopragmatic knowledge
in L2. However, pragmalinguistic and sociopragmatic conventions make reference
to the lexical and grammatical constructions of particular languages, while successfully performing communicative acts often depends on the degree to which
a strategy is conventionalised in a particular language and culture. These parameters, then, point to the necessity of classroom instruction in L2 pragmalinguistics and sociopragmatics. Especially concerning the latter, it has been argued
that the promotion of sociopragmatic awareness is also likely to improve production and comprehension of pragmalinguistic resources (Bou-Franch and GarcesConejos 2003).
The apparent usefulness of the pragmalinguistics/sociopragmatics distinction
in the study of L2 teaching is not proportionately reflected in L2 testing research.
As in the discussion of pragmatic failure above, it is often difficult in practice to determine whether a given error is due to pragmalinguistic or sociopragmatic deficits
(McNamara and Roever 2006: 55). For example, if politeness markers are missing
from an utterance, this could be either because the learner does not know these
markers, or because she is not aware that these markers should be used in the particular situation. The former indicates pragmalinguistic, whereas the latter sociopragmatic, inadequacy. Therefore, it is difficult to devise a test that would assess
pragmalinguistic knowledge to the exclusion of sociopragmatic knowledge, or the
94
Sophia Marmaridou
reverse. The problem seems more acute when testing pragmatic strategies, such as
indirectness. While the linguistic expressions used to encode indirectness belong
to the pragmalinguistic level, the social target to which they are addressed is a sociopragmatic matter. Moreover, sociopragmatic appropriateness seems to be less
dependent on linguistic proficiency than pragmalinguistic knowledge.3 L2 learners
with a high general L2 competence are likely to build their pragmalinguistic competence more easily, but high general L2 competence does not guarantee correct
judgements about sociopragmatic appropriateness.
The problem of testing pragmalinguistics and sociopragmatics independently
from each other also relates to the types of tests devised to test pragmatic knowledge in general, as there is a significant tension between the construction of authentic assessment tasks and practicality (McNamara and Roever 2006: 54). The
two most frequent types are discourse completion tests (DCTs) and role plays. In
attempting to test sociopragmatic knowledge, multiple choice DCTs were shown
not to be reliable, as it is extremely difficult to devise distractors (i.e. wrong
items) that are totally unacceptable by all members of the target language community without being wrong in a very obvious way (Brown 2001). Open DCTs can
only provide evidence of the L2 learners sociopragmatic ability concerning the
particular items in the test (e.g. apologies, requests, formulaic implicatures, etc.)
rather than her general sociopragmatic ability. Additionally, they do not provide
evidence of the learners ability to participate in conversations in real time, or to
take turns, especially given that speakers normally distribute strategies over various turns, monitoring their interlocutors reactions and adjusting their talk accordingly. Although role plays are more reliable in this respect, they are not preferred as
testing devices because they take substantial time to conduct, they are difficult to
monitor, and they require multiple rating. Besides, role play cannot establish context in the way that authentic communication does in the real world, nor does it represent the speakers or the addressees actual face needs in real terms. However, it
is reported that role plays can assess learners ability to make judgments about sociopragmatic appropriateness at discourse level (Cook 2001).
By contrast, it is possible to test L2 pragmalinguistic knowledge practically
and reliably, even though creating items for pragmalinguistic instruments is not
easy. Roever (2006), for example, constructed a test to assess interpretation of implicature, recognition of routine formulae, and knowledge of speech act strategies.
While these three types of knowledge differ, they are based on general pragmatic
and linguistic input. Interpretation of implicature was expectedly found to rely on
the learners proficiency and deduction abilities. The effect of L2 learners proficiency in their knowledge of routines was almost negligible, given that routines are
not creatively constructed, but rather learned as holistic items to be used in specific
situations. As opposed to routines, the production of speech acts and the pragmalinguistic strategies employed therein correlated with level of L2 proficiency, as
expected. Some of the limitations of assessing L2 learners pragmalinguistic
95
knowledge are similar to the ones of testing sociopragmatics. The written format of
the tests restricts the number of contextual cues that would be available in a real encounter. The lack of extended negotiation of speech acts in the context of a DCT is
another limiting factor. However, studies of pragmalinguistic assessment in L2 can
benefit both from the development of testing techniques in general and from
further cross-cultural study of pragmalinguistic parameters of communicative acts
in particular.
It becomes obvious from all the above that the pragmalinguistics/sociopragmatics distinction is better exploited in the area of L2 teaching methodology than
in studies of pragmatic failure, pragmatic transfer, pragmatic development and L2
testing. What the latter have in common is their diagnostic methodology in assessing L2 learners approximations to the target language. The sources of particular
approximations are not easy to detect. By contrast, a theoretical categorisation into
pragmalinguistic and sociopragmatic parameters of language use can fairly easily
be mapped onto teaching methods, materials and awareness raising tasks in which
L2 learners may be involved.
5.
Historical sociopragmatics4
The concepts of pragmalinguistics and sociopragmatics have also informed the

field of study generally known as Historical Pragmatics. Still a relatively young
field of linguistics, historical pragmatics takes a broad, sociologically based view
of patterns of intentional human interaction, as these are determined by the conditions of societies of earlier periods (see Jucker 2006 for a brief overview).5 One approach to historical pragmatics, pragmaphilology, describes the contextual aspects
of historical texts and is essentially synchronic. It refers to the identities of speakers and addressees, their social and personal relationship, the physical and social
setting of text production and reception and the goals of the text (Jacobs and Jucker
1995). Another approach, diachronic pragmatics, concerns form-to-function and
function-to-form mappings across different historical stages of the same language.
Apparently, form-to-function mappings correspond to a pragmalinguistic perspective on the analysis of phenomena such as deixis (Fries 1993), discourse markers
(Brinton 1990), or interjections (Taavitsainen 1995). Diachronic function-to-form
studies, adopting a sociopragmatic perspective, compare the realisation of speech
acts, politeness formulae, text types, or forms of dialogue, at different points in the
development of a language (Arnovick 1999; Jucker and Taavitsainen 2008; Fritz
1995; Biber and Finegan 1992). Since the mid-1990s the investigation of the relevant topics is based on available historical corpora (e.g. the Helsinki Corpus of
English Texts), which are either read and annotated in an old-fashioned and time
consuming way, or analysed by corpus-linguistic methods of data retrieval. Apart
from written genres, such as essays, fiction, personal letters, and manuals for good
96
Sophia Marmaridou
behaviour, speech-based genres have also been analysed, e.g. dialogues from plays
and fiction, court proceedings, political debates, public speeches, and sermons
(Biber and Finegan 1992; Jacobs and Jucker 1995).
Within this framework, historical sociopragmatics focuses on the interaction
between specific aspects of social context and particular historical language use
that leads to pragmatic meanings. More specifically, it concerns language use in its
situational, local context and the ways in which situational contexts generate
norms which interlocutors employ or exploit for pragmatic purposes. A synchronic
approach to historical sociopragmatics consists in showing how language use
shapes and is shaped by context at a particular historical period, while a diachronic
perspective involves the investigation of how shifts in language use affect shifts in
contexts, or how shifts in contexts shape language use over time. An important line
of investigation in historical sociopragmatics concerns the reconstruction of contexts on the basis of historical texts; the latter are viewed as carrying evidence of,
or projecting, their own contexts (Culpeper 2009: 182183).
As already mentioned, the local context, or the sociological context, of language use is of primary concern to sociopragmatics. For example, Nevalainen and
Raumolin-Brunberg (1995) have studied the sociopragmatics of terms of address
in Early English correspondence. Given that the way people address each other in
interaction depends on social variables such as age, status, dialect, etc., the question arises as to the relation between sociopragmatics and sociolinguistics. According to Nevalainen and Raumolin-Brunberg (2003), sociolinguistics comprises
three areas of research, namely, social dialectology, interactional sociolinguistics
and the sociology of language. Parallel to this distinction, Culpeper (2009: 180)
identifies three levels of what may constitute the context against which texts may
be understood: the most local, immediate text and co-text of interlocutors, the
medial level of social situation (including speech events, activity types, frames,
etc.), and the most general level, making reference to national and regional cultures, institutional cultures, etc. As the author notes (Culpeper 2009: 181)
sociopragmatics should primarily, though not exclusively, concern itself with the medial
context and the phenomena that constitute it. Social situations can provide a link between micro, more linguistically-oriented considerations (the typical focus of pragmalinguistics), and macro, more sociologically-oriented considerations (the typical focus
of a field such as Critical Discourse Analysis).6
According to this definition, sociolinguistics is concerned with mapping regular

patterns of usage in interaction, while sociopragmatics is concerned with the ways
in which these regular patterns are used and exploited in particular interactions.
Given the concern of historical sociopragmatics for interactional data and the
obvious non-availability of recorded forms of interaction of earlier periods, the
question arises as to the availability of historical corpora that lend themselves to
sociopragmatic research and the search tools required for this type of investigation.
97
Recent developments in historical corpus linguistics have allowed more pragmatic

research questions in a way that is amenable to corpus-linguistic methodologies
(see, e.g., Kohnen 2008; Jucker et al. 2008); but the need is still felt for the appropriate pragmatic tagging of relevant pragmatic units, such as those of interest to sociopragmatic research. Within this framework, Culpeper and Kyts compilation
of the Sociopragmatic Corpus, a subsection of the Corpus of English Dialogues:
15601760, has made sociopragmatic annotation possible. Facing the issue of the
sociopragmatic variability of context, Archer and Culpeper (2003) devise and implement a tagging system that allows the annotation of speech changes that potentially affect the social meaning of interaction in drama and trial proceedings.7 The
variables include relatively static, sociolinguistic values, such as status (nobility,
gentry, professional, ordinary commoners, lowest groups) and age (young, adult,
old), as well as dynamic values, such as activity role (witness, defendant, customer), kinship role (father, daughter, mother, son, father-in-law), social role (surgeon, baker, friend), and dramatic role (fool, villain, seducer, etc.). The significance of this type of work lies in capturing the utterance-by-utterance interaction
between speakers and their addressees in terms of sociopragmatic variables, which
enhance the dynamic aspects of the analysed texts.
Working with the same corpus, Archer and Culpeper (2009) develop the notion
of keyness as the identification of keywords, key parts-of-speech and key semantic
fields that are statistically characteristic of the speech of dyads in interaction, e.g.
master/mistress with servant and examiner with examinee in trials. Examples include the use of imperative verbs directed to servants by their masters in the partsof-speech category and the use of the semantic domain of documents and writing in
the speech of the same dyad. These social role dyads are a specific part of particular
social situations, the latter constituting the local context relative to all the social
roles and situations that make up the two genres under examination, i.e. drama and
trial proceedings. Therefore, keyness analysis is a tool for the identification of the
characteristic discourse norms of particular local contexts. While the authors use
the term sociophilology to refer to their approach, their concern for the particular
type of annotation and the analysis of the local conditions of language use indicates
the deployment of a sociopragmatic orientation to their work.
The concept of sociopragmatics has also informed further studies focusing on
the social conditions that affect the use of texts. Adopting the model of critical discourse analysis and frame analysis, Wood (2009) examines personal letters from
late 15th century English taken from the two volumes of the Paston Collection, and
Margaret Pastons letters in particular. Wood (2009: 188) addresses the issue of
authorship, which is important in historical sociopragmatics: since some of Margarets letters were penned by her own sons, might we not be justified in ascribing
the language of those letters to them, rather than representative of Margaret Paston
herself? Supporting the view that local contexts are culture specific, she shows
that local context does not only make reference to a particular interactional dyad,
98
Sophia Marmaridou
but also to a more extended speech community, to the discursive practices of that
community, the texts distribution and its consumption. Within this framework, she
shows that on the one hand these letters evidence the manipulation of power structures within a family and, on the other, they indicate Margaret Pastons control
over the form and content of the letters, thereby confirming authorship.
Viewed broadly, historical sociopragmatics includes the macro-level of social,
socio-cultural and sociological factors as well as the micro-level of personal, situational and stylistic factors. Nevala (2009) discusses the concept of person reference and social deixis in Late Modern English letters and journals. Her data consist
of the correspondence of some members of the Lunar Society of Birmingham and
the letters and journals of Agnes Porter, an eighteenth century governess. In particular, she concentrates on the interactional aspects of the referential term friend
as used by public figures and ordinary people, as well as on the way in which selfand addressee-oriented third person reference is used to convey contextual shifts in
interpersonal distance and authority. She demonstrates that the writers in the data
strategically use different terms in order to alternate between specific social positionings, but they also take into consideration the prevalent social and societal constraints.
Examining the letters subcorpus of the Network of Eighteenth-century English
Texts within a relevance theoretic perspective, Fitzmaurice (2009) focuses on the
sociopragmatic construction of implicature and inference in the illicit courtship
correspondence between Edward Wortley and Mary Pierrepont. She argues that
key historical and cultural reference points are necessary in understanding how
communicative practices are embedded in the local context. Such practices are
typically associated with certain activity types and social rituals including courting. Therefore, the understanding of implicated meanings in these letters makes
reference to the cotext created by the discourse of the letter, the situational context
constructed by the exchange of letters, and the broader historical context and the
social constraints in which the correspondence is embedded. Fitzmaurice (2009)
shows that it is at the discourse cotext that the participants in the interaction enact
their sociopragmatic roles and fight for dominance of their relationship.
It transpires from all the above that the concepts of pragmalinguistics and sociopragmatics have been particularly useful in historical pragmatics and enhance
the possibilities of historical corpus linguistics. Also notice that, within the framework of historical sociopragmatics, pragmalinguistic concerns are also accommodated, as, for example, the study of the forms used to perform a particular function
in Nevalas (2009) work. On this understanding, sociopragmatics is broadly construed as referring to both the micro and macro levels of pragmatic analysis. It is
possible that the success in the application of the two concepts under investigation
to historical pragmatics depends partly on the availability of appropriate corpora
and partly on appropriate retrieval methods that facilitate the analysis.
6.
99
Concluding remarks
As already shown, apart from Leechs (1983) and Thomass (1983) contrast and
distinction between pragmalinguistics and sociopragmatics, these terms have been
variously construed in the relevant literature and, used independently of each other,
they have emphasised the social conditions in which utterances are produced
and understood as verbal contributions in acts of communication. Regardless of
whether the pragmalinguistics/sociopragmatics distinction is theoretically justified, analytically possible, or ideologically desirable, it has motivated the corresponding methodological distinction in L2 acquisition research and L2 teaching
and testing in particular. The concepts of pragmatic failure, pragmatic transfer and
pragmatic development, relevant in these fields of research, explicitly make reference to this distinction. The research questions in the pragmatics of L2 teaching
and testing have to do with defining L2 learners verbal approximations to L2 as
detected in their performance and then attributing them to pragmalinguistic or sociopragmatic sources. The focus is on whether L2 learners utterances are appropriate communicative contributions in the social context in which they occur and
whether their inferencing is based on an understanding of social context. Appropriateness is an essentially evaluative term (see Fetzer 2004; 2007). In the framework of L2 acquisition research it refers to the learners speech actions, which are
constantly assessed as being in accordance with the linguistic norms and contextual constraints and requirements of particular communicative acts in the target language and culture. The appropriateness of L2 learners performance is not always
easy to evaluate on the basis of the pragmalinguistics/sociopragmatics distinction.
More specifically, the absence of sharp boundaries between these two concepts
weakens their methodological usefulness in diagnosing the source of pragmatic
failure, pragmatic transfer, or the stage of L2 learners pragmatic development, especially in the context of L2 testing, as already mentioned. It seems easier to maintain the pragmalinguistics/sociopragmatics distinction in the context of L2 teaching, in the sense that the methodology adopted, the materials used, and the tasks in
which learners are engaged can be designed so as to focus on this distinction and
raise the L2 learners awareness of the linguistic and social issues involved in their
use of L2.
In the area of historical pragmatics and historical corpus linguistics the situation is different. The texts examined belong either to the same period (as in synchronic studies) or to different periods (as in diachronic studies).8 Pragmalinguistic
diachronic studies proceed from form to function, whereas sociopragmatic diachronic studies follow the function-to-form process. Historical sociopragmatics
focuses on the interaction of context and language use either synchronically or
diachronically. In the context of historical pragmatics and historical corpus linguistics the distinction between pragmalinguistics and sociopragmatics can be
fruitfully maintained in spite of the lack of sharp boundaries between the two. His-
100
Sophia Marmaridou
torical texts are viewed as definitive and complete instantiations of a complex configuration of parameters, such as coparticipants and their social, interactional and
discursive roles, communicative action, genre, and cultural norms and strategies of
a speech community. Unlike historical pragmatics research, L2 acquisition research is based on texts as approximations of the target language, which makes it
extremely difficult to diagnose just which specific parameter is at issue in evaluating appropriateness.
On the basis of the above, it is fair to say that, even though the pragmalinguistics/sociopragmatics distinction was initially launched as a level of cross-linguistic application of the general pragmatic principles of cooperation and politeness (as in Leech 1983), and even though it was systematically taken up in the
context of L2 acquisition research, it seems to have been more relevant in the area
of historical pragmatics and historical corpus linguistics as a reliable methodological tool in the investigation of the pragmatics of historical texts. Needless to say,
further theoretical justification of the concepts of pragmalinguistics and sociopragmatics, and of the distinction between them, ultimately depends on future research
in pragmatics and related areas of linguistic enquiry.
Notes
1. It is interesting to note here that Olsens views remind one of Wittgensteins (1958) position that linguistic meaning resides in language use and Wierzbickas (1992) introduction
of a Natural Semantic Metalanguage to deal with the cultural relativity of word-meaning
pairings across languages. Evidently, the dependence of linguistic meaning on use in particular socio-cultural settings is a recurring theme in linguistic theory.
2. In fact, interlanguage, as an interim system of L2 learners, has some features of the L1
and some of the L2, but also features that are independent of their L1 and L2 (see Yule
2006: 167).
3. This is compatible with the view that sociopragmatic awareness is likely to improve pragmalinguistic performance.
4. See also Jucker A. H. and I. Taavitsainen (eds.) 2010. Historical Pragmatics, Vol. 8 of the
series Handbooks of Pragmatics (Berlin: Mouton de Gruyter).
5. Within the narrow, Anglo-American tradition to pragmatics, Traugott (2004: 538) defines
historical pragmatics as a usage-based, pragmatically motivated, approach to language
change.
6. Critical discourse analysis is fundamentally concerned with analysing opaque as well as
transparent structural relationships of dominance, discrimination, power and control as
manifested in language (Wodak 2001: 2).
7. These two text types offer interactive, face-to-face, speech-related data that only approximate authentic discourse. Drama consists of imaginary constructed dialogue and trial
proceedings constitute the record of a prior speech event.
8. Relevant to this point is the distinction between diachronic pragmatics and pragmaphilology (Jacobs and Jucker 1995) and diachronic sociopragmatics and sociophilology
(Archer and Culpeper 2009).
101
Acknowledgements
I would like to thank the three reviewers for their helpful suggestions and comments. I am also grateful to Jonathan Culpeper, Marcelo Dascal, Andreas Jucker,
and Carsten Roever for making their work available to me. This paper has benefited from very useful stylistic suggestions made by Eleni Antonopoulou, to whom
I am greatly indebted.
References
2003
Frankfurt am Main: Lang.
2009
Identifying key sociophilological usage in plays and trial proceedings (1640
1760). An empirical approach via corpus annotation. Journal of Historical
Pragmatics 10/2: 286309.
Arnovick, Leslie K.
1999
Diachronic Pragmatics. Seven Case Studies in English Illocutionary Development. Amsterdam: Benjamins.
Bang, Jrgen Chr. and Jrgen Dr
1979
Language, theory, and conditions for production. In: Jacob L. Mey (ed.), Pragmalinguistics. Theory and Practice, 2150. The Hague: Mouton.
Bardovi-Harlig, Kathleen
1999
Exploring the interlanguage of interlanguage pragmatics: A research agenda
for acquisitional pragmatics. Language Learning 49: 677713.
Bardovi-Harlig, Kathleen and Beverly Hartford
1991
Saying no in English: Native and nonnative rejections. In: Lawrence F. Bouton and Yamura Kachru (eds.), Pragmatics and Language Learning, 4157.
Urbana-Champaign, IL: Division of English as an International Language,
University of Illinois.
Bates, Elizabeth
1976
Language and Context. The Acquisition of Pragmatics. New York: Academic
Press.
Beebe, Leslie M. and Tomoko Takahashi
1989
Sociolinguistic variation in face-threatening speech acts. In: Miriam Eisenstein (ed.), The Dynamic Interlanguage, 199218. New York: Plenum.
Bebee, Leslie M., Tomoko Takahashi, and Robin Uliss-Weltz
1990
Pragmatic transfer in ESL refusals. In: Robin C. Scarcella, Elaine Andersen,
and Stephen C. Krashen (eds.), Developing Communicative Competence in a
Second Language, 5573. New York: Newbury House.
Biber, Douglas and Edward Finegan
1992
The linguistic evolution of five written and speech-based English genres from
the 17th to the 20th centuries. In: Matti Rissanen, Ossi Ihalainen, Terttu Neval-
102
Sophia Marmaridou
ainen, and Irma Taavitsainen (eds.), History of Englishes. New Methods and
Interpretations in Historical Linguistics, 5573. Berlin: Mouton de Gruyter.
Blum-Kulka, Shoshana
1991
Interlanguage pragmatics: The case of requests. In: Robert Phillipson, Eric
Kellerman, Larry Selinker, Michael Sharwood-Smith, and Merrill Swain
(eds.), Foreign/Second Language Pedagogy Research, 255272. Clevedon:
Multilingual Matters.
Bodman, Jean and Miriam Eisenstein
1988
May God increase your bounty: The expression of gratitude in English by
native and non-native speakers. Cross Currents 15: 121.
Bou-Franch, Patricia and Pilar Garcs-Conejos
2003
Teaching linguistic politeness: A methodological proposal. International Review of Applied Linguistics 41/1: 122.
Brinton, Laurel
1990
The development of discourse markers in English. In: Jacek Fisiak (ed.), Historical Linguistics and Philology, 4571. Berlin: Mouton de Gruyter.
Brown, James Dean
2001
Six types of pragmatics tests in two different contexts. In: Kenneth Rose and
Gabriele Kasper (eds.), Pragmatics in Language Teaching, 301325. New
York: Cambridge University Press.
1987
Politeness: Some Universals in Language Usage. New York: Cambridge University Press.
Brumfit, Chris and Keith Johnson
1979
The Communicative Approach to Language Teaching. Oxford: Oxford University Press.
Candlin, Chris N.
1976
Communicative language teaching and the debt to pragmatics. In: Clea Rameh
(ed.), 27th Round Table Meeting, 237256. Washington, DC: Georgetown University Press.
Cook, Haruko Minegishi
2001
Why cant learners of Japanese as a foreign language distinguish polite from
impolite speech styles? In: Kenneth R. Rose and Gabriele Kasper (eds.), Pragmatics in Language Teaching, 80102. New York: Cambridge University
Press.
Culpeper, Jonathan
2009
Historical sociopragmatics: An introduction. Journal of Historical Pragmatics
10/2: 179186.
Dascal, Marcelo
1985
Language use in jokes and dreams: sociopragmatics vs. psychopragmatics.
Language and Communication 5/2: 95106.
Dascal Marcelo and Edson Franozo
1989
The pragmatic turn in psycholinguistics: Problems and perspectives. Theoretical Linguistics 15/1,2: 123.
Eelen, Gino
2001
A Critique of Politeness Theories. Manchester: St Jerome.
103
Faerch, Claus and Gabriele Kasper

1989
Internal and external modification in interlanguage request realization. In:
Shoshana Blum-Kulka, Juliane House, and Gabriele Kasper (eds.), Cross-cultural Pragmatics, 221247. Norwood, NJ: Ablex.
Fetzer, Anita
2004
Recontextualizing Context: Grammaticality Meets Appropriateness. Amsterdam: Benjamins.
Fetzer, Anita
2007
If I may say so: Indexing appropriateness in dialogue. In: Anita Fetzer (ed.),
Context and Appropriateness, 115145. Amsterdam: Benjamins.
Fitzmaurice, Susan M.
2009
The sociopragmatics of a lovers spat. The case of the eighteenth-century
courtship letters of Mary Pierrepont and Edward Wortley. Journal of Historical Pragmatics 10/2: 215237.
Freud, Sigmund
1900
The Interpretation of Dreams. London: Harvester Press, Hassocks.
Freud, Sigmund
1905
Jokes and their Relation to the Unconscious. London: The Hogarth Press.
Fries, Udo
1993
Towards a description of text deixis in Old English. In: Klaus R. Grinda and
Klaus-Dieter Wetzel (eds.), Anglo-Saxonica. Festschrift fr Hans Schabram
zum 65. Geburtstag, 527540. Mnchen: Fink.
Fritz, Gerd
1995
Topics in the history of dialogue forms. In: Andreas H. Jucker (ed.), Historical
Pragmatics. Pragmatic Developments in the History of English, 469498.
(Pragmatics and Beyond New Series 35). Amsterdam: Benjamins.
Grice, Paul
1975
Semantics, Volume 3: Speech Acts, 4158. New York: Academic Press.
Harlow, Linda L.
1990
Do they mean what they say? Sociopragmatic competence and second language learners. The Modern Language Journal 74: 328351.
Haverkate, Henk
1994
The dialogues of Don Quixote de la Mancha: A pragmalinguistic analysis
within the framework of Gricean maxims, speech act theory, and politeness
theory. Poetics 22: 219241.
Holmes, Janet
2008
An Introduction to Sociolinguistics. 3rd ed. London/New York: Longman.
House, Juliane and Gabriele Kasper
1987
Interlanguage pragmatics: Requesting in a foreign language. In: Wolfgang
Lrscher and Rainer Schulze (eds.), Perspectives on Language in Performance, 12501288. Tbingen: Narr.
Jacobs, Andreas and Andreas H. Jucker
1995
The historical perspective in pragmatics. In: Andreas H. Jucker (ed.), Historical Pragmatics. Pragmatic Developments in the History of English. (Pragmatics and Beyond New Series 35), 333. Amsterdam: Benjamins.
104
Sophia Marmaridou
Jucker, Andreas H.
2006
Historical pragmatics. In: Jan-Ola stman and Jef Verschueren (in collaboration with Eline Versluys) (eds.), Handbook of Pragmatics 2006, 114. Amsterdam: Benjamins.
Jucker, Andreas H.
2008
Historical pragmatics. Language and Linguistics Compass 2/5: 894906.
Jucker, Andreas H. and Irma Taavitsainen (eds.)
2008
Speech Acts in the History of English. Amsterdam: Benjamins.
2010
Historical Pragmatics. Handbooks of Pragmatics, Vol. 8. Berlin: De Gruyter.
Jucker, Andreas H., Gerold Schneider, Irma Taavitsainen, and Barb Breustedt
2008
Fishing for compliments: Precision and recall in corpus-linguistic compliment
research. In: Andreas H. Jucker and Irma Taavitsainen (eds.),
2008
Speech Acts in the History of English, 273274. Amsterdam: Benjamins.
Kasper, Gabriele
1992
Pragmatic transfer. Second Language Research 8/3: 203231.
Kasper, Gabriele and Kenneth R. Rose
2001
Pragmatics in Language Teaching. In: Kenneth R. Rose and Gabriele Kasper
(eds.), Pragmatics in Language Teaching, 110. New York: Cambridge University Press.
2002
Pragmatic Development in a Second Language. Oxford: Blackwell.
Keenan, Edward L.
1976
The universality of conversational postulates. Language in Society 5: 6780.
Kohnen, Thomas
2008
Tracing directives through text and time: Towards a methodology of a corpusbased diachronic speech-act analysis. In: Andreas H. Jucker and Irma Taavitsainen (eds.),
2008
Leech, Geoffrey N.
1983
Principles of Pragmatics. London/New York: Longman.
1983
Marmaridou, Sophia S. A.
2000
2006
56/2: 4379.
Mey, Jacob L.
1979
Mey, Jacob L.
1993
Morris, Charles W.
1938
Foundations of the theory of signs. In: Otto Neurath, Rudolf Carnap and
Charles W. Morris (eds.), International Encyclopaedia of Unified Science. Volume 1, 77138.
105
Nevala, Minna
2009
Altering distance and defining authority. Person reference in Late Modern
English. Journal of Historical Pragmatics 10/2: 238259.
Nevalainen, Terttu and Helena Raumolin-Brunberg
1995
Constraints on politeness: The pragmatics of address formulae in Early English correspondence. In: Andreas H. Jucker (ed.), Historical Pragmatics. Pragmatic Developments in the History of English, 5776. Amsterdam: Benjamins.
2003
Historical Sociolinguistics. Harlow: Pearson.
Olsen, Svend Erik
1979
Psychopathology, interaction, and pragmatic linguistics. In: Jacob L. Mey
(ed.), Pragmalinguistics. Theory and Practice, 233253. The Hague: Mouton.
Olshtain, Elite and Liora Weinback
1992
Interlanguage features of the speech act of complaining. In: Gabriele Kasper
and Shoshana Blum-Kulka (eds.), Interlanguage Pragmatics, 195208. New
York: Oxford University Press.
Prucha, Jan
1983
Riley, Philip
1979
Towards a contrastive pragmalinguistics. Papers and Studies in Contrastive
Linguistics 10: 5778.
Robinson, Mary Ann
1992
Introspective methodology in interlanguage pragmatics research. In: Gabriele
Kasper (ed.), Pragmatics of Japanese as native and target language Technical
report no. 3: 2782. Honolulu: University of HawaiI at Manoa, Second Language Teaching and Curriculum Centre.
Roever, Carsten
2006
Validation of a web-based test of ESL pragmalinguistics. Language Testing
23/2: 229256.
Rosaldo, Michelle Z.
1982
The things we do with words: Ilongot speech acts and speech act theory in philosophy. Language in Society 11: 203237.
Rose, Kenneth R.
2000
An exploratory cross-sectional study of interlanguage pragmatic development.
Studies in Second Language Acquisition 22: 2767.
Searle, John R.
1979
Selinker, Larry
1972
Interlanguage. International Review of Applied Linguistics 10: 209230.
Sifianou Maria
1992
Politeness Phenomena in England and Greece. A Cross-Cultural Perspective.
Oxford: Clarendon Press.
Taavitsainen, Irma
1995
Interjections in Early Modern English: From imitation of spoken to conventions of written language. In: Andreas H. Jucker (ed.), Historical Pragmatics.
Pragmatic Developments in the History of English, 439465. Amsterdam:
Benjamins.
106
Sophia Marmaridou
Takahashi, Satomi
1996
Pragmatic transferability. Studies in Second Language Acquisition 18: 189
223.
Tahahashi, Satomi
2001
The role of input enhancement in developing pragmatic competence. In: Kenneth R. Rose and Gabriele Kasper (eds.), Pragmatics in Language Teaching,
171199. New York: Cambridge University Press.
Thomas, Jenny
1981
Pragmatic failure. M.A. dissertation, Department of Language and Linguistics, University of Lancaster.
Thomas, Jenny
1983
Cross-cultural pragmatic failure. Applied Linguistics 4/2: 91112.
2004
Historical Pragmentics. In: Lawrence R. Horn and Gregory Ward (eds.), Handbook of Pragmatics, 530561. Oxford: Blackwell.
Trosborg, Anita
1987
Apology strategies in natives/non-natives. Journal of Pragmatics 11: 147167.
Widdowson, Henry G.
1979
Explorations in Applied Linguistics. Oxford: Oxford University Press.
Wierzbicka, Anna
1987
English Speech Act Verbs: A Semantic Dictionary. Sydney/New York: Academic Press.
Wierzbicka, Anna
1992
Semantics, Culture, and Cognition. Universal Human Concepts in CultureSpecific Configurations. New York/Oxford: Oxford University Press.
1958
Wodak, Ruth
2001
What CDA is about. A summary of its history, important concepts and its
developments. In: Ruth Wodak and Michael Meyer (eds.), Methods of Critical
Discourse Analysis, 113. London: Sage Publications.
Wolfson, Nessa
1981
Compliments in cross-cultural perspective. TESOL Quarterly 15: 117124.
Wood, Johanna L.
2009
Structures and expectations. A systematic analysis of Margaret Pastons formulaic and expressive language. Journal of Historical Pragmatics 10/2: 187214.
Yule, George
2006
The Study of Language. 3rd ed. Cambridge: Cambridge University Press.
4.
Metapragmatics
Axel Hbler
Introduction
The term metapragmatics is slippery, many meanings have been ascribed to it,
too many perhaps. A restrictive or selective attitude, however, does not seem appropriate for a handbook article. On the contrary, the survey should accommodate
as many perspectives as are possible and justifiable. The problems mainly arise
from two interrelated facts. (1) The Greek loan prefix meta- has different meanings
(in between, after, later or beyond, among others) and thus lends itself to forming
all sorts of terms. (2) The formative meta- is popular. Wordspy, the internet window
administered by Random, lists 75 new meta-formations.1 Its attraction presumably derives from its particular air of something intellectual, detached and with an
impressively broad view.
To reduce the risk of a sell-out of this formative device, it may be prudent to
control its deployment and avoid it where it is not necessary. A case in point is the
use of the term metapragmatics, where the term pragmatics actually suffices. If we
define pragmatics as the discipline which has interpersonal communication as its
object, then it goes without saying that all abstractions and theoretical conclusions
which help explain communicative behavior form part of this discipline, in fact
they are defining features of the discipline. This holds true for all the answers given
to questions such as how to cooperate, how to be kind, polite, etc. and all the
conditions found to explain aspects of interpersonal communication, i.e., the conditions of thinkability (e.g., the constitutive rules of Searles speech acts), the
conditions of feasibility (e.g., the Gricean maxims) and the conditions of recognizability (e.g., the linguistic and paralinguistic expressive means) (cf. Caffi 2006:
84). To separate such theoretical stances or constructs from the discipline that developed them seems to make little sense. Communication can be act and object,
as Hagemann (1997: 32) subtly observes, pragmatics cannot. It is certainly true
that communication shows properties that can be labeled pragmatic, but such
labeling originates in an external view. This then is a reading of metapragmatics
that deserves mentioning,2 but will not be part of the focus of the article; the theoretical views and notions as such, however, will, of course, be used where necessary.
What remains to be unpacked are the following four readings of the term and
the four corresponding forms of linguistic practice: metapragmatics as the study of
explicit metacommunication (section 1); metapragmatics as the study of implicit
metacommunication (section 2); metapragmatics as the study of peoples abstract-
108
Axel Hbler
ing from interacting (section 3.1); and metapragmatics as metatheory of pragmatics (section 3.2).
1.
Metapragmatics as the (pragmatic) study of explicit

metacommunication
The object of a (meta)pragmatic analysis is metacommunication. In a first go, it

can be defined as communication about (selected aspects of) communication.
Usually, the communication talked about is neither communication as such (the
abstract and general possibilities and impossibilities of communication) nor some
definite previous or future communication, but the ongoing communication. Paul
(1999: 98) characterizes it immanently as practical speech reflection. Quite original is Antons view contrasting it with communication.
[W]e routinely and commonly attend not to speech but simply from it. That is, when we
listen with [sic!] one another in our everyday encounters, we routinely listen from our
speech to the thought so intended. (Anton 1998: 198199)
This behavior contrasts with what happens in particular circumstances, where

we explicitly attend to the speech itself. In Leders [1990] terminology, moments of
breakdown manifest speech dys-appearance.
Reconsider the case of reading a novel: I may be engrossed in the story, simply attending from the sentences to the story, but then, a misspelled word, a foreign word, or perhaps a too-recondite one appears. Now, I consciously and explicitly reflect to the word
[]. Speech, then, is routinely an absent body, an intentional arc which disappears for
the sake of the meaning so intended. (Anton 1998: 199)
What Antons circumscription brings out nicely is a kind of break separating communication from metacommunication, a sort of strained dialogical structure, which
is not attributable to a change in speakers (or voices), but to a change in the level of
communication. This is mainly brought about by a functional change from practical acting to reflecting.3
The change in level is often notable in that it involves a change from ordinary
language to metalanguage. Metalanguage is commonly understood as the language
referring to language; it represents what Verschueren (1998: 55) calls the object
notion of metalanguage.4 It is, however, not an autonomous language (not even
where linguists use a metalanguage). When members of a speech community engage in metacommunication, the metalanguage used does not consist of complicated terminology but of ordinary words, mainly nouns (such as chat or denial) and
verbs (such as promise or argue), whose feature that distinguishes them from other
words is their function to refer not to something in the world out there but to some
aspect of speech.5
Metapragmatics
1.1.
109
On full forms of metacommunication
Though metacommunication is always embedded in primary communication, its

extension can vary considerably. For the purpose of this article, it will be sufficient
to simply differentiate between full and reduced (or abbreviated) metacommunicative forms. A full form consists of at least one metacommunicative utterance;
anything below utterance level can no longer be considered metacommunication
proper, but is, of course, still metacommunicative.6 If primary communication is
the on-line event, then metacommunicative utterances (and more extended forms)
cause such an articulate deviation from it that they appear as off-line, as it were
(cf. Hbler and Bublitz 2007: 12). The boundary between primary communication
and metacommunication is usually not marked in strictly formal terms (except for
those cases in which a speaker refers to some previous utterance within the current
communicative situation by using some speech reporting frame (I said , you
said ); for references to utterances outside the current situation, cf. below). The
shift in level is identified, rather, in pragmatic terms.
The following questions are meant to delineate the main pragmatic parameters
that apply for analyzing metacommunication. They pertain to the issue/topic
chosen, to the intention pursued, and to some frame conditions of metacommunication (for a more detailed treatment and examples of full-length analyses, cf. Bublitz and Hbler (eds.) 2007).7
(1) What are the topics of metacommunication? Though always referring to
particular speech events, either single or serial, the topical scope can be generic as
well as specific. Metacommunicative topics of a more generic kind may, for instance, concern the problem of effability (I dont have words which could adequately express how I feel or Love is just a gross label), the mode chosen (You
dont have to touch me when you want something from me or Could you speak up a
bit?), or questions of participating in a conversation (Why do you always want to
dominate conversation? or You never take part in this kind of talk). They may be
comments that operate with general principles, norms and maxims; thus Grices
maxims find rich application (e.g., the maxim of quality in Are you telling the
truth?, the maxim of quantity in Just give us the gist of it, the maxim of relevance
in Come to the point, please! and the maxim of manner in Let me just finish this
story before we leave) as well as Leechs politeness principle (Stop being so patronizing!).
More specific metacommunicative topics may concern illocutionary functions
(Is that a complaint?), acts of reference (Who are you talking about?) or predication (Thats a very euphemistic formulation or What do you mean by insinuation?). Or they may concern matters of discourse organization. These include
performing or holding back certain speech acts (Give me your permission to do so,
please), announcing a speech act (And now I will disclose my next plans), or steering an ongoing discussion (Weve got three requests to speak; first Lila, then John,
110
Axel Hbler
then Karen). Comments relating to the text structure (such as But the real sensation
is coming now) may occur within longer turns.
(2) What for or why? These are the central questions for any pragmatic approach to communication. The purposes that metacommunicative utterances in
particular pursue are in part prefigured by their topic.8 Thus they concern the organization of discourse, including the textual structure of longer stretches of utterance (So far we have been describing a state of affairs; now we may turn to );
they help secure an adequate understanding, including attempts at rendering a
speech act function more precise (Is this a complaint or an accusation?) and other
kinds of monitoring. Their sense is to prevent or repair misunderstandings.
Another common purpose of metacommunicative comments is evaluative,
mainly critically, in nature, in as far as what is topicalized marks or implies a
negative deviance from what is to be expected (But I am repeating myself). Where
such comments are self-critical, they often represent a preventive measure anticipating criticism from others.
Besides general communicative norms and principles, which provide the standard for (metacommunicatively) evaluating a given utterance, other standards
seem to have been availed of recently as well. They are based on what has become
known under the label of critical language awareness (cf. Fairclough ed. 1992),
which Coupland and Jaworski (2004) characterize (rather critically in tone) thus:
The growing obsession with good communication (Cameron 2000) is filled by a communication industry happily producing prescriptive orders of metadiscourse which are
constitutive of new social orders. These meta-discourses are based on various regimes
of verbal hygiene (Cameron 1995) and codifications of communication skills. (Coupland and Jaworski 2004: 39)
The underlying intentions are certainly honorable: to contribute on the mental

level to the social goal of emancipation.
Farther reaching objectives, i.e., objectives that go beyond the associative link
with the communicative aspect topicalized, may consist in securing the participants receptivity or even in raising conflict, in a constructive or destructive key.
The last example (But I am repeating myself) and its variant (You are repeating
yourself) can well illustrate these purposes. And in a still wider perspective, such
metacommunicative comments may be used for defending or questioning communicative norms (the speaker subscribing to a culture-specific norm of brevity
and conciseness) or even creating/modifying personal identities as communicators
(profiling him- or herself as a critical, assertive person with a social status that entitles him or her to utter things of that sort).
(3) Who is the target? Though comments on utterances may appear completely
depersonalized, the speaker of the utterance commented on is nonetheless the ultimate target of such comments. In principle, any participant in a communication
can be the target; we thus can roughly differentiate between comments directed at
Metapragmatics
111
oneself and comments directed at some other speaker (cf. Hbler and Bublitz 2007:
1516).
In principle, metacommunicative actions are reciprocal and egalitarian,
wherever symmetrical relations between participants in an interaction are given
or should be given (cf. Habermas 1971). There are interactive frames, however,
which do not allow for symmetry; in the classroom or the courtroom, for example,
the rights for metacommunicative utterances are asymmetrically distributed. This
pertains, of course, primarily to the right of commenting on the contribution of the
other party. But even self-commenting may show restrictions, especially where the
possibility for a certain communicative move (such as changing the topic, or
launching into a longer discourse) is not readily available; one may have to choose
the form of asking for permission to add something not really to the point instead
of simply announcing that one is going to do that.
(4) What are the forms of realization? The examples illustrating the various
points have already provided a variety of examples of how a speakers metacommunicative intentions can be realized. Since any metacommunicative activity is
based on a reflexive and diagnostic analysis, a commenting utterance seems to be
the most direct equivalent, either a categorical (I am talking too much or You are
talking too much respectively) or a modalized one (Do I talk too much? Maybe, I
am talking too much or I think you are talking too much respectively). Another option is to operationalize the analysis in terms of a corresponding action to be taken
(I should not talk that much or Stop talking that much, Could you stop talking too
much? respectively). Where the target is not the speaker him/herself, such metacommunicative utterances, especially if they are critical, are likely to provoke
some reaction (Thats ok or Im always supposed to keep my mouth shut!), and this
could even lead to extended sequences.
As to the position of metacommunicative clauses, it is most common that they
follow the utterance that they refer to. Only where the speaker takes him/herself as
target may we find a reversed order, in which the utterance referred follows. This
restriction, of course, is not surprising; it is, after all, only in this circumstance that
the speaker of a metacommunicative utterance knows what it will be that s/he (cataphorically) refers to.
A systematic treatment of how the various aspects interact would certainly be
desirable, but is currently not feasible.
Certain meta-phenomena have so far not been taken into account for good
reasons: Though they consist of reflexive language (in the sense of Lucy 1993) and
though they contribute to the ongoing communication, they do not result in metacommunication. Their common denominator consists in their being reportive in
some way or another, i.e., they refer to instances of direct speech outside the current communicative situation and either reproduce them or sum them up.9
The most common forms are direct and reported speech, which are usually
framed (introduced or rounded off) by a phrase containing a verb of saying (He
112
Axel Hbler
said, or She told me that ), a performative verb (He promised to stop smoking), or one of a more interpretative character (He insisted that ). They lack, to
quote Meyer-Hermann (1978: 128), the defining feature that the object referred to
and predicated about is a communicative interaction (sequence) or part belonging
to the same interactional unit as the metacommunicative speech act. These metarepresentations, as Wilson (2000) calls them, refer to speech events outside a
given communicative situation and can therefore be considered extracommunicative (cf. Ungeheuer 1970). Due to their extracommunicative character, these
speech references get absorbed in the primary discourse either as instrument or object. Newspaper articles are a common source for illustration; examples of another
kind are literary analyses or philosophical language essays.
A phenomenon that is not extracommunicative and still represents hardly an instance of metacommunication is the so-called echo-utterance, where the speaker
echoes the preceding speaker by repeating some linguistic material yet giving a
specific turn to it; cf. Sperber and Wilson (1986), Wilson (2000); Graf (2007).
Echo-questions (Youve got frequent-flyer status with which company?) try to reelicit a certain part of the preceding utterance that the addressee either did not
understand or feels provoked by. Echo-statements such as in the following example
He: Its a lovely day for a picnic
[They go for a picnic and it rains]
She: (sarcastically): Its a lovely day for a picnic, indeed.
(Sperber and Wilson 1986: 239)
usually just convey attitudes10 towards the propositional state of affairs quoted/
echoed. In neither case are we confronted with instances of metacommunication;
the echo-utterances are information-oriented (in that they elicit information or
evaluate information) and do not concern some formal aspect of communication.
Where they cause a break, as in the information-seeking reading of the echo-question above, they do not transpose the ongoing communication onto the meta-level.
Another quotative variant, which is not extracommunicative either but functions within an ongoing communicative situation, is at least a strong candidate for
metapragmatics. The reference is to parody, which is defined as imitating the characteristic style of a speaker (or an author or a work of his/hers) with the intention to
ridicule or achieve a comic effect. We usually associate a literary genre with it, but
parody also occurs in ordinary life and everyday conversation. The illustration that
follows is taken from a piece of literature, but the parodist here is not the author
himself but one personage in his (realistic) drama.
Nick:
George:
Nick:
George:
Nick:
George
I try not to
Get involved. Um? Isnt that right?
Yes thats right.
Id imagine not.
I find it embarrassing.
(sarcastic): Oh, you do, hunh?
Metapragmatics
Nick:
George
113
Yes. Really. Quite.

(mimicking him): Yes. Really. Quite. (Then aloud, but to himself:) Its disgusting!
(Edward Albee, Whos Afraid of Virginia Woolf? New York: Pocket
Books, 1966: 90)
The parodistic mimicking performs the metacommunicative act inexplicitly, but

could be made explicit as is illustrated in the given example by the aside with
which Georges turn ends; both modes convey his disgust at the way in which Nick
expresses himself. As is also shown by the example, parody at least in everyday
conversation where it is used only occasionally shows something similar to, but
weaker than, the break which marks the boundary between communication and
metacommunication: the ongoing dialogic flow comes to a halt. With Bally (1965),
we could characterize this inexplicit type of metacommunication as mode vcu,
contrasting with the mode pur of explicit metacommunication (cf. below).
Stylization, as treated by Coupland (2004), and some forms of irony provide
other instances of such inexplicit metacommunication. They represent doublevoiced utterances in the sense of Bakhtin (e.g., 1981) in that they are moving on
two different levels (the level of primary communication and the level of metacommunication, which consists in exaggerated imitation) and yet are united in one
utterance.
1.2.
On abbreviated forms of metacommunication
The subtitle already links this part directly to section 1.1 where full forms of metacommunication were tackled. The same (meta-)pragmatic aspects are applicable
here as well. They will, consequently, be expounded by the same guiding questions; merely the order will be slightly altered, starting this time with the most distinctive feature, i.e., the formal make-up.
(1) What are the forms of realization? The means allowing for abbreviated
metacommunication are quite easy to delineate. They mainly consist in adverbials
that typically operate on the whole proposition of the given utterance. The most
concise forms are disjuncts (such as frankly); slightly more voluminous are prepositional phrases/clauses (such as in short), infinitive clauses (such as to tell you in
a few words) and participial clauses (such as bluntly speaking). They usually can
be expanded to a fully fledged metacommunicative utterance (Ill be very frank
with you; I will sum it up in a few words; I dont want to fuss around but say bluntly
what I have in mind). Being briefer than their elaborated counterparts, they are
better integrated into the dialogical flow; the communicative break they cause is
more subtle, and at best syntactically marked off from the main clause providing
the propositional content of what is primarily communicated. While the full forms
of metacommunication were characterized as off-line phenomena (seen from the
perspective of ongoing primary communication), the abbreviated variants keep the
114
Axel Hbler
communication on-line, in a manner of speaking. It is therefore no wonder that

they are unlikely to open up metacommunicative (side-)sequences.
(2) What are the topics that the adverbials cover? Quirk et al.s (1985) classifications of adverbs/adverbial provide a comprehensive overview. Of prime relevance are style disjuncts and with some restrictions conjuncts.
According to Quirk et al., style disjuncts convey the speakers comment on
the style and form of what he is saying, defining in some way under what conditions he is speaking as the authority for the utterance (Quirk et al. 1985: 615).
The metalinguistic character can (often) be brought forth explicitly by transforming the disjunct into a clause in which the adverbial (turning into a process adjunct)
is linked to a verb of speaking with I as subject. (e.g., Frankly, I am tired I tell
you frankly, I am tired).11
In line with Quirk et al. (1985: 615 ff.), we can subclassify style disjuncts. One
group of adverbs that is of interest focuses on the manner in which an utterance is
made; with their help, a speaker can either refer to the style of his/her utterance
(e.g., bluntly, briefly, simply) or include some modal overtones (e.g., truthfully).
Recall that longer variants are also available, such as prepositional phrases, or
(in-)finitive and participial clauses (cf. above). The other group of interest consists
of adverbs (and adverbial expressions) which characterize the make-up of a formulation, whether it has to be taken, say, figuratively or literally. Again, some expressions may be tinged with modal overtones. Thus, an utterance like Hawkins
was not, strictly speaking, a traitor may address not only the issue of whether the
word traitor is a suitable term for Hawkinss behavior but also the issue of whether
or not Hawkins is a traitor. In addition, there are some adverbs of degree (belonging to the category of subjuncts) that can serve the same metalinguistic purpose:
compromisers (such as kind of, sort of) and approximators (such as almost, virtually) provide indications as to the reliability and adequacy of the lexical expression chosen. In uttering, for example, He is kind of a traitor or He virtually
stole the money the speaker may want to let the listener know that the verdictive
terms traitor and steal respectively are not to be taken literally, but come close.
These adverbials can, in general, be linked up again with some conversational
principle, norm or maxim. A speaker using truthfully in his/her utterance echoes
Grices maxim of quality, using briefly his maxim of quantity, while with figuratively he alludes to his maxim of relevance (by flouting it), and with frankly to his
maxim of manner.12 Adverbs like sort of pay tribute to Leechs politeness principle
in as far as they may render the utterance more palpable.
The second set of adverbials that in part can be interpreted in a metacommunicative key are conjuncts/conjunctions. They are text-structural markers that
specify how an utterance fits into the surrounding discourse in which what is to
follow is systematically connected to what has gone before (Halliday and Hasan
1976: 227). Not all of the 7 conjunctive roles (and sub-roles) between which Quirk
et al. (1985: 634 ff.) distinguish (i.e., listing, summative, appositional, resultive, in-
Metapragmatics
115
ferential, contrastive, transitional) are metacommunicative. The ones that do not

qualify for inclusion are those that establish the co-textual link purely in terms of
content, not of form. The ones whose use allow for a pragmatic interpretation as
being metacommunicative are adverbs of enumerative listing (such as first, second, third; to begin with, next, to conclude), summative expressions (such as in
conclusion, to summarize), contrastive adverbs with a reformulatory (such as
more accurately, alias, in other words) or replacive meaning (as in alternatively,
rather; better, worse) and finally, adverbials, such as incidentally or by the way,
which mark discoursal transition.13
The preceding classification of metacommunicative adverbials shows more
openly perhaps than the treatment of full metacommunicative utterances does a
clear-cut topical dividing line between personal and textual thematic foci. With respect to narratives, McNeill (1992: 185 ff.) differentiates in my opinion quite felicitously between two levels outside the narrative proper: the meta-level, which
includes the language clues characterizing and structuring the narrative as text, and
the para-level which relates to the narrator listener axis, comprising elements of
an interpersonal nature (e.g., providing orientation to the listener, comments and
attitudes of the narrator etc.) Applied to communication in general, we could distinguish between a meta-communicative level that would pertain to the structural
aspects of dialogues and turns and the para-communicative level that would account for those non-primary forms of communication which are person-centered
and focus on the interactional aspects involving the speaker and/or hearer.14 More
recent treatments draw similar distinctions, though less categorically. del (2006)
starts out from the notion of metadiscourse embracing various configurations that
enfocus either the text, the writer (or speaker), and the reader (or hearer) and distinguishes in a second step between metatext, which relates to the text or code
proper, and writer-reader interaction. Also Hyland (2005) uses metadiscourse
as cover-term; he defines it as basically interpersonal and then draws a functional
distinction between means that help to guide the reader (or listener) through the
text (interactive function) and those that involve the reader (or listener) in the text
(interactional category). Some such terminological differentiation could help to
disentangle the complexity of the concept at issue.
(3) The question as to who is the target (and who has got the right to freely select the target) is easy to answer for the adverbials with a metacommunicative
function. Where a speaker uses style disjuncts, s/he has necessarily him/herself as
target in that his/her metacommunicative comment refers pro- or retrospectively to
his or her own primary conversational contribution. Whereas the speaker as metacommunicative target is oftentimes quite transparent with such style disjuncts (cf.
frankly speaking), s/he seems to withdraw from being target where conjuncts that
are reflexive of discourse structure come into play. The sociological question concerning the right to select the target, in this case, boils down to the question of
whether any speakers in any communicative situation may take the opportunity to
116
Axel Hbler
comment on him- or herself as communicator in such brief (adverbial) terms. This

is such an elementary (and minimalistic) question that it can either be answered in
strictly categorical moral terms along the line of Habermas or in empirical terms on
the basis of rich data sets.
(4) The final question to face concerns, again, the raison detre, the why or
what for? The purposes that a pragmatic view can ascribe to these reduced forms of
metacommunication cannot be expected to be essentially different from those
identified for the extended forms with the speaker as target. But they can be expected to be somewhat reduced, according to the iconic principle of a proportionality between quantity of form and quantity of meaning. This seems, indeed, to be the
case. In one way or another, they help the speaker optimize the ongoing primarylevel conversation. The style disjuncts serve him/her as (prophylactic) lubricant to
avoid clashes with interactants (and subsequent repairs). The textual conjuncts facilitate the reception of the primary message emitted.
2.
Metapragmatics as the study of implicit metacommunication
Auer (1986: 22) has hailed it as a new research paradigm and Verschueren (1998:
60) holds that it is the central area, the proper domain of metapragmatic studies:
peoples metapragmatic communicative behavior based on their metapragmatic
awareness of situation and context as co-determinants of communication. No specific explicit repertoire is available to meet this meta-function. The approach conceives metalanguage as a dimension of language use and does not relate it to the
object notion of metalanguage (Verschueren 1998: 55); cf. section 1. What we look
for and find in utterances, instead, are indicators that testify to a metapragmatic
awareness of speakers, a kind of self-monitoring activity which constantly calibrates (to use a Silversteinian term) the communicative intention with situational
and contextual conditions. The significance is highlighted by Silverstein, to whom
Verschueren repeatedly refers in his argumentation.
Without a metapragmatic function simultaneously in play with whatever pragmatic
function(s) there may be in a discursive interaction, there is no possibility of interactional coherence, since there is no framework of structure here, interactional text
structure in which indexical origins or centerings are relatable one to another as aggregated contributions to some segmentable, accomplishable event(s). (Silverstein
1993: 3637)
The corresponding type of reflexive language are indexical forms, which change
their value depending on the actual event of speaking (Lucy 1993: 10). The specific segment of reflexive language consists of shifters, contextualization cues
(prosodic patterns, code switching, etc.) or implicit voices (cf. Verschueren 1998:
61).15
Metapragmatics
117
Communicatively speaking, there is, consequently, nothing that could be paraphrased in explicit metacommunicative terms (cf. 1.2). And likewise, no neat
break or boundary between a primary and a secondary level of communication can
be found, and no turn to off-line communication. But at least one metacommunicative feature is preserved: a reader/listener will be lead to attend to the words instead of attending from them, to recall Antons (1998) subtle play with prepositions
(quoted in section 1).
The attending to words, more specifically to indexical words and their metapragmatic dimension, shall be further elaborated and illustrated in two points. The
first point concerns the so-called shifters themselves, e.g., personal pronouns, deictic expressions, or tenses whose meanings shift due to their dependency on the
individual single text and situation.16 The pronoun you, for example, acquires its
specific referential meaning according to the communicative situation. This situation-dependent meaning assignment then is the result of the metapragmatic reasoning, on the speakers as well as the receivers side. More interesting and complex are such metapragmatical assignments in the case of some text-cohesive
devices, i.e., reference, substitution, or ellipsis. Halliday and Hasan (1976) provide
characterizations in which the metapragmatics behind their use and understanding
is (unvoluntarily) well highlighted. All three are forms of presupposition, devices
for identifying something by referring it to something that is already there known
to, or at least recoverable by, the hearer (Halliday and Hasan 1976: 144).17 With
respect to reference, we read that [b]oth exophoric and endophoric reference embody an instruction to retrieve from elsewhere the information necessary for interpreting the passage in question (Halliday and Hasan 1976: 33). More concretely,
a reference item signals supply the appropriate instantial meaning, the referent in this
instance, which is already available (or shortly to become available); and one source of
its availability is the preceding (or following) text. (Halliday and Hasan 1976: 227)
In substitution and ellipsis (as a special case of substitution), the presupposition is

at the level of words and structures. When a substitute is used, it signals that the
actual item required, the particular word or group or clause, is recoverable from the
environment (Halliday and Hasan 1976: 145).
The second point pertains to contextualization cues, whose metapragmatic dimension is as Gumperz (1982: 131) states rarely consciously noted. He gives
the following definition:
[A] contextualization cue is any feature of linguistic form that contributes to the signaling of contextual presuppositions. Such cues may have a number of such linguistic realizations depending on the historically given linguistic repertoire of the participants. The
code, dialect and style switching processes, some [] prosodic phenomena [] as well
as choice among lexical and syntactic options, formulaic expressions, conversational
[] strategies can all have similar contextualizing functions (Gumperz 1982: 131).
118
Axel Hbler
Among his examples, we find one where a graduate student is supposed to interview a black housewife in a low income, inner city neighborhood. After an appointment has been arranged over the phone by someone from the university, the
student goes to the black home and is met by the husband, who opens the door,
smiles and says to him So yre gonna check out ma ol lady, hah? What he utters is
a formulaic phrase identifiable through co-occurrent selections of phonological,
prosodic, morphological and lexical features (Gumperz 1982: 133); it is marked
for ethnicity and signals both expectations about what is to be accomplished and
about the form that replies must take (Gumperz 1982: 134). That such expectations exist comes best to the fore where they are not met by the interactant and
subsequently cause irritation, break-down or the like. In the example, the student
does not meet the husbands expectations by (inappropriately) replying Ah, no,
I only came to get some information. They called from the office. Thereupon the
husband dropped his smile and disappeared without a word while calling his
wife.
Sometimes, the metapragmatic reasoning behind an utterance may not be interactionally oriented, but may rather serve the maintenance and construction of identity (cf. Tajfel 1982). Identity is not only to be understood as a fixed set of properties or operations residing in the individuals cognitive make-up, where language is
regarded as a largely docile medium through which dimensions of a persons
identity [] may receive occasional expression (Wooffitt and Clark 1998: 107); it
is also (and above all) an accomplishment with respect to interactional and inferential concerns generated by the trajectory of verbal exchanges [], [whose] ascription is inextricably tied to the details of talk-in-interaction (Wooffitt and Clark
1998: 107). Take the utterance Hang on, love. The term of address used could be
indicative of an informal (cross-gender) relationship between speaker and addressee. In the actual instance, however, the addressee is not the speakers wife, but
a (female) mediator the couple has turned to in their divorce case. The speaker here
constitutes his relationship with [the addressee] in a way which is at odds with the
mediator-disputant relationship they have thus far collaboratively sustained
(Greatbatch and Dingwall 1998: 128). With this utterance, the speaker is questioning the addressees professional identity.
What the metapragmatic devices under consideration all have in common and
whereby they differ from the reflexive types taken into consideration in the preceding two sections is the mode of reflection. Bally (1965) distinguishes between the
mode pur and the mode vcu; it is the latter one which we have been confronted
with in the present section, whereas we were concerned with the former one previously. Ballys distinction is worth quoting in greater detail.
Si le langage nest pas une cration logique, cest que la vie dont il est lespression na
que faire des ides pures. Si lon me dit que la vie est courte, cet axiome ne mintresse
pas en lui-meme, tant que je ne le sens pas, tant quil nest pas vcu [] (Bally 1965:
15)
Metapragmatics
119
If language is not a logical creation, it is due to the fact that life, whose expression it
represents, does not simply produce pure thoughts: If I am told that life is short, I am not
interested in this idea unless I feel it, unless I experience it.
The following illustration fits well in our context, since the Italian pur(e) is a discourse marker.
Le jugement intellectuel La terre tourne se change en jugement de valeur dans la bouche
de Galile seciant devant ses juges: E pur si muove! Ce nest plus une vrit scientifique, cest laffirmation dune valeur attache cette vrit [] (Bally 1965: 15)
The logical judgement The globe revolves becomes an evaluative judgement in the
mouth of Galileo screaming into the face of his judges E pur si muove! Then it is no longer
a simple scientific truth, rather it is the affirmation of a value attached to this truth.
Uncovering the metapragmatic dimension of language and language use in indexical forms is the objective of the corresponding field of scientific activity, which, in
turn, derives its name from the very objective. Strictly speaking, metapragmatics is
in this sense then the study of the metapragmatic dimension of language and
speech rather than the study of metacommunication. If we were to continue in the
attempt at terminologically distinguishing the various conceptualizations of metapragmatics, the term syncommunicative could well serve this purpose; the prefix
syn- (of Greek origin like the prefixes meta- and para- in the complementary expressions) conveys quite adequately the fact that this reflexive level is not isolatable but inextricably linked to the primary level of communication.18
3.
Beyond interacting
Leaving ongoing conversations and taking a more generic view instead opens up
perspectives to be dealt with now. Observing communication in its general design
and different patterns and describing and systematizing the observations is the perspective that constitutes a first-level step away from conversation; theorizing about
the conditions, presuppositions and implications of such endeavors constitutes a
second-level step. While the second level is reserved for the professional linguist
(cf. 3.2), the first level is open to both the professional and the ordinary conversationalists (cf. 3.1).
3.1.
Metapragmatics as the study of peoples abstracting from interacting
Through their participation in all sorts of conversations, members of a speech community gather in the course of time knowledge about conversation in general, their
genres, patterns, styles, norms, etc. Since this knowledge relates to the pragmatic
dimension of communication, it may be specified as metapragmatic knowledge.
Some similar metapragmatic knowledge may be obtained by the scientist who
120
Axel Hbler
studies the conversations of this speech community as an outsider yet in the role of
a participating observer. In other words, it is hardly possible to clearly distinguish
between the scientific metapragmatic knowledge and the metapragmatic knowledge of the users of a language. The assumption of a continuum between both
stances, as made by Caffi (2006: 87), might therefore seem quite appropriate;19 the
difference consists merely in a smaller or greater distance from the object, viz.: the
use of the language and a lower or higher degree of explicitness.
Folk theories, however, call for the mediating linguist, the linguist, that is, who
elicits folk knowledge or elaborates on their shorthand versions. Gumperz (1982)
identifies as mediating linguists in particular those working in the fields of discourse analysis or ethnography of communication. Folk-linguistics is a separate
discipline (cf. Niedzielski and Preston 2000), but has roots in both. In any case the
task is to come to grips with the varying ways in which linguistic behavior is conceptualized by those engaged in it (Verschueren 1998: 6061). Of the folk-linguistic topics, speakers awareness of and knowledge about pragmatic issues is, of
course, particularly relevant for the current context. Two areas and modes (cf. Preston 1996) shall be considered in greater detail.
(1) A major source of information about peoples metapragmatic knowledge
could be the lexical repertoire they have at their disposal for referring to aspects of
speech; it is not an autonomous language, but a subset of the ordinary language.
Speech act verbs are a favorite research area (where folk taxonomies merge with
the empirical-conceptual approach of speech act theories and the ethnography of
speaking, cf. Caffi 2006: 85). But other lexical expressions relating to other selected aspects of communication could also be common ground where folk- and
empirical-conceptual linguistics meet. Lexical expressions, in general, can be interpreted as a means that a speech community uses for coding cognitive and cultural models, which consist in more or less coherent sets of concepts for structuring
experience (cf. also Gee 1999: 40 ff.). Experience also includes acting by communicating, and this, in turn, calls for adequate conceptualizations of a metapragmatic kind and their storage in words.
One of the most comprehensive collections of English speech act verbs is the
dictionary compiled by Ballmer and Brennenstuhl (1981).20 Its comprehensiveness
derives from two decisions. (a) They apply a much wider definition of speech act
verb by including all those verbs which designate (aspects of) speech activities
(Ballmer and Brennenstuhl 1981: 3), no matter whether they can be used performatively or not; thus lie and persuade, for example, are speech act verbs just as
well as admit or promise. (b) They include literal as well as metaphorical expressions, as, for example, contradict and reject. Of particular interest is the authors ordering of the gathered material into a huge word field and its many subfields,
because these can be understood as identifying the metapragmatic aspects of communication that the English speaking community takes recourse to. The word
fields in their entirety unfold a global metapragmatic conceptualization whose
Metapragmatics
121
main constituents delineate models entering into an integral system of speech activities. Each model has as its linguistic correlate a set of verbs, which again have
substructures. The enaction model, for example, consists of verbs, of which wish,
aim at, or anticipate express VOLITION , alert s. o., indicate, or refer to PUTTING INTO
FOCUS , while expressions of offering, asking, commissioning, ordering, warning
etc. are instances of ENACTION , and verbs like accept, refuse, or obey instances of
REACTIONS ON ENACTIONS . At the most refined level, we find sets of verbs that show
a high degree of synonymy; for the enactment category SURPRISING , for instance,
the following verbs are listed: amaze, astonish, astound, bewilder, disconcert, flabbergast, shock, stagger, startle, strike, surprise. Taken together they delineate, if
we will, how the English language community conceptualizes acts of surprising
somebody by words; seen in a differentiating way, each single item contributes a
special nuance to the overall-concept, in comparison to the others.
Along these lines, Goossens (1987) has made an analysis of some such nearsynonymous verbs. His treatment of the speech act verbs say, tell, talk and speak
can well illustrate how much an in-depth analysis is able to reveal about peoples
metapragmatic models, though it explicitly tries just to answer a typically Fillmorean question, namely, how do linguistic action verbs frame the scene of linguistic (inter)action? (Goossens 1987: 95). Thus, native English speakers assign
different roles to receivers, encoding in tell and say the receptor who plays a more
passive part, and in talk the interactor who plays a more active, participative part;
in the case of speak, the English speaker accounts for the two roles by using distinct prepositions, with where the receptor acts as interactor, to where as receptor
(cf. Goossens 1987: 103). The four verbs reveal, furthermore, that native English
speakers differentiate messages according to the degree to which they have been
condensed. This is reflected in the facts that, for example, say freely combines
with direct enunciations, while tell is considerably restricted in taking them
(Goossens 1987: 103); the differentiation is in principle even reflected where the
distributional pattern is reversed, in other words where more extreme condensations go with tell, but not with say (e.g., tell the truth vs. *say the truth).
Lexical expressions that belong to certain metaphorical concepts offer particularly rich information about peoples cognitive metapragmatic models.21 Metaphors have in general attracted much attention by linguists because they seem to
offer insights into how the human mind operates. As Lakoff and Johnson state in
their pivotal study of 1980: [M]etaphor is pervasive in everyday life, not just in
language but in thought and action. Our ordinary conceptual system, in terms of
which we both think and act, is fundamentally metaphorical in nature (Lakoff and
Johnson 1980: 3); and metaphors that conceptualize communication in one respect
or another have been paid attention to right from the very beginning.
For an illustration, we may turn to two of the most widely known examples,
i.e., the so-called conduit metaphor, identified by Reddy (1979) and differentiated
later by Vanparys (1995) and Semino (2006), and the argument is war concept
122
Axel Hbler
proposed by Lakoff and Johnson (1980) and modified by Grady (1998) and Semino
(2006). The results of these investigations, derived from a close analysis of sets of
metaphorical expressions in general use and subsequently supplemented by corpus
data, can be understood as reconstructions of native English speakers conceptualizations of certain aspects of how they communicate and what their form of communication is like. These conceptualizations are metapragmatic in nature.
As far as communication between communicators in general is concerned, the
dominant meta-pragmatic model people have developed shows three main constituents: the speaker/sender puts an idea-object into a word-container and sends it
(through a conduit) to the receiver, who extracts from the word-container the ideaobject.
Discussion/argumentation represents a specific form of communication/discourse. The corresponding metaphors widely in circulation reveal that English
speakers share a (metapragmatic) model of argumentative discourse that centers
around physical conflict; debatants have two positions, attacking the others
opinion, while defending ones own; we have phases of maneuvering and retreat;
the outcome amounts to truce or victory/surrender.
(2) Another important source for gaining insights into peoples understanding
of communication would consist in information that researchers are able to elicit
directly from members of the speech community; this is a method which is often
applied by discourse analysts, where the aim is to produce ideotypical descriptions [of concrete discourses] that can be dissected into significant components
(Gumperz 1982: 157). But the results are on the whole quite unsatisfying; ordinary
conversationalists are obviously poor descriptors (Gumperz 1982: 157),22 and
this should hold true even more, were they confronted with a more abstract and
complicated task. Alternatively, an experimental way of eliciting metapragmatic
knowledge may be feasible (similar to the method that sociologists, e.g., Goffman,
occasionally deploy, for discovering certain norms which people subscribe to in
relations in public, i.e., by confronting them with transgressions), but has not, to
my knowledge, been pursued. The use of questionnaires seems to be the relatively
more reliable method.
The study by Simon-Vandenbergen (1995) shall be used to further illustrate the
point at stake, even if it is not exactly made for such a purpose. Its aim was to obtain assessments of linguistic behavior from English native speakers. The method
consisted in having natives evaluate selected aspects and forms of communicative
behavior through evaluating corresponding (metaphorical) expressions, which, as
we saw, embody parts and aspects of folk models of behavior. The questionnaire
applied included expressions covering social functions of talk (e.g., conversation,
chat, or prattle), turn-taking (e.g., run on, cut someone short, step in), topic management (e.g., bring up, move unto, meander), and manner of speaking (e.g., rattle,
babble, or air, thunder, or tell someone flat out, wrap up ones meaning). The questions/tasks that were to stimulate statements about peoples metapragmatic views
Metapragmatics
123
had three formats: 1. (social) assessments in terms of positive negative neutral,

2. (aesthetic) assessments in terms of light heavy neither, and 3. the same as
format 1 but supplemented by (a) a question asking why the subject rated the expression the way s/he did and (b) the request to give an example sentence.
While certainly all tasks can be interpreted as throwing light on peoples understanding of communication, it is the supplementary task (a) which should elicit the
most; unfortunately, Simon-Vandenbergen does not provide any cue as to what and
how subjects met this task. Nevertheless, some of the results she obtained allow
conclusions. Speech tempo is a good case in point. This is the picture emerging for
native English conversationalists (cf. Simon-Vandenbergen 1995: 108): Speed is
an important criterion; high and low speeds are often negatively evaluated as too
fast and too slowly respectively, in which case they tend to be talked about in
metaphorical terms.
One could, of course, object and claim that the negative evaluations elicited by
the questionnaire were just triggered by the metaphoricity of the expressions themselves. Any such objection, however, would not be justified. As Simon-Vandenbergen already seems to imply, it is the perceived marked value of the communicative phenomenon in relation to norms and standards in a given community that
triggers metaphorical expressions and assessments, and not the metaphorical expression that triggers a corresponding evaluation. Besides, in the case of speed,
there is even some independent support from social-psychological research for the
claim that the negative evaluations relate, indeed, to speed and are not simply triggered by the metaphoricity of the lexical expression. Smith et al. (1975), for
example, had subjects listen to recordings of speakers performing one and the same
text with different speeds; in reality, it was just one speaker whose natural speech
tempo had been technically manipulated. The subjects had to judge the speakers
as to benevolence, a social criterion which aligns well with Simon-Vandenbergens positive-negative scale. The authors found that [t]he benevolence/rate
plot reveals an inverted U-relationship, with the mean for normal voices [] being
rated more benevolent (Smith et al. 1975: 150), while high- and low-speed
speakers scored significantly low in benevolence.
The insights obtained by metapragmatic research outlined so far can be enriched (and maybe even relativized) by widening the scope so as to also include
possibilities of variation, always applying the same analytical parameters.23 Of the
three main types of variation, i.e., social, developmental, and historical, the social
variant has not yet received due attention. Illustrations will therefore be restricted
to the other two, turning to the lexicological source first.
As Geeraerts and Grondelaers (1995) point out,
if cognitive models are also cultural models, they are also cultural institutions, and as
such, they carry their history along with them: their institutional nature implies their historical continuity. It is only by investigating their historical origins and their gradual
transformation that their contemporary form can be properly understood. (1995: 177)
124
Axel Hbler
In line with this stance, Arnovick (1999) has examined changes in the conceptualization of promising:
The frequency of promises rendered defective at inception or follow-through has created a practical (if not moral) dilemma, a problem with promises. People do not always
trust the promises they hear. As a result, the need to convince the hearer is often assumed by the speaker. To perform a promise that satisfies the hearer, the speaker often
has to emphasize his or her commitment. (Arnovick 1999: 60)
She notes that even a performative use of promise, which is already a very strong
form of promising (cf. Traugott 1997), is nowadays often felt to be insufficient and
to require some further declarations, such as those observed with George Bush:
Read my lips, I guarantee it, Believe you me. A simple I will nowadays works
only in a promissory situation and an extra-linguistic institution for its force (Arnovick 1999: 58), in contrast to OE times when both the use of sculan (expressing a
subjects obligation) and wilan (expressing a subjects intention) were binding on
the speaker. Nowadays, promises usually undergo relativizations and rationalizations: promises merely express honorable intentions; changing circumstances
can render vows unrealistic; insincerity is politically motivated, deception is committed for the sake of the task pursued. Such views are further warranted by references to independent proposals (by Muck 1989 and Rawls 1955) for constructing
theories of promising that incorporate a limit on speaker responsibility (Arnovick 1999: 60).
The historical and culture-bound relativity of metaphorical concepts on the
pragmatics of language can well be illustrated by two investigations of mine. In
Hbler (1998), I showed that the predominant spatial metaphorization relating to
the notion of express, which represents the first constituent within the conduit
paradigm resulted from a change that had taken place around the turn from the 16th
to the 17th centuries. The argument centered around the change in the use of prepositions going with the notion of express; it was a change from the instrumental
with to the spatial in (words). The former conceptualization aligns well with
and foreshadows the tools-paradigm proposed by Reddy (1979) as a possible alternative to the predominance of the conduit metaphor. And in Hbler (ms), I examined data from the 16th and 17th centuries which show that argument was conceived
during that period in less fierce and more playful terms, more specifically in terms
of a tennis match; Gradys (1998) findings can be taken as a diluted late version of
this early concept.
A discoursal source is tapped on by Gotti (2006). He shows for the 16th/17th
century that neologizing (in contrast to the practice nowadays) was still a matter to
talk and write about; it was the time when a scientific language developed in England, by taking over elements from the common code. The following passage can
be read as an instance of the public discourse about it, particularly about using new
words of foreign origin in English.
Metapragmatics
125
But now as touching myne entent in writing this treatise in the English. Though this
cause might seme sufficient to satisfy many men that I am an englysh man, and therefore
may more easely and plainly write in my natyue tonge, rather then in any other: yet vnto
them that know the hardness of the mater, this answer shuld seme vnlykely: considering
that it is more harder to translate into such a tonge, wherein the arte hath not ben written
before, then to write in those tongues that are accustomed, and (as I might say) acquainted with the termes of the science (Recorde 1547, quoted in Gotti 2006: 218)
Recorde was a physician and mathematician. His topic in this passage (probably
from the preface) is metalinguistic and concerns the (un-)availability of English
words for certain concepts. But the discourse itself could be a case for metapragmatics (in the sense discussed in section 1); the passage is metacommunicative in
that the writer raises the problem of effability in English and thus justifies his use
of un-English words in the treatise to follow. In the current context, however, the
quotation can be seen as documenting metapragmatic knowledge about the limits
and chances of informing an uninformed readership about specific and new states
of affairs. Representing the metapragmatic knowledge of a member of an educated
16th-century English speech community, the document could also give grounds for
comparing it with (in-)effability treatments by speakers from different social-cultural backgrounds; but I do not know of any such investigation.
The developmental type of variation in metapragmatic knowledge is, in
contrast, well investigated. For the purpose of illustration, the article by Bernicot
and Laval (1996) has been chosen, because it focuses, again, on promises. Their
subjects are children of three age groups, around 3, around 6 and around 10 years
of age. The childrens metapragmatic knowledge is defined as their capacity to talk
and think about acts of keeping and breaking promises and tested by examining
their verbal comments stimulated by a series of cartoon stories; each story consists
of a promise (by a child or a parent) and the subsequent fulfilment or non-fulfilment; the effect of sadness (as sign of dissatisfaction) and happiness (as sign of satisfaction) respectively, which the (non-)fulfilment has on the person to whom the
promise was made, has then to be determined by the subjects themselves and subsequently justified. The authors summarize their results thus:
Metapragmatic knowledge was found to evolve with age. At the ages of 3 and 6,
childrens metapragmatic knowledge mainly concerns the execution of the action. At
age 6, the listeners desires start being added in cases where [this] preparatory condition
is not satisfied. At age 10, explanations pertaining to action accomplishment completely
disappear, and explanations about the speakers intentions alone or about both the
speakers intentions and the listeners desires appear. (Bernicot and Laval 1996: 120)
The metapragmatic knowledge of adults would at least also comprise the sincerity
condition of promises, i.e., that the speaker intends to accomplish the future action
(cf. Gibbs and Delaney 1987).
126
3.2.
Axel Hbler
Metapragmatics as metatheory of pragmatics
So far, we have dealt with cases where speakers provide the data/material, which
linguists then process, refine (as was the case with metacommunication) or elaborate on (as in the case of speakers theoremes). We now leave the sphere where linguists occupy themselves with other peoples products and turn to their own activity, subjecting it to reflection.24 That portion of such reflections which becomes
manifest constitutes a professional theoretical dialogue (mainly in the written
medium) about pragmatics, with the aim of defining the field, supporting, supplementing, or justifying certain theoretical stances or challenging them. Such theoretical discussions show, in principle, two orientations. They either operate introspectively or involve external considerations, and sometimes these perspectives
merge. Since pragmatics as such is already the most complex language-related discipline, it is easy to understand that the meta-discourse about it is immense and
open-ended; problems abound. The few topics that will subsequently be addressed
can, therefore, do nothing else but give some indications.
The problems already start on the most elementary of levels, including even the
question what pragmatics is all about. What, in other words, is the object of pragmatics? Is it human communication or just the verbal part of it? The answer may
depend on the answer given to an even more elementary question: How to define
the scope of activity for a linguist working in the field of pragmatics? Is it the discipline that defines the scope or is it the object as such? In that particular case,
pragmatics originated in philosophy but was subsequently fully absorbed by linguistics; should it therefore remain a linguistic discipline and restrict its activities
accordingly? According to what? Was it not one of the major figures of linguistics,
de Saussure, who embedded linguistics in semiotics, which consequently would
license any occupation with the non-verbal as well? If, on the other hand, it is the
object which defines for scholars of pragmatics the scope of their activities, then
the non-verbal with its phonetic and kinesic domains would be included anyway
because natural communication, even any single utterance, can only be fully
understood as a trimodal phenomenon. It is not even justifiable to claim that the
verbal part plays the main role because intonation and/or gestures oftentimes modify in an important way what is verbally said; they sometimes even override the
verbally mediated meaning. Even if one were to subscribe to the view that, ideally
speaking, the nonverbal modes should be attended to as well, it could still be argued that such a holistic approach is simply not practicable; several disciplines
should therefore participate and join in a common endeavor. But how much compartmentalizing does the subject matter tolerate? Interdisciplinary research may
be a good answer, but transdisciplinarity may still be unavoidable up to a certain
degree.
Linguistics and communication research did not fall out of the sky; its predecessor was rhetoric, the art of persuasion. Traces of this tradition of thinking can
Metapragmatics
127
still be found in Searle (1969) in that he keeps the link alive that exists between a
speech act and its effect, the perlocution. Though the ensuing research paradigm
widened the scope so as to treat the speech act as part of speech act sequences, it
may cause wonder why perlocution never played a role. Why has research in indirect communication strategies been restricted to what the speaker plots with the
addressee underdetermined; why has research not systematically investigated the
(perlocutionary) effects that certain strategies are likely to cause in receivers? Because social psychology is doing this job? At some US universities, departments of
Communication (not tantamount to mass communication) have been established
whose activities are at the cross-roads of various lines of research, including linguistics, rhetoric and persuasion, and social psychology. Is that a more adequate
discipline for coming to terms with interpersonal communication? Or what do they
lack that the (European) language-biased approach has to offer?
Questions of involving other disciplines do not only represent boundary issues (Caffi); they may also be raised and pursued with pleasure when they help
dignify central views or important findings of ones own discipline. The dignifying effect results from their proving that the views or results are not isolated, but
align well with what is considered to be more fundamental in the sense that it provides the foundations of ones own discipline, its overarching rationale. A good
case in point is the ever growing theoretical horizon associated with pragmatics.
The cognitive turn in the humanities necessitated a cognitive foundation also for
pragmatics. Since the cognitive sciences, however, have (neuro-)biological foundations, to what degree are they to be paid tribute? Is it enough to allude every now
and then to such deep connections? Is it acceptable to do so only when it seems opportune, i.e., when it serves the argument at issue, and neglect them otherwise?
McNeills treatment of (everyday conversational) narratives provides an interesting example. A central part of his production theory, which (to varying degrees)
takes all three modes of realization into consideration (i.e., language, intonation,
and gestures), operates with a model of self-organization that obviously has its
roots in the theory of autopoietic (or self-referential, self-organizing) systems,
which goes back to Maturana (1970a, 1970b). McNeill, however, does not discuss
(metapragmatically!) these roots. Was it because the theory is difficult for students
of interpersonal communication to digest?25 Where are the limits, and how far do
we have to go, in how much detail?
Many of the questions raised (as well as others that could be raised) still await
answers. Whatever the answers are or will be, the fundamental nature of these issues (and even the questions themselves) are rarely free of any ideological bias,
whether one wants to admit it or not. A sensibility for the ideological dimension of
research in pragmatics cannot be taken for granted; it seems, in fact, that it still
needs to be called in. [A] constant monitoring of linguistic rhetoric in view of the
ideological underpinnings of theories and analyses is not a luxury but a prerequisite for the advancement of linguistic pragmatics (Verschueren 1998: 67). How
128
Axel Hbler
subtle the dangers of ideological pitfalls can sometimes be is well highlighted by

Caffi.
[O]ne has to avoid the risk of new forms of hypostasis [], brought about by certain,
apparently harmless, metalinguistic uses. For instance, the use of a definite noun phrase
can be idealizing: the risk of an idealization exists whenever one refers to the speaker,
who could end up by being an hyperuranian subject, or a lunar effet-sujet. (Caffi 2006:
83)
But more tangible issues have, of course, to be subjected to such self-examination

as well. What is the researchers stance to the speech community that s/he wants to
investigate about, even towards the individual informant? Does the stance have
any political implications? And on a methodological level, are the questions asked
(orally or by means of a questionnaire) politically correct?
A subspecies of this kind of metatheoretical problem consists in the question of
whether the methods of tackling pragmatically relevant data is neutral enough to
cope with different yet comparable sets of data. The problem of culturally biased
approaches is addressed here as well. Among the English speech act verbs, for
example, we find none that would have lexicalized any specification as to time and
location (as in Tenejapa Tzeltal, where concepts like talk that has taken place just
a short time ago or talk occurring in a grassy area are lexicalized), but we do find
English speech act verbs of directing, such as appoint, convene or ordain, which
include specifications as to social setting and institution (cf. Verschueren 1987:
131133). A simple speech act classification would not do because it would neglect the specificities. In the face of such shortcomings, Verschueren proposes to
avoid attempts at classifying linguistic action verbs (and the corresponding acts)
and [] concentrate instead on the careful contrastive analysis of the semantic dimensions involved in the lexicalization of linguistic action (Verschueren 1987:
126). The descriptive framework he offers is supposed to allow for accommodating
(by way of comparing) intra-linguistic, cross-linguistic and cross-cultural differences.
A (partial) translation of Verschuerens program into action is Meys analysis
of a (by now familiar) speech act, promise.
[T]he context in which a promise is made is of the utmost importance for its status as a
promise and for its binding effects. Take the case of a young person promising his or her
parents not to smoke before the age of eighteen. In this case, the societal conditions surrounding the execution of such a promise can be exceedingly difficult (peer-group
pressure, work conditions, etc.). (Mey 1993: 170)
But it serves the author to underline a metapragmatic argument: the view that a
philosophical speech act theory la Austin (21990) and Searle (1969) may not suffice for linguistic purposes.
Both Austin and Searle operate on the one sentence one case principle. That is to say,
in order to illustrate their theory, they use examples that are characteristic of what they
Metapragmatics
129
see as the case under discussion, such as, e.g., a particular kind of speech act. I will, in
the following, show the shortcomings of the case approach (Mey 1993: 170).
The example also demonstrates that metapragmatic discourse does not always
have to be the raison detre for an entire (theoretical) text. It often has, rather, a
more limited purpose, i.e., to profile ones own theoretical assumptions against the
backdrop of what is widely held to be the case. This pattern is pervasive and can be
found in almost all scholarly work in whatever discipline; but it is of course only in
pragmatics that such metatheoretical discourse is metapragmatic. The illustration
to follow is taken from a prominent source.
A more radical difference between Grices approach and relevance theory [as proposed
by Sperber and Wilson, A.H.] is this. Grices principle and maxims are norms which
communicators and audience must know in order to communicate adequately. Communicators generally keep to the norms, but may also violate them to achieve particular
effects; and the audience uses its knowledge of the norms in interpreting communicative
behaviour. The principle of relevance, by contrast, is a generalisation about ostensiveinferential communication. Communicators need no more know the principle of relevance to communicate than they need to know the principles of genetics to reproduce
[] (Sperber and Wilson 1986: 162).
The quotation is an excerpt from a short metapragmatic passage of one and a half
pages, where Sperber and Wilson confront their concept of relevance with that of
Grice (e.g. 1975). The metapragmatic discourse in this case is even sparked off by
a metalinguistic problem, i.e., the fact that both parties use one and the same term,
but associate it with different concepts. The passage, however, is not metacommunicative because concepts, methods and the like are the issue, not the communicative form or aspects of it. This is a feature that generally applies to any form in
which pragmatics is theoretically reflected. It is a type of reflection that is at best
accidentally language-reflexive. This sets it apart from all the other phenomena
that have been taken account of in this article and aligns it with the genre metatheory. Metapragmatics may be, in the end, an all too generous term.
Notes
1. On July 31, 2009.
2. This is basically what Caffi (2006) takes account of as metapragmatics 2.
3. Applying the typology of metamessages proposed by Jakobson (1971 [1957] 130133),
we are mainly concerned with type (M/M), i.e., messages about messages. Certain attention will also be given to type (C/M), that is to those metamessages in which so-called
shifters (as the code involved) incorporate a message about a message. Messages about
code (M/C) as well as the meta-message type (C/C), where proper names are the protagonists, fall outside the scope of metacommunication proper. Cf. also Lucy (1993: 16).
4. In section 2, a different concept will be introduced, which conceptualizes metalanguage
as a dimension.
130
Axel Hbler
5. Metalanguage in the sense outlined, however, is part of but not equivalent to reflexive
language as described by Lucy (1993) just as speech reflexive activity/ reflexive language use is not tantamount to metacommunication.
6. Such reduced forms will be tackled separately below.
7. Quite extensive studies along these lines have been conducted earlier by Meyer-Hermann (1980 [unpublished]). Lumping together topic and function, he identified 72
metacommunicative speech act types (on the basis of French and Portuguese material).
The 11 most frequently used types are: requesting the performance of speech acts, acknowledging the performance of speech acts, announcing the performance of speech
acts, describing communicative interaction, guiding discussions, criticizing communicative interaction, restating a communicative function more precisely, eliciting receptive readiness, securing comprehension, evaluating communicative interaction, and anticipating sanctions prophylactically (cf. Welte 1990: 178180). They all have been
included in one way or another in the present outline.
8. Meyer-Hermann (1980), in fact, tackles functions only, the topics being implied.
9. Quoting, in fact, is the phenomenon that Jakobsons meta-message type (M/M) was obviously meant to mainly accommodate.
10. As Wilson (2000: 148) points out, [the] attitudes conveyed by echoic utterances can be
very rich and varied: the speaker may indicate that she agrees or disagrees with the original, is puzzled, angry, amused, intrigued, skeptical etc., or any combination of these.
11. Often the disjunct may have a paraphrase with a verb of saying taking the form of a
finite clause (If I may say so without offence, your writing is immature).
12. Hagemann (1997) conducted a detailed investigation on German diction-characterizing adverbs in Gricean terms.
13. I am not sure about the status of adverbs such as well which can either function as discourse initiators (cf. Quirk et al. 1995) or continuation markers (cf. Halliday and Hasan
1976).
14. The topical distinction as such has been highlighted, under different labels, by various
authors; cf. Bamford and Biondi (2005: XVII), Hagemann (1997: 2728), Geissner
(1981: 206). Watzlawick et al. (1968) and Bateson (1972) focus exclusively on the interpersonal dimension and reserve for this aspect the term metacommunication.
15. In Jakobsons typology (cf. footnote 3), they would represent metamessages of the type
(C/M).
16. Jakobson (1971: 131), who adopted the concept from Jespersen (1922: 123), discusses
shifters at some length.
17. Halliday and Hasans use of the term presupposition differs, of course, from its usual
semantic-pragmatic reading; they use it in the wider sense that something is to be supplied, or understood (Halliday and Hasan 1976: 144).
18. Cf. also analogous term formations such as synsemantic.
19. But caution is advised. Speakers awareness [] does not have to match the linguists
metapragmatic descriptions. Naively confusing these two may contribute to the furthering of a specific kind of folk metapragmatics inherent in western linguistics itself. Assuming that such nave confusion can be avoided, at least some of my further comments
will be based on results obtained in this line of research (Verschueren (1998: 6061).
20. In certain respects complementary, but of a narrower scope, is Schmid (2000) with his
investigation of shell nouns, a subset of which overlaps with notions taken account of
by Ballmer and Brennenstuhl (1981).
Metapragmatics
131
21. For obvious reasons, I simply take both kinds for granted, though it is certainly a matter
worth discussing; cf. Kvecses (2005: 193 ff.) and his discussion of the question
whether cultural models can exist without prior metaphorical understanding, as claimed
by Quinn (1991).
22. When participants report on actual verbal encounters, they tend to do so by mentioning
some item of content, or by referring to what people were getting at (Gumperz 1982:
157).
23. For questioning the parameters, cf. section 3.
24. This overlaps with what Caffi characterizes (under metapragmatics 1) as concern for
the epistemological implications of what researchers assume pragmatics is about (Caffi
2006: 83).
25. The reader may test it, in the original or in a version such as Benseler et al. (eds.) 1980,
which has communication scientists as its target audience.
References
del, Annelie
2006
Metadiscourse in L1 and L2 English. Amsterdam: Benjamins.
Anton, Corey
1998
About talk: The category of talk-reflexive words. Semiotica 121/3,4:
193212.
Arnovick, Leslie K.
1999
Diachronic pragmatics. Seven case studies in English illocutionary development. Amsterdam: Benjamins.
Auer, Peter
1986
Kontextualisierung. Studium Linguistik 19: 2247.
Austin, John L.
1990
How to do things with words. Edited by James O. Urmson and Marina Sbis.
2nd. ed. Oxford: Oxford University Press.
Bakhtin, Michail M.
1981
The dialogic imagination. Edited by Michael Holquist translated by Emerson,
Caryl and Michael Holquist. Austin, TX: University of Texas Press.
Ballmer, Thomas and Waltraud Brennenstuhl
1981
Speech act classification. A study in the lexical analysis of English speech activity verbs. Berlin: Springer.
Bally, Charles
1913 [1965] Le langage et la vie. Genve: Ed. Atar, [reprinted Genve: Librairie Droz].
Bamford, Julia and Marina Biondi
2005
Introduction. In: Julia Bamford and Marina Biondi, Dialogue within discourse
communities. Metadiscursive perspectives on academic genres. Tbingen:
Niemeyer. IXXIV.
Bateson, Gregory
1972 [1955] Steps to an ecology of mind. New York: Balantine.
Benseler, Frank, Peter M. Hejl and Wolfram K. Kck (eds.)
1980
Autopoiesis, communication and society. The theory of autopoietic system in
the social sciences. Frankfurt, New York: Campus.
132
Axel Hbler
Bernicot, Josie and Virginie Laval

1996
Promises in French children: Comprehension and metapragmatic knowledge.
Journal of Pragmatics 24: 101122.
Bublitz, Wolfram and Axel Hbler (eds.)
2007
Metapragmatics in use. Amsterdam: Benjamins.
Caffi, Claudia
2006
Metapragmatics. Encyclopedia of Language and Linguistics. Amsterdam:
Elsevier. 8288.
Cameron, Deborah
1995
Verbal hygiene. London: Routledge.
Cameron, Deborah
2000
Good to talk? Living and working in a communication culture. London: Sage.
Coupland, Nikolas
2004
Stylised deception. In: Adam Jaworski, Nikolas Coupland and Dariusz Galasinski (eds.), Metalanguage. Social and ideological perspectives, 249274.
Berlin: Mouton de Gruyter.
Coupland, Nikolas and Adam Jaworski
2004
Sociolinguistic perspectives on metalanguage: Reflexivity, evaluation and
ideology. In: Adam Jaworski, Nikolas Coupland and Dariusz Galasinski (eds.),
Metalanguage. Social and ideological perspectives, 1551. Berlin: Mouton de
Gruyter.
Fairclough, Norman (ed.)
1992
Critical language awareness. London: Longman.
Gee, James P.
1999
Discourse analysis. London, New York: Routledge.
Geeraerts, Dirk and Stefan Grondelaers
1995
Looking back at anger: Cultural traditions and metaphorical patterns. In: John
R. Taylor and Robert E. MacLaury (eds.), Language and the cognitive construal of the world, 153179. Berlin: Mouton de Gruyter.
Geissner, Hellmut
1981
Sprechwissenschaft. Theorie der mndlichen Kommunikation. Knigstein im
Taunus: Scriptor.
Gibbs, Raymond W. and Suzanne M. Delaney
1987
Pragmatic factors in making and understanding promises. Discourse Processes
10: 107126.
Goffman, Erving
1972
Relations in public: Microstudies of the public order. New York: Harper and
Row.
Goossens, L.
1987
Dealing with linguistic action verbs in depth. In: Jef Verschueren (ed.), Linguistic action: Some empirical-conceptual studies. Norwood, NJ: Ablex.
95106.
Gotti, Maurizio
2006
Metalinguistic considerations in 17th-century specialized discourse. In: Anna
Duszak and Urszula Okulska (eds.), Bridges and barriers in metalinguistic
discourse, 217231. Frankfurt: Peter Lang.
Metapragmatics
133
Grady, Joseph
1998
The Conduit metaphor revisited: A reassessment of metaphors for communication. In: Jean-Pierre Koenig (ed.), Discourse and cognition: Bridging the
gap. Stanford, CA: CSLI Publications. 205218.
Greatbatch, David and Robert Dingwall
1998
Talk and identity in divorce mediation. In: Charles Antaki and Sue Widdicombe (eds.), Identities in talk, 121132. London: Sage.
Graf, Claudia
2007
Echo und Zitat. In: Elke Brendel, Jrg Maibauer and Markus Steinbach (eds.),
Zitat und Bedeutung. Linguistische Berichte, Sonderheft 15: 89109.
Grice, H. Paul
1975
semantics. Vol. 3: Speech acts. New York: Academic Press.
Gumperz, John J.
1982
Discourse strategies. Cambridge: Cambridge University Press.
Habermas, Jrgen
1971
Vorbereitende Bemerkungen zu einer Theorie der kommunikativen Kompetenz. In: Jrgen Habermas and Niklas Luhmann, Theorie der Gesellschaft oder
Sozialtechnologie, 101141. Frankfurt: Suhrkamp.
Hagemann, Jrg
1997
Reflexiver Sprachgebrauch: Diktumscharakterisierung aus Gricescher Sicht.
Opladen: Westdeutscher Verlag.
Halliday, Michael A.K. and Ruquaia Hasan
1976
Hbler, Axel
1998
The expressivity of grammar. Grammatical devices expressing emotion across
time. Berlin: Mouton de Gruyter.
Hbler, Axel
o.J.
The conceptualization of argument in terms other than war. A socio-cultural
case study. (Ms)
Hbler, Axel and Wolfram Bublitz
2007
Introducing metapragmatics in use. In: Wolfram Bublitz and Axel Hbler
(eds.), Metapragmatics in use, 126. Amsterdam: Benjamins.
Hyland, Ken
2005
Metadiscourse. Exploring interaction in writing. London, New York: Continuum.
Jakobson, Roman
1971 [1957] Shifters, verbal categories, and the Russian verb. In: Selected writings II:
Word and language, 130147. The Hague: Mouton.
Jespersen, Otto
1922
Language. London: Allen and Unwin.
Kvesces, Zoltan
2005
Metaphor in culture. Universality and variation. Cambridge: Cambridge University Press.
1980
Metaphors we live by. Chicago: The University of Chicago Press.
134
Axel Hbler
Leder, Drew
1990
The absent body. Chicago: The University of Chicago Press.
Lucy, John A.
1993
Reflexive language and the human disciplines. In: John A. Lucy (ed.), Reflexive language. Reported speech and metapragmatics. Cambridge: Cambridge
University Press. 932.
Maturana, Humberto R.
1970a
Biology of cognition. Report 9. Biological Computer Laboratory, Department
of Electrical Engineering. University of Illinois, Ill.
1970b
Neurophysiology of cognition. In: Paul L. Garvin (ed.), Cognition: A multiple
view, 323. New York: Spartan.
McNeill, David
1992
Hand and mind. What gestures reveal about thought. Chicago, London: The
University of Chicago Press.
Mey, Jacob L.
1993
Pragmatics. An introduction. Oxford, Cambridge: Blackwell.
Meyer-Hermann, Reinhard
1978
Aspekte der Analyse metakommunikativer Interaktionen. In: Reinhard MeyerHermann (ed.), Sprechen Handeln Interaktion, 103142. Tbingen: Niemeyer.
1980
Studien zur Funktion von Metakommunikation am Beispiel portugiesischer
und franzsischer Alltagssprache. Habilitationsschrift Universitt Bielefeld.
(Ms.)
Muck, Bruce
1989
Promising rules and intentions. PhD Diss. University of Wisconsin.
Niedzielski, Nancy A. and Dennis R. Preston
2000
Folk linguistics. Berlin: Mouton de Gruyter.
Paul, Ingwer
1999
Praktische Sprachreflexion. Tbingen: Niemeyer.
Preston, Dennis R.
1996
Whaddayaknow? The modes of folk linguistic awareness. Language Awareness 5: 4077.
Quinn, Naomi
1991
The cultural basis of metaphor. In: James Fernandez (ed.), Beyond metaphor:
The theory of tropes in anthropology. Stanford, CA: Stanford University Press.
5693.
Quirk, Randoph, Sidney Greenbaum, Geoffrey Leech and Jan Svartvik
1985
A comprehensive grammar of the English language. London, New York: Longman.
Rawls, John
1955
Two concepts of rules. Philosophical Review 64/1: 332.
Reddy, Michael
1979
The conduit metaphor: a case of frame conflict in our language about language.
In: Andrew Ortony (ed.), Metaphor and thought, 164201. Cambridge: Cambridge University Press.
Metapragmatics
135
Schmid, Hans-Jrg
2000
English abstract nouns as conceptual shells. Berlin: de Gruyter.
Searle, John R.
1969
Speech acts. Cambridge: Cambridge University Press.
Semino, Elena
2006
A corpus-based study of metaphors for speech activity in British English. In:
Anatol Stefanowitsch and Stefan T. Gries (eds.), Corpus-based approaches to
metaphor and metonymy, 3662. Berlin: Mouton de Gruyter.
Silverstein, Michael
1993
Metapragmatic discourse and metapragmatic function. In: John A. Lucy (ed.),
Reflexive language, 3358. Cambridge: Cambridge University Press.
Simon-Vandenbergen, Anne-Marie
1995
Assessing linguistic behaviour. A study of value judgements. In: Louis
Goossens, Paul Pauwels, Brygida Rudzka-Ostyn, Anne-Marie Simon-Vandenbergen and Johan Vanparys (eds.), By word of mouth, 71124. Amsterdam:
Benjamins.
Smith, Bruce L., Bruce L. Brown, William J. Strong and Alvin C. Rencher
1975
Effects of speech rate on personality perception. Language and Speech 18:
145152.
1986
Relevance. Communication and cognition. Oxford: Blackwell.
Tajfel, Henri
1982
Social identity and intergroup relations. Cambridge: Cambridge University
Press.
Traugott, Elizabeth
1997
Subjectivication and the development of epistemic meaning: The case of
promise and threaten. In: Toril Swan and Olaf J. Westvik (eds.), Modality in
Germanic languages: Historical and comparative perspectives, 185210. Berlin: de Gruyter.
Ungeheuer, Gerold
1970
Kommunikative und extrakommunikative Betrachtungsweisen in der Phonetik. In: Proceedings of the Sixth International Congress of Phonetic Sciences,
7386. (Prag 1967). Prague.
Vanparys, Johan
1995
A survey of metalinguistic metaphors. In: Louis Goossens, Paul Pauwels,
Brygida Rudzka-Ostyn, Anne-Marie Simon-Vandenbergen and Johan Vanparys (eds.), By word of mouth, 134. Amsterdam: Benjamins.
Verschueren, Jef
1987
Metapragmatics and universals of linguistic action. In: Jef Verschueren (ed.),
Linguistic action. Some empirical-conceptual studie,125147. Norwood, NJ:
Ablex.
Verschueren, Jef
1998
Notes on the role of metapragmatic awareness in language use. In: Adam
Jaworski, Nikolas Coupland and Dariusz Galasinski (eds.), Metalanguage.
Social and ideological perspectives, 5373. Berlin: Mouton de Gruyter.
Watzlawick, Paul, Janet H. Beavin and Don D. Jackson
1968
Pragmatics of human communication: A study of interactional patterns, pathologies, and paradoxes. London: Faber and Faber.
136
Axel Hbler
Welte, Werner
1990
Alltagssprachliche Metakommunikation im Englischen und Deutschen. Frankfurt: Lang.
Wilson, Deirdre
2000
Metarepresentation in linguistic communication. In: UCLWPL 11: 127161.
Wooffitt, Robin and Clark Colin
1998
Mobilizing discourse and social identities in knowledge talk. In: Charles
Antaki and Sue Widdicombe (eds.), Identities in talk, 107120. London: Sage.
Part II
Theoretical foundations
5.
The rise of pragmatics: a historiographic

overview
Wataru Koyama
1.
Introduction
This article investigates the theoretical, philosophical, social institutional, and interpersonal genealogies of several key concepts of contemporary linguistic pragmatics, such as presupposition, implicature, speech act, and performativity, as well
as perhaps more fundamental ones, viz., context, indexicality, function, referential
and non-referential practice, metapragmatics, and reflexivity. More specifically,
after briefly dealing with the genealogy of the concepts context and indexicality in Section 2, the article discusses the genealogies of presupposition and
effect (presupposing and creative functions) in Section 3, function (multifunctionality) in Sections 4, 5, and 6, and performativity, metapragmatics, and
reflexivity in Section 7.
In so doing, this article tries to articulate a historiographic overview of how linguistic pragmatics was established and developed before it became an institutionalized international discipline. It specifically focuses on two critical developments since the late nineteenth century that led to the rise of pragmatics in the
1970s: i.e., the linguistic turn, to be followed by the pragmatic turn. The
former, which took place from the 1880s to the 1930s, gave rise to structural-functionalist linguistics. This historical transformation was closely linked to the neoKantian movement, marking the empirical turn of Kantian philosophy, and the
phenomenological movement, focusing on the synchronic here and now of communicative events (see Sections 5 and 6). On the other hand, the linguistic turn also
led to the emergence of analytic philosophy, with its focus on the propositional,
logical-semantic aspects of language. Structural linguistics, interacting with analytic philosophy, took the formalist turn in the 1940s, turning away from context
and (communicative) function and moving to concentrate on the (morpho)phonological, (morpho)syntactic, and semantic aspects of language. Gradually, however,
especially in the 1950s and 60s, there was increasing awareness of the importance
of context, often discussed under the rubrics of indexicality and pragmatics, even
in logico-analytic philosophy (see Section 2). This led to the pragmatic turn and the
institutional establishment of linguistic pragmatics in the 70s, linguistics and philosophy moving from the analytically-oriented study of de-contextualizable regularities of language toward the empirical study of the contextualized use of language. Thus, since the 1970s, the theoretical notions articulated or suggested by
the older traditions of structural-functional linguistics, such as context, indexical-
140
Wataru Koyama
ity, function, and metalanguage, have been re-discovered and further explored. In
addition, the pragmatic turn has allowed the convergence of pragmatics with the
empirical social sciences, which are traditionally concerned with actions and interpretations in context. As is described in Section 3, this convergence characterizes
the current state of linguistic pragmatics, where the earlier emphasis on regularities
in language use is increasingly supplemented with the focus on the pragmatically
achieved (trans)formation of group identities and power-relations of language
users in specific sociocultural contexts.
In such a way, this article tries to present a historiographic account showing
that the true home of linguistic pragmatics lies in the Continental tradition (cf.
Huang 2007: 45), which tries to provide a contextual, pragmatic view of language
and communication in our life-worlds, not in the Anglo-American tradition,
which sees pragmatics as a narrowly delimited component supplementary to the
ahistorical, universalizing studies of syntax and semantics, abstractable from referential regularities. As this article will demonstrate, pragmatics came out of the
Continental tradition in the late nineteenth century and, after the formalist lapse at
the mid-twentieth century, the discipline has been moving back from the AngloAmerican to the Continental tradition, further exploring referential and non-referential practices, i.e., what we do, with or without words, in our life-worlds (see
Section 5 for details).
2.
Genealogy of the concepts context and indexicality
Let us begin our historiographic reconstruction with one of the most basic notions
of pragmatics, namely, context, which is closely related to the phenomenological
notion of life-world and which includes not only the tokens of linguistic types
but also communicative-event participants and their pragmatic acts (cf. Mey
[1993] 2001: 206235). It would not be too far-fetched to say that pragmatic
means contextual(ized), as the discipline itself is differentiated by its focus on
context, from structural linguistics and semantic logic, which are mainly concerned with de-contextualized regularities abstractable from referential practices.
Clearly, the historiography of pragmatics must be written around this theme: how
context has been discovered as a theoretical topic, and theoretically dealt with (cf.
Nerlich and Clarke 1996).
Of course, all empirical studies of language or communication have been
at least implicitly concerned with context, so that we might argue that modern
theorization on context(uality) started with the emergence of modern empiricism,
as suggested by John Lockes (16321704) treatise on human understanding
and communication (cf. Locke [1689] 1975), which anticipated Thomas Reids
(171096) pragmatic commonsense philosophy and thus paved the way for the
development of modern semiotics, pragmati(ci)sm, and pragmatics.1
141
Yet, from the linguistic (structural) point of view, which focuses on de-contextualizable linguistic categories, it is, of course, personal and demonstrative pronouns, moods, modalities, tenses, and other deictic categories (deixis) or indexical
types that most clearly show contextuality, often called context-dependency (cf.
Fillmore 1975; Levinson 1983: 5496); hence, the specifically linguistic origins
of todays linguistic pragmatics might be found in empirical investigations and
theorizations of the deictic categories. Thus, the roots of linguistic pragmatics
may be sought, as Nerlich and Clarke (1996) have done, in the structural-functionalist traditions: i.e., the (Austro-)German tradition, which led to Karl Bhlers
(18791963) theory of indexical expressions (Zeigwrter), the deictic field (Zeigfeld; vs. Symbolfeld), and Organon model, developed in 193334; the French
one, culminating in Benvenistes (190276) theory of pronouns, subjectivity, and
smantique, or what would become known as nunciation (Benveniste, 1966); and
the British tradition, which eventually led to Lyonss (1977) semantics (in the
broad sense, including linguistic pragmatics) and Hallidays text linguistics, the
latter synthesizing the Firthian theory (London School) and the Praguean theory
of referential (informational) communication and multi-functionalism (see Section 6).2
We shall follow these functionalist movements in Europe, anticipating the
worldwide rise of pragmatics in the latter half of the twentieth century, later in this
article (see Sections 46). Here, we should take a brief look at the development on
the other side of the Atlantic. That is, in the United States, the Praguean functionalist theory became combined with the Peircean theory of semiotic pragmati(ci)sm
by Jakobson (18961982), who articulated a contextually-anchored, that is, speech
event-centered account of grammatical categories, including what Bally called
actualisateurs or Jespersen (18601943) named shifters (Jespersen 1922),
which Jakobson (1957) characterized as denotational-indexicals. As is well
known, the notion of indexicality came from Peirce (see note 1), who defined it as
the kind of relationship between signs (e.g., tokens of linguistic types) and their
objects (e.g., contextual referents) that is pragmatically motivated by the empirical
principle of contiguity in the context of signification (communication). Note that
indexicality is a general semiotic notion, characterizing not only deixis, but
also pointing gestures, gazes, spatial arrangements (e.g., proximity), and any other
modes of communicative practices based on the principle of contiguity in the contexts of communicative events (cf. Peirce [18931910] 1955). Clearly, indexicality
is a key to the proper theorization of context and contextualized communication, as
duly recognized by the micro-sociologist (ethnomethodologist) Garfinkel (see
Section 3) and, later, the social pragmatician Mey ([1993] 2001: 5260).
The significance of indexicality for pragmatics had been also acknowledged, in
a different disciplinary context, in Bar-Hillel (1954), which would lead to Richard
Montagues (193071) attempt at a formal pragmatics. As this indicates, Bar-Hillel
was an analytic philosopher generally working in the logical syntax framework
142
Wataru Koyama
of Rudolf Carnap (18911970), closely associated with the behavioristic pragmatist Charles Morris (190179), i.e., two figures involved in the logical-empiricist
project of Unified Science (Einheitswissenschaft), or the Unity of Science Movement, led by Otto Neurath (18821945; a member of the Vienna Circle), in the
early-to-mid-twentieth century (cf. Carnap [1934] 1937; Morris 1938). In particular, Morris, partially following Peirces semiotic pragmati(ci)sm (see immediately
above), had articulated, in 1937, the famous semiotic triad of syntax, semantics,
and pragmatics,3 to be adopted by Carnap. Coming out of such a tradition, Bar-Hillels work on indexicality, pragmatics (as wastebasket of syntax and semantics),
and implying (what Grice would later call implicature) in the 1940s and 1950s
(cf. Bar-Hillel 1946; Chapman 2005: 92) anticipated, in retrospect, not only Montague and Grice, but also the pragmatic turn in analytic philosophy in the 1960s
and 70s, which saw Kripkes (1972) theory of rigid designation, causal theory
of reference on proper names (cf. Putnam 1975), speakers reference (true reference; Donnellan 1966, 1978; Kripke 1979), and Kaplans (1978) theory of
Dthat on demonstratives. With these developments, by which analytic philosophy became increasingly focused on indexicality and pragmatics, the discipline
turned post-analytic and came to merge with neo-pragmatism in the 1970s. Also,
whereas the Carnapian logical syntax tradition was associated with neo-Bloomfieldian structural linguistics and its derivatives like Chomskyan generativism, the
newer, post-analytic developments became associated with Generative Semantics
and the emerging discipline of linguistic pragmatics in the 1970s (cf. Donnellan
1978; Davis 1991).
Thus, indexicality, along with its associated concepts, has been a key element
for the constitution of pragmatics and the proper theorization of context. Further,
the Peircean notion of indexicality and Bhlers ideas of deictic field (Zeigfeld) and deictic center (origo) became theoretically combined in the linguisticanthropological tradition, derived from not only Boas (18581942) and Sapir
(18841939), but also Peirce, Jakobson, Benveniste, and Kuryowicz (1895
1978), and led by a disciple of Jakobsons: namely, Silverstein (b. 1945). As documented by Pressmans (1994) contribution to the historiography of pragmatics
(also, Caton 1987), Silverstein (1976a) demonstrated, in his famous Jakobsonian
work on case-marking and the Noun Phrase (NP) hierarchy, that grammatical categories are universally, systematically anchored onto the deictic center, located at
the center of the context of communicative interaction (i.e., Jakobsons (1957)
speech event). Further, Silverstein (1976b) made clear that this pragmatic, indexicality-based theory of linguistic structure was part of the general linguistic-anthropological theory of sociocultural communication, covering not only linguistic
structure, but also referential and non-referential communicative practices, the
latter pertaining to social indexicality, or the indexing of the group identities and
power-relations of communicative-event participants. Thus, indexicality has been
shown to be a key concept in the pragmatic study of contextuality, across the struc-
143
tural, referential, and non-referential domains of language and communication (see

Section 6, below, for more on indexicality; cf. Hanks 1990; Duranti and Goodwin
1992; Agha 2007).
3.
Genealogy of the concepts (contextual) presupposition and effect
Having briefly seen how the notions of indexicality and deictic center have
been made use of to account for the communicative event-centered constitution of
context and linguistic structure, we now move to explore historical developments
in the theoretical understanding about pragmatic processes that take place in, and
(re-)create, context. Here, it appears that the intuitive notions of situational (contextual) appropriateness and effectiveness (or efficacy) of actions have led to,
and still lie behind, more elaborate theorizations of this theme, in terms of, for
example, presupposition and perlocution(ary effect), respectively.
Let us first deal with theoretical notions related to contextual appropriateness.
As noted immediately above, the concept of presupposition, be it semantic or
pragmatic (cf. Karttunen 1974), is a theoretical elaboration of the intuitive notion
appropriateness. The concept was developed by Frege (18481925; see Section 5) and Russell (18721970; a leading figure in the logicist tradition of British
analytic philosophy), and was later picked up by Peter Strawson (19192006; a
major figure in the Oxford Ordinary Language School). Subsequently, the
schools theoretical framework and concepts were discovered and integrated into
linguistics in the late 1960s and 70s by (post-Chomskyan) Generative Semanticists: viz., George Lakoff, McCawley, Robin Lakoff, Ross, Sadock, et al. (cf. Lakoff 1989; Harris 1993; Koyama 2000). As these were immediate predecessors of
linguistic pragmatics (which became established in the 1970s; cf. Mey [1993]
2001: 1927),4 presupposition became a main theme in the new discipline, especially in the 1970s and early 1980s (cf. Gazdar 1979: 89128; Levinson 1983:
167225). Generally speaking, since the 1970s, the studies of presupposition have
been characterized by the increasing focus on pragmatic, rather than semantic, presupposition and the increasing awareness of the centrality of pragmatic context in
matters of presupposition in general.
Similarly, the notion of felicity condition, apparently concerning contextual
appropriateness and derived from the works of Austin (191160) and Searle (b.
1932), became a prominent theme in the new discipline. (Austin was the leading
figure of the Oxford philosophy of language, and Searle was an American-born
student of Austin and Strawson at Oxford in the 1950s). As Austin must have been
at least dimly aware, the theme of felicity condition, which concerns the contextual conditions that must obtain in order for a (speech) act to occur validly, was
actually an old theme in law (see note 7), and, we should also observe, in other disciplines concerned with (judicial or other) rituals, such as sociology and anthropol-
144
Wataru Koyama
ogy, which had the long tradition of investigating the social or cultural (contextual)
appropriateness of rituals and other actions (see Section 7). Thus, in the 1970s and
early 80s, felicity condition, an old theme recently re-thematized by Austins
disciple Searle, became a topic not only in the new discipline of pragmatics, but
also in the social sciences (see immediately below), thereby facilitating the interaction and partial merger between pragmatics, micro-sociology, and linguistic
anthropology. In other words, it appears that this theme may have helped the new
discipline to move in the direction of sociocultural: that is, toward the microsociology of conversation analysis (cf. Levinson 1983: 284370) and, beyond
that, linguistic anthropology, social semiotics, social pragmatics, critical pragmatics, and critical discourse analysis, which have been gaining prominence in
pragmatics since the late 1970s and especially 80s and 90s (cf. Mey [1993]
2001). Note that, in the 1970s and early 80s, when felicity condition was hotly
discussed, the micro-sociologist Goffman (192282) wrote an important article on
it (Goffman 1983), and a variety of concepts coming out of his uvre, such as
face (cf. Brown and Levinson [1978] 1987), frame (cf. Tannen 1986), and
footing (cf. Irvine 1996), began to exercise great influence among pragmaticians, discourse analysts, and linguistic anthropologists (cf. Goffman 1967, 1974,
1981, 1983). On the other hand, in the same period, the more individual-psychological aspects of felicity condition, such as the speakers intention and sincerity,
articulated by Searle, became critically examined from cross-cultural perspectives
by linguistic anthropologists (cf. Rosaldo 1982; Du Bois 1987; Duranti 1993),
working in the tradition of ethnography of speaking/communication (see Section 6, below, for details; cf. Gumperz and Hymes [1972] 1986; Bauman and
Sherzer [1974] 1989). This move anticipated the increasing appreciation of crosscultural differences in pragmatics especially since the 1980s (cf. Mey [1993] 2001:
262288).
Indeed, the social anthropological concern with situational (contextual) appropriateness of actions goes back to a founder of British social anthropology,
Malinowski (18841942), who advanced the notions of context of situation
and context of culture,5 adopted in the 1930s by his close associate, J. R. Firth
(18901960), i.e., the leading figure of the London school of functional linguistics.
The modern anthropological idea, advanced by Malinowski, Radcliff-Brown
(18811955), and Boas (18581942) in the early twentieth century, that any action,
be it linguistic or non-linguistic, should be understood in relation to the situational
or cultural context in which the action takes place, was intertwined with the anthropological doctrine of cultural relativism, which urged that actions be seen and
understood from natives point of view, from within, or emically in the terminology of Pike (1967: 3772). Beyond the disciplinary matrix of twentieth-century social/cultural anthropology, the idea of the cultural relativity of actions, especially linguistic actions, was also shared by Wittgenstein (18891951) and Peter
Winch (192697), the former sometimes considered a precursor of the Ordinary
145
Language School, including Austin, Grice, and Searle, and subsequently, linguistic
pragmatics (cf. Levinson 1983: 227).
Importantly, the (situational or cultural) contextualist doctrine, advanced by
modern anthropology and later adopted by much of linguistic pragmatics, not only
emphasized the role of context in determining the meanings of actions, but also
tended to take context to be an independent macro-social variable determining the
meanings of micro-social actions, the latter considered as dependent variables. In
other words, the doctrine failed to pay sufficient attention to how actions create
and change the contexts (situations and cultures) in which they take place. This
static view of culture was perhaps most pronounced in Radcliff-Browns Durkheimian functionalism, where the meanings of actions were explained by the
(macro-)social functions they served. In this view, which came close to macrosocial determinism, the meanings of micro-social actions (what we do, with or
without words) were equated with the functions they served in the macro-social
order (contextual reality) presupposed to exist independently of actions.
To be sure, in Malinowskis functionalism, which was less Durkheimian and
more Freudian than Radliff-Browns (cf. Kuklick 1991), it was not the macro-social order, but the psychological needs of socially contextualized individuals that
were taken as presupposed variables, teleologically explaining the meanings
(functions) of their actions; all the same, the meanings of actions were functionally explained by the conformity (fit, or appropriate correspondence) of the actions with their presupposed ends, be they psychological or (macro)sociological.
The functionalist doctrine was espoused by much of linguistic functionalism
and the nascent discipline of linguistic pragmatics in the 1970s. For instance, in
these disciplines, actions (utterances, or speech acts) were often explained or theorized by appealing to the notions such as (social or psychosocial) maxims and
conditions, which were, as in Malinowskian functionalism, taken to be independent variables explaining the meanings (implicatures) or efficacy of utterances and
other practices. This functionalist way of seeing speech and other social actions
had been dominant in the post-World War Two behavioral sciences, which became
challenged in the social sciences in the 1960s and 70s, when the idea that society
and culture are contingently (re-)created in and by daily actions and, thus, can be
changed by actions, gained momentum, especially in the anti-establishment,
youth, peace, and (second-wave) feminist movements. In the social sciences, even
before the 1960s, the ethnomethodologist Harold Garfinkel (b. 1917) had begun
frontally to attack the functionalist doctrine, attributed to Durkheim (18581917),
and explore how micro-social, contextualized actions create, and indeed lie at the
foundation of, social organization (cf. Heritage 1984). Interacting with the microsociological studies carried out by Goffman (mentioned above, in this Section),
ethnomethodology subsequently saw a new development called conversation
analysis, explored by Sacks, Jefferson, Schegloff, and their followers, and influencing pragmatics (cf. Levinson 1983: 284370). Meanwhile, since the 1970s,
146
Wataru Koyama
second-wave feminism, critical of the functionalist view of society, has given rise
to womens studies, gender studies, and later, queer studies, a series of developments in which feminist linguistics/pragmatics have been involved (cf. Lakoff
1973; Cameron and Kulick 2003), and in which the term performativity, taken
from Austin, has played a crucial role (cf. Butler 1990). As this indicates, the
speech act theories of Oxonian philosophy and early pragmatics were not exclusively concerned with the question of contextual appropriateness. Rather, they investigated conditions, maxims, and other presupposed contextual variables, at
least partially so as to explain the efficacy, or the creative (performative) function,
of actions, the intuitive appreciation of which was, indeed, the point of departure
for them. And, it was this insight of performativity that attracted theoretical interests of many literary critics, historians, social scientists, and feminists, to pragmatics (see Section 7 for more on performativity). Thus, interacting with them,
pragmaticians moved to construct pragmatic theories focusing on the creative,
rather than presupposing, function of language use, especially as related to powerrelations and group identities of language users. As a result, not only feminist pragmatics, but also critical linguistics/pragmatics, associated with the Frankfurt
School, the New Left, and neo-Marxism, came into being in the late 1970s (cf. Mey
1979, 1985, [1993] 2001), in critical opposition to the functionalist view of society.
Eventually, the 1980s and 90s saw the rise of critical discourse analysis in Western
and Central Europe. Even sociolinguistics, in the wake of criticisms against the
functionalist doctrine (cf. Williams 1992), has begun to move in the direction of
social constructivism, investigating how social realities and systems are dynamically created by discursive interactions (cf. Coupland, Sarangi, and Candlin 2001).
4.
Genealogy of the idea (communicative) functions

(multi-functionality)
Thus, appropriateness and effectiveness, that is, the presupposing and creative
functions of interactional practices, constitute one dimension of pragmatic processes in context, around which the history of pragmatics can be described. Crosscutting this is another dimension of pragmatic processes in context: viz., the dimension of what one might call the functional (indexical) orientations of
communication, such as the referential (descriptive), expressive (emotional), and
appeal functions (cf. Bhler [1934] 1990). Jakobsons (1960) six-functions
model, one of the most solid models of the multi-functionality of communication,
identified the emotive, conative, referential, metalingual (metalinguistic), poetic,
and phatic functions, respectively characterized by the six factors constituting
the event of communicative practice, i.e., addresser, addressee, referent, code,
message, and channel (contact; see Section 6 for more details). Indeed, the term
function has been usually used in this sense in linguistic functionalism and prag-
147
matics, and the history of pragmatics has been often written by tracing the history
of studies dealing with these functions in linguistics, philosophy, psychology, law,
and the social sciences (cf. Nerlich and Clarke 1996). Although Austins theory of
speech act seems to have arisen more or less independently of these functionalist
studies, which had blossomed from the 1880s to the 1930s (see Section 5), it is
clear that they anticipated the basic insight of Wittgenstein and the Ordinary Language School (including Austin, Warnock, Urmson, Grice, Strawson), to be absorbed by Generative Semantics and linguistic pragmatics from the late 1960s onward, that language includes many functions other than the referential function or
propositional meaning, which appears to have been the exclusive concern of logic,
structural linguistics, and generative syntax.
That is, as Nerlich and Clarke (1996) have shown, the social, psychological,
linguistic, and philosophical studies of the non-referential functions (multi-functionality) of language and communication, were vigorously pursued especially
from the 1880s to the 1930s. This was the period of the linguistic turn (cf. Rorty
1967; Apel 1994), before the rise of formalist structural linguistics, which concentrated on the formal and semantic, but not pragmatic or indexical aspects of language, in the 1940s. To be sure, the functionalist traditions were kept alive in the
works of Jakobson, Benveniste, Kuryowicz, Pragueans, and (neo-)Firthians, to
name but a few important figures and schools, but the dominance of formalists
(such as neo-Bloomfieldians and Chomskyans) in structural and generative linguistics, associated with the philosophical project of logical syntax, in the United
States and beyond, continued until the 1970s (see Section 2). As a result, the earlier
functionalist theories and their achievements became largely forgotten, so that,
when Generative Semanticists re-discovered the importance of context (pragmatics) and started to rebel against the orthodox Chomskyan doctrine of autonomous syntax in the late 1960s, they made appeal to the philosophical theories of
Austin, Grice, Strawson, and Searle, but not to the older functionalist traditions (cf.
Harris 1993). Linguistic pragmatics, which arose in the 1970s in the wake of Generative Semantics, was no less strongly influenced by the philosophical theories,
but it also actively interacted with the offspring of the older functionalist traditions,
i.e., a variety of functionalist approaches to grammar, text, and discourse, which
saw a renaissance in the 1970s (cf. Bolinger 1968; Comrie [1981] 1989; Halliday
and Hasan 1976; Leech 1983: 4678). Thus, the importance of the notions advanced by the older traditions, especially deictic field (Zeigfeld), indexicality,
and metapragmatics (the latter two derived in part from Jakobsons (1957, 1960)
code-and-message model and metalingual function, respectively; cf. Silverstein 1976b), has become more fully appreciated, and a number of attempts have
been made to explore and elaborate on the implications of these important notions,
especially in the recent literature of pragmatics (see Sections 2 and 7; cf. Mey
[1993] 2001; Verschueren 1999; Bublitz and Hbler 2007).
148
5.
Wataru Koyama
The linguistic turn and the philosophy of language: functionalism,

phenomenology, and neo-Kantianism
The older traditions of functionalism, as noted above, blossomed from the 1880s to
the 1930s, the period when philosophy took the linguistic turn. Here, focusing on
the disciplinary development of philosophy, which was closely linked to the rise of
semantics and pragmatics, we note that the late nineteenth century saw the returnto-Kant movement, as can be seen in the Kant-inspired pragmatic philosophy of
Peirce in the United States and, especially, the epistemology of neo-Kantians in
Germany, whose opponents included the logico-mathematician Frege (cf. Coffa
1991). This suggests that logico-analytic philosophy, other kinds of the philosophy
of language, neo-Kantianism, and phenomenology were, directly or indirectly, all
involved in the linguistic turn, and interacted with each other, as well as with the
older traditions of functionalism (cf. Richardson 1998; Friedman 2000). Let us
briefly explore historical connections among them, taking Frege as our starting
point.
Needless to say, Frege (18481925) was one of the founders of modern predicate logic and semantics, and his distinction between sense (Sinn) and reference
(Bedeutung) anticipated the later division of semantics and pragmatics, articulated
by Morris and Bar-Hillel (see Section 2, above). No less important, Freges differentiation between proposition and force (Kraft), anticipated by Condillac among
others,6 was a precursor of the distinctions between constatives and performatives,
and between locutionary, illocutionary, and perlocutionary forces, advanced by
Austin, who was a translator of Freges Die Grundlagen der Arithmetik (The
Foundations of Arithmetic). As is well known, these distinctions underlay the formula F(p) (where p and F stand for proposition and illocutionary force, respectively), articulated by Austins student, Searle (cf. Nerlich and Clarke 1996:
22, 76, 202204). Interestingly, the later Searle (1983) advanced a phenomenologically-oriented theory of speech, as suggested by the title Intentionality (a notion
derived from Husserl (18591938), Brentano (18381917), and, ultimately, the
Scholastic concept of intentio); and Searles master, Austin, intended his philosophy of language to be a kind of linguistic phenomenology.7
Thus, there appears a link between Freges logic(ism) and Husserlian phenomenology, although Husserl, who started his career as a philosopher of mathematics
and logic, moved away from logic towards transcendental phenomenology and,
eventually, the phenomenology of the life-world (Lebenswelt). Subsequently, phenomenology inspired various kinds of the philosophy of language: e.g., not only
Austins and Searles, but also the French dialectician Merleau-Pontys (1960),
who exchanged his views with Oxonians, viz., Ryle, Austin, and Urmson, at a colloquium in 1958. Further, phenomenology also influenced the critical social theory
of Habermas (b. 1929; cf. Mey [1993] 2001: 315), who has adopted the notion of
Lebenswelt, and, earlier, the sociology of Alfred Schutz (18991959), who carried
149
it from Vienna to New York. In North America, phenomenological sociology was

then carried out by Berger and Luckmann (cf. Mey [1993] 2001: 115) and by ethnomethodologists, including Garfinkel (cf. Giddens [1976] 1993; Bourdieu 1980).
As linguistic pragmatics has been variously influenced by ethnomethodology, conversation analysis, and Critical Theory, it can be said that the phenomenological
orientation has entered linguistic pragmatics not only from the Ordinary Language
School (see immediately above), but also from phenomenological sociology.
Hence, the phenomenological strand is fairly strong in linguistic pragmatics. After
all, the discipline was established in reaction against formalist structuralism in linguistics, which was, unlike the phenomenologically-oriented functional theories of
Bhler, Jakobson, and the Prague School (cf. Holenstein 1977), non- or even antiphenomenological.
Thus, although analytic logic and phenomenology may appear polar opposites,
there is a linkage between them, as indicated by the careers of Husserl and Austin,
as well as by the history of linguistic pragmatics, which has been, since the late
1970s, moving from de-contextualized logic and maxims towards contextualized
events, i.e., social interactions in historical and cultural life-worlds.
Further, this strongly suggests a certain affinity between linguistic pragmatics
and the philosophical movement of neo-Kantianism. Here, observe that, just like
pragmatics, neo-Kantianism seems located between logic and phenomenology, as
witnessed by the fact that Peirce, who belonged to the Return-to-Kant Movement,
invented a semiotically-based predicate logic, distinct from Freges arithmetic
one, and called his pragmati(ci)sm phenomenology and phaneroscopy.8 Also,
one may note that a neo-Kantian philosopher par excellence, namely Cassirer
(18741945), who belonged to the Marburg School of neo-Kantianism and emigrated to New York, contributed an important paper (Cassirer 1945) on the status
of structural linguistics to the New York-based journal Word, run by Martinet
(190899), Jakobson, and other structural-functionalists, who were opposed to
native-born neo-Bloomfieldian formalists in the United States. This points to the
linkage between neo-Kantianism and functionalist linguistics, and the latter was,
as we have seen, intertwined with phenomenology, especially in the works of Jakobson and Bhler.9
As a matter of fact, Cassirers career, characterized by the application of
(neo)Kantian philosophy, in particular epistemology, onto empirical investigations
into myths, languages, and the history of ideas, was part of the historical transformation, or empirical turn, of Kantian philosophy, which moved from pure and
practical (praktisch in Kants terminology) reason, i.e., critical-philosophical,
transcendental principles of epistemology and praxiology, towards more empirical,
historical, anthropological, pragmatic (pragmatisch) studies of knowledge and actions, at the turn of the twentieth century. Crucially, we should see that, in general,
when knowledge and actions are studied empirically, it becomes clear that they are
mediated by language and communication. Here, recall from Section 2, that Locke
150
Wataru Koyama
(16321704), one of the founders of modern empiricism, felt compelled to write a

treatise on human understanding, in particular, language, communication, and semiotics; and Thomas Reid (171096), deeply influenced by Locke (as well as by
Hume, a Scotsman like Reid), was a precursor of modern pragmatics and, perhaps,
phenomenology (cf. Mulligan 1987: 3334). Subsequently, Peirce (18391914),
influenced by Locke, Reid, and Kant (17241804), developed a semiotic philosophy of communication, which he called pragmati(ci)sm, based on Kants term
pragmatisch (see immediately above). The term was adopted by the behavioristic
pragmatist Morris, who coined the term pragmatics in his famous trichotomy of
communication (see Section 2). We still operate with Morriss terminology, i.e.,
syntax, semantics, and pragmatics; and this fact suggests the intermingling of not
only the terminological but also the conceptual genealogies of todays linguistic
pragmatics and the neo-Kantian movement in philosophy.
Thus, the empirical turn of (neo)Kantianism seems to lie at the roots of linguistic pragmatics. Here, notice that the heyday of neo-Kantianism in philosophy
coincided with the linguistic turn, i.e., the period in which the older traditions of
functionalism blossomed, from the 1880s to the 1930s (cf. Nerlich and Clarke
1996). This cannot be by chance. By the linguistic turn, judgment, a concept integral to Kantian philosophy, became replaced with proposition in post-Fregean
logic, and, more generally, with language in twentieth-century philosophy, including logico-analytic philosophy and the Ordinary Language School;10 at the
same time, (neo)Kantianism turned empirical, giving rise to pragmati(ci)sm and
the empirical science of language, i.e., linguistic functionalism, studying the synchronic, living state of language in action, context, and communication. This
movement, replacing the nineteenth-century historical comparative linguistics
(over)systematized by Neo-Grammarians, was what was called (functionalist)
structuralism (cf. Jakobson 1973), which would become overshadowed by formalist structuralism, the latter turning away from context and pragmatic functions to
formal structure and semantics in the 1940s and 50s.
We shall soon come back to the synchronic turn in the studies of language,
corresponding to the linguistic turn in philosophy: i.e., the rise of structuralism as
part of linguistic functionalism, between the 1880s and the 1930s. But, first, we
must observe that the empirical turn of (neo)Kantian philosophy meant the emergence of the empirical sciences of not just language, but also the mind, society, and
culture: i.e., modern twentieth-century psychology, sociology, and anthropology.
Note that William James (18421910), Wundt (18321920), Max Weber (1864
1920), Simmel (18581918), Durkheim, Mauss, Boas, Malinowski, and RadcliffBrown, who are usually counted as founders of these disciplines, worked in this
period, from the 1880s to the 1930s, and some of them were explicitly associated
with pragmatism and philosophical neo-Kantianism, while others directly or indirectly communicated with them. Further, whether or not they were pragmatists
or neo-Kantians in the narrow sense, their empirical sciences and theories gen-
151
erally shared the epistemological framework of neo-Kantianism, such as the methodologically-based distinction between nomothetic sciences, investigating laws
and other regularities, and idiographic (or historical) sciences, exploring unique
and singular, sociohistorically contextualized phenomena (cf. Stocking 1996). Together with the related opposition between collectives and individuals, this distinction underlay, among others, the Saussurean notions of langue and parole, the
proto-Weberian Methodenstreit concerning the status of economics as a socio-historical (idiographic) or ahistorical (nomothetic, classical Newtonian-Laplacean)
science, and the Durkheimian understanding of sociology in opposition to individualist psychology.
Thus, (neo-)Kantianism was the philosophical foundation for the rise of the
human sciences (cf. Foucault 1966), including linguistics and, later, pragmatics.
The linguistic turn in philosophy and the (functionalist) structuralist turn in linguistics were part of the (broadly conceived) neo-Kantian movement, which also
included the rise of phenomenology. That is, in the studies of language, the linguistic turn at the beginning of the twentieth century consisted of the move towards
the synchronic, living state of language in action (i.e., pragmatics); and the rise
of structuralism should be understood as part of the functionalist, phenomenological movement, oriented towards actions and other happenings taking place hic et
nunc, or the dectic center (origo) of such pragmatic occurrences (see Section 2). As
Jakobson was never tired of pointing out (cf. Jakobson 1973; Holenstein 1977),
structuralism, i.e., functionalist (vs. formalist) structuralism, was not opposed to
phenomenology; they constituted a unified whole (holistic Gestalt), as it were. The
philosophy underlying structuralism was the neo-Kantian one, represented by the
idea of the primacy of (human) (ap)perception, advanced against nineteenthcentury positivism and mechanical empiricism, which had culminated in the NeoGrammarians comparative (historical) linguistics in the late nineteenth century,
and which had earlier served to suppress the proto-linguistic turn, achieved by
Bernhardi and Humboldt in Germany, Condillac and the Idologues in France, and
Reid and the followers of his commonsense philosophy such as Dugald Stewart
(17531828) at the turn of the eighteenth to the nineteenth century (see notes 1 and
2, above; cf. Nerlich and Clarke 1996). The primacy of human perception, or subjective apperception, clearly underlies the structuralist notions of phoneme and
sound pattern, advanced in contradistinction to mechanical physical sounds,
studied by the empiricist, positivist science of nineteenth-century phonetics.
Further, phonemes were considered part of not just subjective (individual), but
also intersubjective, sociocultural reality (see Section 6). The insiders of a community sharing (by degrees) a linguistic structure, i.e., members of a linguistic
community, (ap)perceive the (more or less) same phonemes, but not phones;
they have phonemes as part of their psychological reality (cf. Sapir 1933), that
is, intersubjective phenomenological reality, a pattern of (ap)perception not shared
by the outsiders of the (Durkheimian) socio-cultural, linguistic-structural commu-
152
Wataru Koyama
nity. Thus understood, it becomes less mysterious why the phenomenologist Merleau-Ponty (1945, 1960) appreciated both Gestalt psychology and Saussures linguistics, celebrating the latters discovery of synchrony and parole, and why
Saussure was closely associated with Bral (18321915), a founder (precursor) of
modern twentieth-century semantics and pragmatics, in Paris (cf. Aarsleff 1982;
Nerlich 1992). For all of them, including even Saussure, language was, au fond, an
intersubjective, phenomenological reality, pragmatically unfolding in front of our
eyes, that is, emerging at the synchronic hic et nunc of communication. Thus, the
historical and theoretical roots of modern structural linguistics and linguistic pragmatics converge in the rise of structuralism, i.e., functional structuralism, in the
period from the 1880s to the 1930s, that is, the days of the older traditions of linguistic functionalism, to be eclipsed when structural linguistics became dominated
by neo-Bloomfieldian, Chomskyan, and other kinds of dogmatic formalism after
the achronic (vs. synchronic) turn in the 1940s.
6.
Intersubjectivity and interaction: functionalism, Ordinary Language

philosophy, and the social turn in pragmatics
Intersubjectivity involves not only (1) the speaker (addresser) and (2) the hearer
(addressee), but also (3) the channel (contact) between them, (4) the sign (message) that is sent from the former to the latter, (5) the (intersubjectively) shared
code(s) for interpretations of the sign, and (6) the (intersubjectively) understood
referent(s) of the sign, located in the micro- or macro-context of the interaction between the speech- or communicative-event participants, such as the speaker and
addressee. This is, of course, the contextual, pragmatic matrix of communication,
including language, which was articulated in the Praguean functionalist theory of
Jakobson (1960), mentioned in Section 4. As part of the code, linguistic structure is
embedded in this functional matrix of communication. In other words, structure
emerges through communicative practices; there is no structure (langue) without
pragmatics (parole). According to this Jakobsonian theory, as noted earlier, there
are six basic functions of communication, corresponding to the six aspects of the
contextual matrix: viz., (1) emotive, (2) conative, (3) phatic, (4) poetic, (5) metalingual (metalinguistic), and (6) referential functions.11 Although linguistic structure, as part of the code, is primarily used to interpret the sign (token) to identify
(i.e., intersubjectively construct) its referent(s) and thus primarily involved in the
metalinguistic, poetic, and referential functions (respectively corresponding to the
code, sign, and referents), it inevitably interacts with the other functions as well,
because it exists only as part of the code pragmatically used for intersubjective interpretations of the sign indexically exchanged in the context of communication.
Thus, language, including linguistic structure, i.e., (morpho)phonology, (morpho)syntax, and semantics, should be understood, theorized, and analyzed by tak-
153
ing into proper consideration the six functions, as well as the six aspects, of communicative practices (cf. Silverstein 1976a, 1976b, 1993).
But, of course, the contextual, functional matrix of communication, articulated
in the six-functions model, is not limited to, or centered around, linguistic structure.
For example, the code includes not just linguistic structure, but also maxims, interpretive rules of thumb, passing theories, metapragmatic frames, and models
of interpretation vis--vis (not necessarily linguistic) signs, including gestures, facial expressions, and other paralinguistic phenomena (cf. Bolinger 1968), all of
which occur in specific, sociohistorical contexts of communicative practices (see
Section 7; cf. Mey [1993] 2001: 206235). And such contextualized practices
index12 not just referents (in the referential function) and the semantic and pragmatic codes being used (in the metalinguistic function). They also index the social identities of, and power-relations among, communicative-event participants
(e.g., in the emotive and conative functions). That is, contextualized practices
index not just referents, but also social identities and power-relations, which can be
read in the emergent texts of sociocultural practices of communication. Such
texts are not to be understood as static objects, but as the dynamically emergent,
socioculturally interpretable organizations of referential and non-referential (interactional) practices (mainly corresponding to the poetic, phatic, and metapragmatic functions), that is, organizations studied in conversation analysis (cf.
adjacency pairs; Levinson 1983: 284370), discourse analysis (cf. frames of interaction; Goffman 1974), linguistic anthropology (cf. Silverstein and Urban 1996;
Blommaert 2005), and, of course, pragmatics (cf. Mey [1993] 2001).
Thus, in a sense, the Praguean/Jakobsonian functional theory provided a general framework in which todays empirical approaches to language, listed immediately above, operate (also see Section 2). As this indicates, the rise of structuralism
in the late nineteenth and early twentieth centuries served to prepare the proper theorization of the functional, contextual matrix of language, which would emerge in
the 1970s after the formalist lapse from the 1940s to 60s. This lapse, however,
made obscure the pragmatic, functionalist foundation of structural linguistics. Indeed, by the late 1960s, when Generative Semanticists, who were immediate predecessors of linguistic pragmaticians, started to rebel against formalist structuralism (Chomskyan generativism) and delve into the realm of pragmatics, North
American linguistics had become more closely linked to the philosophy of language, especially logico-analytic philosophy, than to the early twentieth-century
functionalist traditions of structuralism, so that, in trying to find theoretical tools to
deal with pragmatic phenomena, Generative Semanticists discovered and adopted
the theories of speech act, implicature, and presupposition, articulated by Ordinary
Language philosophers like Austin, Grice, and Strawson (see Section 3, above; cf.
Levinson 1983). The publication of Austins (1962; second edition, 1975) William
James Lectures, originally delivered in 1955 at Harvard, which is located in the
same city as MIT (Cambridge, Mass.), i.e., the home of generativism, made the
154
Wataru Koyama
theories of the Ordinary Language philosophers more visible to generativists interested in pragmatic matters (also cf. Grice 1967; Chapman 2005: 100). And Searle
(1969) provided a desideratum, i.e., the desired systematization of speech act theory, making it amenable to the linguists formal, syntactic treatment of pragmatic
phenomena.13 As Generative Semantics yielded to linguistic pragmatics in the
1970s, the early years of the new discipline were marked by the preoccupation with
the issues articulated by the Ordinary Language philosophers, and many of the culture-specific assumptions of their philosophy, i.e., post-Lockean modern British
philosophy, concerning human communication, seem to have been directly borrowed into linguistic pragmatics.
Here, recall that Locke, a founder of modern British empiricism, discussed,
like Saussure, the arbitrary nature of human language (communication), but, unlike the latter, Locke meant by the term the volitional, intentional, willful character of the individuals use of language, whereas Saussure meant the conventional
(non-volitional), extensionally unmotivated (intensional) yet socially existing (i.e.,
Durkheimian) association of the signifiant and the signifi. As this indicates, British and Anglo-American philosophy, especially the empiricist tradition thereof
(see Section 2), has tended to theorize communication (more generally, human
practice) in terms of volitional, intentional actions or, in sociological terms of Max
Weber, the purposive function (cf. Silverstein 1976b; Apel 1994: 63). Ordinary
Language philosophy, an empiricist philosophy of language and communication
concerned with what we do with words and what is meant, is no exception. It
sees language and communication as purposive activities, as explicitly articulated
by G. E. M. Anscombe, G. H. von Wright, and Grice, who tried to explain the social, interactional fact of communication in terms of what we intend to do or what
we intend to mean, that is, individuals conscious intention (cf. Chapman 2005).
Notwithstanding, as Weber and other social scientists have shown, actions have
consequences (perlocutionary effects) which are unintended by agents (subjects),
and which may be more important, in terms of historical consequences (i.e., pragmatic effects), than actors mere intentions (cf. Silverstein 1976b). Further, any
practice at least partially consists of unintentional, non-conscious elements; obviously, we are not conscious of everything we do. For example, we, as ordinary
language users, are usually more aware of surface segments (continuous, segmentable units), such as phonemes, words, and expressions, than of allophones and
discontinuous syntactic units (cf. Lucy 1992); yet, the latter two are often part of
dialectal/sociolectal varieties and thus index their users social identities, pertaining to class, gender, status, etc., and power-relations among them, as can be readily
seen in the sociolinguistic literature, Labovian or otherwise. Thus, there are nonpurposive pragmatic functions in discourse, which are too important to be left out
in any theory of pragmatics.
In addition, cross-cultural comparisons suggest that many societies place much
less value on speakers intentions in the evaluation of (sociocultural) meanings
155
of their linguistic practices than the culture of Ordinary Language philosophers

does, a point poignantly raised and substantiated by linguistic anthropologists,
who have been critical of the intentionalist and individualist bias of the modern
Standard Average European speech act theories ever since the 1970s and early
80s.14 Subsequently, partly as a result of this cross-cultural (anthropological)
critique, linguistic pragmatics has come to appreciate and focus on cross-cultural
differences in speech acts and, more generally, linguistic practices and their
metapragmatic understandings, the latter now discussed under the rubric of language ideology.15
No less important, the individualist assumption of the speech act theories was
undermined by empirical studies of discourse, which showed that the meaning(s)
of an action (speech act) cannot be determined out of context, i.e., without considering a series of interactions which contains the action (cf. Hancher 1979; Levinson 1983). This means that, to the extent that the meaning(s) of an action is
determinable, it is determined by the contextualized interaction of speech-event
participants, not solely by the individual speakers intention. Contributions of conversation analysis (cf. Levinson 1983: 284370) and discourse analysis (cf. Mey
[1993] 2001: 191), focusing their attention on interaction rather than on solitary individual actors minds, have been important in this interactional turn (social
turn) of pragmatics, which has characterized the discipline since the 1980s or so
(cf. Thomas 1996). Pragmatics today is much more interaction- or discourse-centric than in the 1970s, when the new discipline was more intention- and individualcentric under the influence of Ordinary Language philosophy and modern British
empiricism.
7.
Genealogy of performative utterances and other multi-functional

co-incident happenings in the speech event
Let us finally move to the question of performativity, another central theme in linguistic pragmatics. Clearly, this is a complex notion, only a part of which is accounted for by the creative function (efficacy), discussed in Section 3. That is, in
addition to the creative function, performativity involves self-referentiality (cf.
Recanati 1979),16 which is itself a composite notion, made up of reflexitivity and
(transparent) referentiality. The latter means the (relatively) transparent relationship between the referential function (what is said) and the non-referential, interactional function (what is done), as observed when we use a referentially explicit
metapragmatic utterance, such as When I said that, I meant it to be a compliment
(cf. Caffi 1984; Mey [1993] 2001: 173205). On the other hand, reflexivity as
such is just the mutual indexing of some semiotic elements, linguistic or non-linguistic, co-occurring in the event of communication. For example, a pointing gesture and a token of the deictic expression this book, occurring together, mutually
156
Wataru Koyama
index each other and function reflexively, as the gesture and the linguistic utterance
coincide in the same context (deictic, indexical field; see Section 2). Mutual, reflexive indexing also obtains when a defiant smile, or a serious demeanor, occurs
with an utterance Im sorry, or when a judge, properly dressed or completely undressed, utters I declare you guilty. Clearly, these are examples of what Gregory
Bateson ([1955] 1972) called metacommunication (cf. Ruesch and Bateson
1951; Caffi 1984) and what Gumperz (1982) called contextualization cue (cf.
Auer and Luzio 1992; Duranti and Goodwin 1992), notions coming out of the anthropological, psychiatric, and ethological traditions, independently of Austins
discussion of performatives.
Austins (1962) notion of performativity, following the discussions by the phenomenologist Adolf Reinach (see note 7) and others in the study of law, did not
clearly separate these two dimensions of performativity or explore each dimension in depth. Originally, linguistic pragmatics was under the influence of Austins
way of formulating the problem, but it has gradually moved to see that the problem
of performativity should be dealt with by separating out different principles underlying the problem, such as the principle of metapragmatics, which is becoming a
major topic in our discipline (cf. Ducrot 1984; Verschueren 1999; Mey [1993]
2001). Like the cognate notion metacommunication, the concept of metapragmatics came out of the anthropological literature, in particular Silverstein
(1976b), who coined the term to designate the pragmatic (vs. semantic) aspect of
the metalingual function,17 articulated by his teacher, Jakobson (1957, 1960),
who, in turn, derived it from the discussion of metalanguage (vs. object language) by the mathematician David Hilbert (18621943), Kurt Gdel (190678),
Tarski and other logicians, e.g., Russell, the early Wittgenstein, and Carnap (cf.
Apel 1994). The metapragmatic function obtains when a sign, be it an utterance, a
gesture, or whatever, indexes another sign (speech, non-linguistic action, etc.) in
context. When the two signs co-occur (coincide) in the same context, they are said
to show reflexively calibrated metapragmatics (cf. Lucy 1993; Mey [1993]
2001). According to Silverstein (1993), the reflexive metapragmatic function
serves to create texts, be they referential or non-referential (see Section 6 on
text). For instance, the reflexive co-indexing of referential signs, such as (thematic) anaphor and its (rhematic) antecedent, in the same unit of discourse creates
referential textuality, as studied in Praguean functional structuralism, text linguistics, and discourse analysis (cf. Halliday and Hasan 1976). Similarly, the reflexive co-indexing of non-referential signs, such as the two units of an adjacency
pair (e.g., request-compliance or -rejection; cf. Levinson 1983: 284370), in the
same phase of discourse creates interactional, non-referential textuality.
Clearly, the idea of textualization came from, among other places, Bhlers
notion of Sprachwerk language work and, in particular, Jakobsons poetic function (see Section 6), i.e., the communicative function that is focused on messages
(i.e., signs) occurring in the context of interaction and thus gives textual organiz-
157
ation to them, as can be most clearly seen in parallelisms and repetitions abundantly observed in poetry and other rhetorically elaborate verbal arts (e.g., folklore
and myths; cf. Bauman 2004). Importantly, basic units of poetic text are often nonreferential segments, such as meters and rhymes, in contradistinction to referential
units such as rhemes, themes, and syntactic segments like phrases, clauses, sentences. Thus, poetic texts are, at least partially, non-referential texts, i.e., texts
of non-referential interaction. And hence, poetic texts tend to become endowed
with non-referential meanings, such as emotional (emotive) and persuasive
(conative) forces (functions). The same characteristics are found in rituals,
which also typically show the abundance of parallelisms and repetitions in the linguistic and non-linguistic, interactional modes (cf. Caton 1990). To be sure, in ordinary conversations too, we also find such interactional poetic parallelisms and
repetitions especially when they contain what Goffman (1967) called interaction
rituals (cf. greetings and other adjacency pairs, mentioned above). Yet, the characteristic is most salient in rituals, in which, we should recall, Austin (1962) discovered (explicit) performatives: e.g., christening rituals, legal rituals, etc. Performative utterances are the part of rituals characterized by the use of little ritual
one-liners, i.e., performative sentences, which are, thus, socially recognized formulaic types, recursively co-occurring with the non-linguistic elements of ritual
settings (cf. Volosinov [1929] 1973; Silverstein 1993).
Thus, performatives show reflexivity, i.e., the reflexive metapragmatic
function, which is observed more broadly in rituals, poetry, metacommunication,
contextualization cues, etc. On the other hand, performatives involve another
principle, that is, referentiality, or relatively pronounced degrees of transparency between the referential function (what is said) and the non-referential, interactional function (what is done), as noted above. Clearly, it is this principle that
has been the principal focus of pragmatic discussions on performativity and
speech act, most clearly observable in Searles theory of indirect speech act and
Grices theory of implicature. These two notions were formulated to account
for non-explicit or implicit relationships between the referential and non-referential functions of utterances, and they have been main themes of linguistic pragmatics since the 1970s. Much of linguistic pragmatics still operates with these
concepts. For example, Brown and Levinsons ([1978] 1987) politeness theory
and, inter alia, Sperber and Wilsons Relevance Theory seem squarely located in
this tradition.
Clearly, the theme of referential explicitness vs. implicitness goes back to the
old tradition of rhetoric, where literal expressions and tropes were understood in
such terms (cf. Ricur [1975] 1977). When Austin discovered non-referential
functions in explicit performatives, he did not depart from this age-old tradition, as
he focused on non-referential functions only insofar as they appear together with
the referential function, as in performatives, where the two kinds of function
happen to coincide. Just as in ancient rhetoric, Austin and his followers moved
158
Wataru Koyama
from the referential function, as studied in logic and formalist structuralism, to

non-referential functions. That is, Austins was a reference-centric approach to
pragmatic functions, the severe limit of which is indicated by the comparison with
multi-functionalist approaches to pragmatics found in, e.g., Jakobsons theory (see
Section 6), its developments in linguistic anthropology (see Section 2 and immediately above), and the theory of pragmatic act, advanced in social pragmatics (cf.
Mey [1993] 2001: 206235; also see Section 3, above). The rise of social pragmatics and linguistic anthropology in linguistic pragmatics since the 1990s seems
to suggest that the discipline is not only recovering but also further elaborating the
multi-functionalist, social interactionist insights of the older traditions of linguistic
functionalism, which blossomed from the 1880s to the 1930s (see Section 2).
Notes
1. More specifically, Reids philosophy influenced Benjamin Smart (17861872), Adolphe
Garnier (180164), and Henry Sidgwick (18381900), and developed into Charles
Peirces (18391914) semiotic pragmati(ci)sm, Charles Morriss (190179) behaviorist
pragmatism, Sidgwicks student George Stouts (18601944) empiricist and behavioristic philosophy, the latters student G. E. Moores (18731958) commonsense empiricism, and Grices (191388) theory of non-natural meaning and implicature in the
nineteenth and twentieth centuries (cf. Nerlich and Clarke 1996: 2021, 105106; Chapman 2005: 22, 25, 7172).
2. The German tradition ran from Bernhardi (17691820) and Wilhelm von Humboldt
(17671835) to Philipp Wegener (18481916), the later Karl Brugmann (18491919),
Anton Marty (18471914), and Bhler, whereas the French one ran from the Idologues
to Bral, (18321915), Meillet (18661936), Mauss (18721950), Bally (18651947),
Guillaume (18831960), and Benveniste, to mention only a few of the major figures.
3. This was preceded by the medieval trivium (artes sermocinales), i.e., grammar, logic,
and rhetoric. Note that Morriss predecessor Peirce was thoroughly familiar with Scholastic philosophy.
4. The Journal of Pragmatics was launched in 1977 (cf. Haberland and Mey 1977). Subsequently, major textbooks of pragmatics became published in the early 1980s (cf. Leech
1983; Levinson 1983), the International Pragmatics Association (IPrA) was founded in
1986, and Pragmatics was started in 1991, while the second wave of major, and more
Continental-European (vs. Anglo-American), textbooks appeared in the 1990s (cf. Mey
[1993] 2001; Verschueren 1999).
5. These ideas had been suggested and, in part, advanced by Philipp Wegener (18481916).
6. For instance, before Condillac (171480), Grammaire gnrale et raisonne had made
the distinction between dictum and modus, roughly corresponding to proposition and
force, respectively.
7. Also, note that Adolf Reinach (18831917), whose theory of social acts was one of the
precursors of Austins theory of speech acts and performativity, was a phenomenologist. Reinach was a student of Husserls and, with Johannes Daubert (18771947), belonged to the Munich School of phenomenology. The Scottish commonsense philos-
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
159
ophy of Thomas Reid (see note 1), which underlay Ordinary Language philosophy and
much of twentieth-century British empiricism, also exercised influence over the early
phenomenologists, such as Brentano (18381917), one of the predecessors of Reinach.
(Reinach was under the influence of the (individualistic) functionalist Anton Marty
(18471914), a student of Brentanos.) Further, it should be underlined that Reid, Reinach, and Austin all used the philosophy of law as a common source of reference for their
philosophies of language (see Section 7, below; cf. Mulligan 1987: 3334; Nerlich and
Clarke 1996: 12, 104, 211215).
Peirce and Husserl were both influenced by Bolzano (17811848), who had been, in
turn, influenced by Johann Lambert (172877) and Locke; the latter two also directly influenced Peirce (cf. Nerlich and Clarke 1996: 21).
Bhler was also closely associated with the Vienna Circle, including Carnap. As this
suggests, and as Friedman (2000) has shown, logical positivism (of the Vienna Circle)
and phenomenology shared common themes (e.g., intersubjectivity, construction of the
world, etc.) and common roots (e.g., neo-Kantianism). For instance, the early Carnap
made appeal to Gestalt psychology and epistemological holism, which were intimately
associated with structuralism, phenomenology, and especially, the neo-Kantian notion
of constitution.
Generally speaking, the linguistic turn in philosophy was made up of (1) the Russellian critique of ordinary language by means of logico-mathematical analysis, (2) the
Carnapian critique of metaphysical dogmas (later, epistemology), such as realism and
phenomenalist positivism, by means of logic, and (3) the Austinian critique of philosophy, including logic, by means of the commonsense analysis of ordinary language.
Austins philosophy of language was anticipated by Scottish commonsense philosophy,
especially Reid, who influenced Peirce, Moore, and Grice, among others (see note 1,
above).
See Section 7, below, for the metalinguistic (more specifically, metapragmatic) and
poetic functions.
See Section 2 for indexicality.
Recall Rosss and other Generative Semanticists performative hypothesis (cf. Sadock 1974; Levinson 1983: 247263, 271273).
See Section 3, above; cf. Silverstein 1976b; Rosaldo 1982; Du Bois 1987; Duranti 1993.
Cf. Schieffelin, Woolard, and Kroskrity [1992] 1998; Verschueren 1999; Mey [1993]
2001: 173205, 262288; Jaworski, Coupland, and Dariusz 2004.
As Nerlich and Clarke (1996) have pointed out, the property of self-reference (soirfrence) had been discovered, in 1929, by Erwin Koschmieder (18951977), who
started to discuss Koinzidenzfall the case of coincidence in the 1940s.
See Section 6, above.
References
Aarsleff, Hans
1982
From Locke to Saussure: Essays on the Study of Language and Intellectual
History. Minneapolis, MN: University of Minnesota Press.
Agha, Asif
2007
Language and Social Relations. Cambridge: Cambridge University Press.
160
Wataru Koyama
Apel, Karl-Otto
1994
Karl-Otto Apel: Selected Essays, Volume 1: Towards a Transcendental Semiotics. Edited by Eduardo Mendieta. Atlantic Heights, NJ: Humanities Press.
1992
Austin, John L.
1962
How to Do Things with Words: The William James Lectures Delivered at Harvard University in 1955. Edited by James O. Urmson. Oxford: Clarendon
Press.
Bar-Hillel, Yeoshua
1946
Analysis of correct language. Mind 55: 328340.
Bar-Hillel, Yeoshua
1954
Indexical expressions. Mind 63: 359376.
Bateson, Gregory
[1955] 1972 A theory of play and fantasy. In: Gregory Bateson, Steps to an Ecology of
Mind, 177193. New York: Ballantine Books.
Bauman, Richard
2004
A World of Others Words: Cross-Cultural Perspectives on Intertextuality. Oxford/Malden, MA: Blackwell.
Bauman, Richard and Joel Sherzer (eds.)
[1974] 1989 Explorations in the Ethnography of Speaking, 2nd edition. Cambridge:
Benveniste, Emile
1966
Problmes de linguistique gnrale, I. Paris: Gallimard.
Blommaert, Jan
2005
Discourse: A Critical Introduction. Cambridge: Cambridge University Press.
Bolinger, Dwight
1968
Aspects of Language. New York: Harcourt Brace Jovanovich.
Bourdieu, Pierre
1980
Le sens pratique. Paris: Minuit.
[1978] 1987 Politeness: Some Universals in Language Usage. Cambridge: Cambridge
University Press.
2007
Metapragmatics in Use. (Pragmatics and Beyond, n. s. 165.) Amsterdam: Benjamins.
Bhler, Karl
[1934] 1990 Theory of Language: The Representational Function of Language. Amsterdam: Benjamins. [Sprachtheorie: Die Darstellungsfunktion der Sprache.
Jena: Fischer.]
Butler, Judith
1990
Gender Trouble: Feminism and the Subversion of Identity. London: Routledge.
Caffi, Claudia (ed.)
1984
Special Issue on Metapragmatics. Journal of Pragmatics 8(4).
Cameron, Deborah and Don Kulick
2003
Language and Sexuality. Cambridge: Cambridge University Press.
161
Carnap, Rudolf
[1934] 1937 The Logical Syntax of Language. London: Kegan Paul. [Logische Syntax
der Sprache. Vienna: Springer.]
Cassirer, Ernst A.
1945
Structuralism in modern linguistics. Word 1: 99120.
Caton, Steven C.
1987
Contributions of Roman Jakobson. Annual Review of Anthropology 16: 223
260.
Caton, Steven C.
1990
Peaks of Yemen I Summon: Poetry as Cultural Practice in a North Yemeni
Tribe. Berkeley/Los Angeles/London: University of California Press.
Chapman, Siobhan
2005
Paul Grice, Philosopher and Linguist. Basingstoke/New York: Palgrave Macmillan.
Coffa, J. Alberto
1991
The Semantic Tradition from Kant to Carnap: To the Vienna Station. Cambridge: Cambridge University Press.
Comrie, Bernard
[1981] 1989 Language Universals and Linguistic Typology: Syntax and Morphology,
2nd edition. Chicago: University of Chicago Press.
Coupland, Nikolas, Srikant Sarangi and Christopher N. Candlin (eds.)
2001
Sociolinguistics and Social Theory. Harlow: Pearson Education.
Davis, Steven (ed.)
1991
Pragmatics: A Reader. Oxford/New York: Oxford University Press.
Donnellan, Keith S.
1966
Reference and definite descriptions. Philosophical Review 75: 281304.
Donnellan, Keith S.
1978
Speaker reference, descriptions and anaphora. In: Peter Cole (ed.), Syntax and
Semantics, Volume 9: Pragmatics, 4768. New York: Academic Press.
Du Bois, John W.
1987
Meaning without intention: Lessons from divination. IPrA Papers in Pragmatics 1: 80122.
Ducrot, Oswald
1984
Langage, mtalangage et performatifs. In: Le dire et le dit, 117148. Paris:
Minuit.
Duranti, Alessandro
1993
Intentions, self, and responsibility. In: Jane H. Hill and Judith T. Irvine (eds.),
Responsibility and Evidence in Oral Discourse, 2447. Cambridge: Cambridge University Press.
1992
Rethinking Context: Language as an Interactive Phenomenon. Cambridge:
Fillmore, Charles J.
1975
Santa Cruz Lectures on Deixis, 1971. Bloomington, IN: Indiana University
Linguistics Club.
Foucault, Michel
1966
Les mots et les choses: une archologie des sciences humaines. Paris: Gaillmard.
162
Wataru Koyama
Friedman, Michael
2000
A Parting of the Ways: Carnap, Cassirer, and Heidegger. Chicago/La Salle, IL:
Open Court.
Gazdar, Gerald
1979
Pragmatics: Implicature, Presupposition, and Logical Form. New York: Academic Press.
Giddens, Anthony
[1976] 1993 New Rules of Sociological Method: A Positive Critique of Interpretative
Sociologies, 2nd edition. Stanford: Stanford University Press.
Goffman, Erving
1967
Interaction Ritual: Essays on Face-to-Face Behavior. New York: Anchor
Books.
Goffman, Erving
1974
Frame Analysis: An Essay on the Organization of Experience. New York:
Harper and Row.
Goffman, Erving
1981
Forms of Talk. Philadelphia: University of Pennsylvania Press.
Goffman, Erving
1983
Felicitys condition. American Journal of Sociology 1: 153.
Grice, H. Paul
1967
Logic and conversation. Unpublished manuscript of the William James Lectures, Harvard University. Published in Studied in the Way of Words. Cambridge, MA: Harvard University Press [1989].
Gumperz, John J.
1982
Discourse Strategies. Cambridge: Cambridge University Press.
Gumperz, John J. and Dell Hymes (eds.)
[1972] 1986 Directions in Sociolinguistics: The Ethnography of Communication. Oxford: Basil Blackwell.
Haberland, Hartmut and Jacob L. Mey
1977
Editorial: Pragmatics and linguistics. Journal of Pragmatics 1(1): 116.
Halliday, Michael A. K. and Ruqaiya Hasan
1976
Hancher, Michael
1979
The classification of cooperative illocutionary acts. Language in Society 8:
114.
Hanks, William F.
1990
Referential Practice: Language and Lived Space among the Maya. Chicago:
Harris, Randy Allen
1993
The Linguistic Wars. Oxford: Oxford University Press.
Heritage, John
1984
Holenstein, Elmar
1977
Jakobsons contribution to phenomenology. In: Daniel Armstrong and Cornelis H. van Schooneveld (eds.), Roman Jakobson: Echoes of his Scholarship,
145162. Lisse: The Peter de Ridder Press.
163
Huang, Yan
2007
Pragmatics. Oxford/New York: Oxford University Press.
Irvine, Judith T.
1996
Shadow conversations: The indeterminacy of participant roles. In: Michael
Silverstein and Greg Urban (eds.), Natural Histories of Discourse, 131159.
Chicago: University of Chicago Press.
Jakobson, Roman
1957
Shifters, verbal categories, and the Russian verb. Cambridge, MA: Harvard
University Russian Language Project.
Jakobson, Roman
1960
Closing statement: Linguistics and poetics. In: Thomas A. Sebeok (ed.), Style
in Language, 350377. Cambridge, MA: MIT Press.
Jakobson, Roman
1973
Main Trends in the Science of Language. London: George Allen and Unwin.
Jaworski, Adam, Nikolas Coupland and Galasinski Dariusz (eds.)
2004
Metalanguage: Social and Ideological Perspectives. Berlin: Mouton de
Gruyter.
Jespersen, Otto
1922
Kaplan, David
1978
Dthat. In: Peter Cole (ed.), Syntax and Semantics, Volume 9: Pragmatics,
221243. New York: Academic Press.
Karttunen, Lauri
1974
Presupposition and linguistic context. Theoretical Linguistics 1: 181194.
Koyama, Wataru
2000
How to be a singular scientist of words, and worlds, and other (possibly) wonderful things. Journal of Pragmatics 32: 651686.
Kripke, Saul
1972
Naming and Necessity. Cambridge, MA: Harvard University Press.
Kripke, Saul
1979
Speakers reference and semantic reference. In: Peter A. French, Theodore E.
Uehling Jr. and Howard K. Wettstein (eds.), Contemporary Perspectives in the
Philosophy of Language, 627. Minneapolis, MN: University of Minnesota
Press.
Kuklick, Henrika
1991
The Savage Within: The Social History of British Anthropology, 18851945.
Cambridge: Cambridge University Press.
Lakoff, Robin T.
1973
Language and womans place. Language in Society 2(1): 4580.
Lakoff, Robin T.
1989
The way we were. Journal of Pragmatics 13: 939988.
Leech, Geoffrey N.
1983
1983
Locke, John
[1689] 1975 Essay on Human Understanding. Edited by Peter H. Nidditch. Oxford:
Oxford University Press.
164
Wataru Koyama
Lucy, John A.
1992
Language Diversity and Thought: A Reformulation of the Linguistic Relativity
Hypothesis. Cambridge: Cambridge University Press.
Lucy, John A. (ed.)
1993
Reflexive Language: Reported Speech and Metapragmatics. Cambridge University Press: Cambridge.
Lyons, John
1977
Semantics (2 volumes). Cambridge: Cambridge University Press.
Merleau-Ponty, Maurice
1945
Phnomnologie de la perception. Paris: Gallimard.
1960
Signes. Paris: Gallimard.
Mey, Jacob L.
1979
Zur kritischen Sprachtheorie. In: Jacob L. Mey (ed.), Pragmalinguistics:
Theory and Practice, 411434. The Hague: Mouton.
Mey, Jacob L.
1985
Whose Language?: A Study in Linguistic Pragmatics. Amsterdam: Benjamins.
Mey, Jacob L.
[1993] 2001 Pragmatics: An Introduction, 2nd edition. Oxford/Malden, MA: Blackwell.
Morris, Charles W.
1938
Foundations of the theory of signs. In: Otto Neurath, Rudolf Carnap, and
Charles W. Morris (eds.), International Encyclopedia of Unified Science,
Volume 1, No. 2, 77137. Chicago: University of Chicago Press.
Mulligan, Kevin (ed.)
1987
Speech Act and Sachverhalt: Reinach and the Foundations of Realist Phenomenology. Dordrecht: Martinus Nijhoff.
Nerlich, Brigitte
1992
Semantic Theories in Europe: From Etymology to Contextuality. Amsterdam:
Benjamins.
Nerlich, Brigitte and David D. Clarke
1996
Language, Action, and Context: The Early History of Pragmatics in Europe
and America, 17801930. Amsterdam: Benjamins.
Peirce, Charles S.
[18931910] 1955 Logic as semiotic: The theory of signs. In: Justus Buchler (ed.),
Philosophical Writings of Peirce, 98119. New York: Dover.
Pike, Kenneth Lee
1967
Language in Relation to a Unified Theory of the Structure of Human Behavior,
2nd, rev. edition. The Hague: Mouton.
Pressman, Jon F.
1994
Pragmatics in the late twentieth century: Countering recent historiographic
neglect. Pragmatics 4(4): 461489.
Putnam, Hilary
1975
Philosophical Papers, Volume 2: Mind, Language, and Reality. Cambridge:
Recanati, Franois
1979
La transparence et lnonciation: pour introduire la pragmatique. Paris:
Seuil.
165
Richardson, Alan
1998
Carnaps Construction of the World: The Aufbau and the Emergence of Logical Empiricism. Cambridge: Cambridge University Press.
Ricur, Paul
[1975] 1977 The Rule of Metaphor: Multi-Disciplinary Studies of the Creation of
Meaning in Language. Toronto: University of Toronto Press.
Rorty, Richard M. (ed.)
1967
The Linguistic Turn: Recent Essays in Philosophical Method. Chicago: University of Chicago Press.
1982
Ruesch, Jrgen and Gregory Bateson
1951
Communication: The Social Matrix of Psychiatry. New York: Norton.
Sadock, Jerrold M.
1974
Toward a Linguistic Theory of Speech Acts. New York: Academic Press.
Sapir, Edward
1933
La ralit psychologique des phonmes. Journal de Psychologie Normale et
Pathologique 30: 247265.
Schieffelin, Bambi B., Kathryn Woolard and Paul V. Kroskrity (eds.)
1998
Language Ideologies: Practice and Theory. New York/Oxford: Oxford University Press. [Originally, 1992, Special Issue on Language Ideologies,
Pragmatics 2(3).]
Searle, John R.
1969
Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge
University Press.
Searle, John R.
1983
Intentionality: An Essay in the Philosophy of Mind. Cambridge: Cambridge
University Press.
1976a
Hierarchy of features and ergativity. In: Robert M. W. Dixon (ed.), Grammatical Categories in Australian Languages, 112171. Canberra: Australian Institute of Aboriginal Studies.
1976b
Shifters, linguistic categories, and cultural description. In: Keith H. Basso and
Henry A. Selby (eds.), Meaning in Anthropology, 1155. Albuquerque, NM:
University of New Mexico Press.
1993
Reflexive Language: Reported Speech and Metapragmatics, 3358. Cambridge: Cambridge University Press.
Silverstein, Michael and Greg Urban (eds.)
1996
Natural Histories of Discourse. Chicago: University of Chicago Press.
Stocking, Jr., George W. (ed.)
1996
Volksgeist as Method and Ethic: Essays on Boasian Ethnography and the
German Anthropological Tradition. (History of Anthropology, Volume 8.)
Madison, WI: University of Wisconsin Press.
166
Wataru Koyama
Tannen, Deborah (ed.)

1986
Framing in Discourse. Oxford/New York: Oxford University Press.
Thomas, Jenny
1996
Verschueren, Jef
1999
Volosinov, Valentin N.
[1929] 1973 Marxism and the Philosophy of Language. Cambridge, MA: Harvard University Press.
Williams, Glyn
1992
Sociolinguistics: A Sociological Critique. London: Routledge.
6.

Winfried Nth
1.
Semiotics as a general framework of pragmatics
Levinson (1983: 15) gives the standard account of how semiotics, the general
study of signs, is considered to be fundamental to linguistic pragmatics. The author of the Cambridge Textbook on Pragmatics begins with Charles W. Morris
(19011979), the founder of a general semiotics first outlined in 1938. Morriss
science of signs (1938: 12) is both broader and narrower than the scope of modern linguistic pragmatics.1 It is broader insofar as it is not restricted to the study of
verbal communication but deals with signs in all their forms and manifestations,
whether in animals or men, whether normal or pathological, whether linguistic
or nonlinguistic, whether personal or social (1964: 1), acoustic, visual, olfactory,
gustatory, or tactile. It is also broader since pragmatics, as conceived by Morris,
comprises branches of language studies which today fall into the domain of other
disciplines of linguistics, such as psycholinguistics, sociolinguistics, neurolinguistics, and much besides (Levinson 1983: 2). It is narrower insofar as Morris
aims at a science of signs on a biological basis and specifically within the framework of the science of behavior (1946: 80) indebted to the positivist paradigm of
psychological and social behaviorism. As Levinson (1983: 24) points out, the
scope of pragmatics became still narrower in the frameworks of logical positivism
and analytical philosophy of ordinary language: Carnap (1942) restricts pragmatics to the logical study of language use, for Bar-Hillel (1954) pragmatics is the
study of language use involving indexical words, and Montague (1970: 68) conceives pragmatics as a contextual theory of truth conditions founded on intensional
logic.
Sign behavior, according to Morris, involves three main factors: that which
acts as a sign [the sign vehicle], that which the sign refers to [the designatum], and
that effect on some interpreter in virtue of which the thing in question is a sign to
that interpreter [the interpretant] (1938: 3). Based on this triad, Morris (1938:
67) defines semiotics as a field of study of the following three domains corresponding to three well-known branches of modern linguistics: syntax (or syntactics), the study of the relation between sign vehicles, semantics, the study of the relations between sign vehicles and their designata, and pragmatics, the study of the
relation between sign vehicles and their interpreters (cf. Posner 1985).
20th century semiotics has developed in many directions, and some of the
schools of semiotic research have largely ignored the pragmatic dimension of sign
processes. This is especially true of structuralist semiotics in the tradition of Fer-
168
Winfried Nth
dinand de Saussure (18571913) and Louis Hjelmslev (18991965), whose focus

is entirely on language as a system and not on language use. It is no wonder, therefore, that Jacob L. Mey, once Hjelmslevs PhD student in Copenhagen, in his Introduction to Pragmatics, disregards semiotics entirely as a framework of linguistic pragmatics and that Dan Sperber and Deirdre Wilson, while reducing
semiotics to a theory of codes, dismiss its relevance to pragmatics altogether
(Sperber and Wilson 1986: 7, 2).
The structuralist and behaviorist heritage of semiotics has been an important
obstacle to a more fruitful interdisciplinary dialogue between general semiotics
and linguistic pragmatics. For a long time, it has impeded the spread of the seminal
ideas which the semiotic writings of Charles S. Peirce (18391914) offer to linguistics in general (Nth 2002a) and pragmatics in particular. Only relatively recently has the relevance of these ideas to pragmatics become acknowledged,2 but
so far mostly only by philosophers of language and hardly by linguists (e. g., Levinson 1983: 1, fn. 1).
When Morris conceived the triadic subdivision of semiotics into syntactics, semantics, and pragmatics, he was clearly influenced by Peirce. Peirce had also proposed a triadic framework for the study of signs, but the three branches of semiotics which Peirce set up describe a much wider field of study. Peirce was
inspired by the medieval canon of the three liberal arts, the trivium of grammar,
logic, and rhetoric.3 In contrast to the Saussurean tradition, in which linguistics is a
branch of semiology (cf. Nth 2000: 7273), Peirce considers linguistics and semiotics as two different branches of the system of sciences. Whereas linguistics, in
Peirces detailed classification of the sciences of 1902 (CP4 1.203283), is a
branch of the empirical or special sciences, semiotic(s), the quasi-necessary, or
formal, doctrine of signs, belongs to the normative and formal sciences (CP 2.227,
c. 1897). In contrast to the empirical sciences, which study what is in the actual
world, semiotics, as a normative science, aims at inquiring into what must be the
character of all signs and what would be true of signs in all cases (CP 2.227).
The normative approach is particularly evident in logic since to decide whether a
statement or conclusion is true or false is a normative decision. This is why, to
Peirce, semiotics was only another name for logic (CP 2.227, c. 1897).
The three branches of semiotics, in Peirces redefinition of the trivium of the
liberal arts, are the following: (1.) speculative (or formal) grammar, (2.) logic
(proper), and (3.) pure, formal or speculative rhetoric.5 Speculative grammar, the
precursor of Morriss syntactics, is the study of signs as such; it inquires into the
general conditions of signs being signs (CP 1.444, c. 1896), and it is a general
theory of the nature and meanings of signs, whether they be icons, indices, or symbols (CP 1.191, 1903). Of the branches of modern linguistics, not only syntax, but
also phonology and morphology belong to this branch of semiotics. Logic proper,
the second branch of semiotics, which Peirce also calls critic, is the precursor of
Morriss semantics and certainly also comprises linguistic semantics. According to
169
Peirce, it deals with the relations of signs to the objects that they represent, being
the science of what is quasi-necessarily true of signs [] in order that they []
may be true; [ it] is the formal science of the conditions of the truth of representations (CP 2.229, c. 1897).
Speculative rhetoric, the precursor of Morriss pragmatics and with it of linguistic pragmatics, deals with the relations of signs to their producers and interpreters; it studies the effects created by signs in the participants of the process of
semiosis. Speculative rhetoric is the study of the necessary conditions of the
transmission of meaning by signs from mind to mind, and from one state of mind to
another (CP 1.444, c. 1896); it deals with the formal conditions of the force of
symbols, or their power of appealing to a mind (CP 8.342, 1908). Speculative
rhetoric is broader than Morriss pragmatics in two respects. First, to study how
signs are transmitted from one state of mind to another means that this branch of
semiotics does not only study signs in communication but also in thought not communicated to any other interpreter. Thinking has pragmatic dimensions since it always proceeds in the form of a dialogue a dialogue between different phases of
the ego (CP 4.6, 1906). Second, speculative rhetoric is also conceived as a rhetoric of fine art, a rhetoric of practical persuasion, and a rhetoric of science (EP
26: 329, 1904). Peirce had the vision that the rhetoric of science, which he also
called methodeutics (Santaella 1999), was destined to grow into a colossal doctrine which may be expected to lead to most important philosophical conclusions
(CP 3.454, 1896).
2.
Semiotic foundations of communication
Linguistic pragmatics studies how language is used in communication, especially,

how verbal messages are exchanged between speakers and hearers in dialogue.
Merely incidentally transmitted utterances, monologues, and other forms of language use without an addressee are of no concern to the field of study (cf. Levinson
1983: 16).
Communication is a central issue of semiotics; its study is also called semiotics
of communication (Santaella and Nth 2004). The general study of signs and sign
processes may thus provide a theoretical framework for linguistic pragmatics.
Among its topics of relevance to linguistic pragmatics are: (1) models and functions of communication, (2) the instrumental view of language in communication,
(3) the problem of intentionality, (4) the nature of cooperation in conversation, and
(5) the dialogical nature of the sign and its relevance to the study of communication.
170
2.1.
Winfried Nth
Models and functions of communication
Most theories of pragmatics take a very rudimentary model of communication as

the point of departure of their analyses of speech acts. With reference to Grice
(1957), Levinson considers the triad S-U-R, sender (speaker, writer), utterance,
and receiver (hearer, reader), to be fundamental: Communication consists of the
sender intending to cause the receiver to think or do something, just by getting
the receiver to recognize that the sender is trying to cause that thought or action (Levinson 1983: 16). The utterance, in this triad, is a linguistic token; U has
meaning, which becomes mutual knowledge to S and R in the process of communication, but initially, U may have two kinds of meaning, possibly in conflict, a
speaker meaning and a sentence meaning (Levinson 1983: 16).
Leech, in his Principles of Pragmatics, proposes a flow diagram of communication according to which an S uttering U to H is engaged in three kinds of semiotic
transaction: (1) an interpersonal transaction, or discourse, (2) an ideational transaction or message transmission, and (3) a textual transaction or text (Leech 1983:
59). As to its three constituents, S-U-H, and its notion of message transmission, this
model is certainly inspired by the technical model of the flow of information from
S to H by Shannon and Weaver (1949). This model is inadequate for the study of
human communication, since it is linear, unidirectional, and assigns a merely passive role to the recipient of the message, making the sender the only active agent in
communication. It disregards Hs autonomy in the process of interpreting Ss message as well as the important elements of circularity by which S and H are connected
in communication through feedback loops by which H contributes to the production
of Ss message from the moment of its articulation on (cf. Nth 2000: 24446).
A much cited communication model in linguistic pragmatics is the semiotic
model of the six factors and functions of communication which Roman Jakobson
proposed in extension of Bhlers triadic organon model of language (Bhler
1933). Jakobson defines the factors involved in a process of communication as follows: An addresser sends a message to an addressee. This message refers to a context (the referent of the verbal message) seizable by the addressee. Addresser
and addressee have a common code, such as the vocabulary and grammar of the
language in which they communicate, and finally, there is a contact which serves as
a physical channel and psychological connection between the addresser and the
addressee, enabling both of them to enter and stay in communication (Jakobson
1960: 353).
Each of these six factors is the determinant of a specific communicative function; one and the same message may have several functions, but one is usually predominant. Messages which focus on the referent (context), such as informative, descriptive, or narrative texts, have a predominantly referential function. The
emotive or expressive function is predominant when the focus is on the speakers
attitude in relation to the meaning of the message. The use of interjections, a tremb-
171
ling voice, or the frequent use of the pronouns I or me are examples of signs of an
utterance in which the emotive function of language is predominant. The conative
function is oriented toward the addressee; its most typical grammatical expressions
are the vocative and the imperative. The phatic function predominates in messages
which, without necessarily conveying any particular meaning at all, serve to establish, to prolong, or to discontinue communication, to check whether the channel
works (Hello, do you hear me?), to attract the attention of the interlocutor or to
confirm his continued attention (Are you listening?) (1960: 355). The metalinguistic function predominates when language is used to refer to language and verbal communication; metalanguage is language about language, e.g., as in grammar
books or in the discussion of questions of terminology. Finally, the poetic function,
according to Jakobson, predominates when the focus of verbal communication is
on the qualities of language irrespective of its message, e.g., on the sound pattern,
rhythm, or meter of a text.
Although Peirce considers the model of the sign to be more fundamental than
the model of communication, it is not true that the founder of modern semiotics had
little to say about the topic of communication, as Habermas (1995) believes. Elements of a Peircean theory of communication have meanwhile been reconstructed.7 In his manuscript on Pragmatism of 1907, Peirce defines communication
of a speaker with a hearer as follows: Signs mostly function between two minds,
or theatres of consciousness, of which one is the agent that utters the sign (whether
acoustically, optically, or otherwise) while the other is the patient mind that interprets the sign (EP 2: 403). In this description, the characterization of communication as a theater, a staged scenario of acting agents, is more than a mere metaphor. Actors on stage do not act autonomously, but they represent actions and act
on behalf of other agents. But who should be the agents in communication if not S
and H? This question is the topic of sections 2.2 and 2.3.
2.2.
Are verbal signs instruments of those who communicate?
Most approaches to linguistic pragmatics are based on the assumption that verbal
messages are instruments by which senders seek to influence receivers. Language
use is regarded as a means to some end. Searle (1969: 48), e.g., writes: The sentence uttered by a speaker who knows its meaning then provides a conventional
means of achieving the intention to produce certain illocutionary effects in the
hearer. Leech (1983: 36), too, postulates that speaking is a means to an end and
hence an instrument for a purpose: spoken or written messages are transferred or
transmitted from a speaker to a hearer, and the speakers purpose in this transaction is to convey a particular illocutionary force to the hearer (Leech 1983:
5860).
The view that verbal signs are used as instruments has been defended since antiquity. It is central to one of the earliest treatises on the philosophy of language,
172
Winfried Nth
Platos Cratylus. In this dialogue on the relation between words and the objects for
which they stand, Socrates compares the practical utility of the craftsmans tools
with the utility which words have for a speaker. A tool, says Socrates, is naturally
adequate and appropriate to the task for which it serves. To the weaver, the right
and natural (391E) instrument is the shuttle, to the blacksmith, the anvil and the
hammer serve best to shape the iron, and to the shoemaker, the awl is the natural
tool for piercing leather. As any practical action requires its proper tool, so does the
act of speaking. The speakers appropriate instruments for communicating their
ideas are their words (388a-389c). Practical instruments have a practical utility,
whereas words serve a semiotic utility. As the weavers shuttle is useful for separating the warp from the woof, so is the word useful to the speaker who wants to
communicate or to organize his or her ideas.
In the history of semiotics, the instrumental theory of the sign has lived on in
many variants (cf. Nth 2009). In the context of pragmatics, Wittgensteins (1953)
revival of the topic is of special interest. Instrument, or tool, are appropriate
characterizations of verbal signs in Wittgensteins philosophy of language since
these terms illustrate that the proper study of verbal signs is the study of their use.
What language and instruments have in common is that they are both used and
have functions: Think of the tools in a box: there is the hammer, pliers, a saw,
screwdriver, a glue-pot, nails and screws. The functions of words are as diverse
as the functions of these objects ( 11). Whereas Socrates focuses on the utility of
verbal signs in communication, Wittgensteins focus is on their meaning and the
way meaning is revealed in the use of signs. Not only is language in general an instrument, but also its concepts are instruments ( 569). Instruments as well as
language have their sense in their employment: Look at a sentence as an instrument and its sense as its employment ( 421).
Although the view that language serves as an instrument in communication is a
proverbial commonplace, Peirce contests it. A tool is only a means, i.e., an efficient
cause, but not an end, i.e., not a final cause, in the craftsmans process of producing
an effect. Words used in communication, by contrast, are more than mere tools
since they have purposes of their own, which is not their speakers purpose. What a
word means is not determined by its user. The truth conditions of a proposition are
not determined by the utterer of the sentence which it contains either; they are determined by the reality to which it refers and by the consequences the utterance
has. Whether an argument is valid or not is determined by the logic of language and
not by the person who argues, and whether a speech act is felicitous or not depends
on whether it fulfills certain felicity conditions that cannot be ignored by the utterer. Hence, to use words means to be constrained by restrictions determined by
language and the consequences of its use.
Instead of saying that it is the speaker who produces the utterance in accordance with the rules of syntax, semantics, and pragmatics, we can also say that these
rules determine the speaker to speak the way he or she speaks, but the pragmatist
173
account of language and meaning goes a step further. According to Peirces pragmatic maxim (see 3.1), it is not the rules of language and logic that determine our
utterances, but the anticipated feedback which the success or failure of our utterances gives us whenever we speak. The anticipation of the effects of our words
makes us utter them in this or that way. Nor can it be argued that the utterers
thoughts in which his or her intention to speak is conceived use the ensuing utterance as instruments to put the speakers intentions into effect. Utterances which express thoughts are interpretants of these thoughts, which make the interpretants the
final causes of the thoughts. Since the utterance is itself a sign with the purpose of
creating further interpretants in the hearers mind it cannot be merely instrumentally (or efficiently) caused by the thoughts in which they were conceived.
Nothing could be less in accordance with Peirces semiotic premises than a
dualist assumption of internal thoughts becoming merely externalized as spoken or
written words in communication. Sperber and Wilson (1986: 1), for example, make
this assumption when they reflect on the relationship between their own thoughts
and their expression in writing as follows: what we put down on paper are little
dark marks []. As for our thoughts, they remain where they always were, inside
our brains. Peirce, by contrast, argues that signs exert their semiotic agency inside
as well as outside of human brains, and perhaps even more so outside than inside.
He underlines the latter view by saying that it is even much more true that the
thoughts of a living writer are in any printed copy of his book than that they are in
his brain (CP 7.364, 1902). Hence, thought-signs do not remain contained in the
container of a human brain just as written words do not remain fixed in their meaning once they are written down. Words are symbols which live and grow outside of
human brains in the semiotic effects which they have, and this growth takes place
both in thought and in communication. What they have in common is their dialogical nature:
Before the sign was uttered, it already was virtually present to the consciousness of the
utterer, in the form of thought. But [] a thought is itself a sign, and should itself have
an utterer (namely, the ego of a previous moment) to whose consciousness it must have
been already virtually present, and so back. Likewise, after a sign has been interpreted,
it will virtually remain in the consciousness of its interpreter, where it will be a sign []
and, as a sign should, in its turn, have an interpreter, and so on forward. (EP 2: 403,
1907; see below, 2.4 for the argument of the infinite regress.)
In sum, if signs have purposes, or final causes, and exert a semiotic agency of their
own, they cannot be the mere instruments, that is, instrumental (or efficient)
causes, of other signs; and since both thoughts and utterances are signs, utterances
cannot be the instruments of the thoughts in which they were first conceived.
Although the sign is not an instrument, instruments are necessary for the dissemination of signs. We need a pen and paper to write down words, and our voices
are necessary instruments by which our words reach their hearers, but these words
themselves are not our instruments. Not words are our semiotic instruments, but
174
Winfried Nth
our voices, tongues and lips, pens, paper, typewriters, telephones, or computers
are. The instruments by which thought-signs exert their semiotic agency are the
neurons, the brain lobes, and the perceptual organs, eyes and ears, by means
of which thoughts are processed. This is probably Peirces most radical thesis:
our brain is not the only cause of our utterances; it is merely the efficient cause
of the thoughts and of their expression. The ideas on which this approach to semiosis is based have parallels in current quasi-biological positions of the philosophy
of mind known as memetics and teleosemantics; the biolinguistic assumption is
here that words in culture spread like genes spread in biological species (cf. Nth
2009).
In order to deconstruct the autonomy of the speaking subject, Peirce, for the
sake of the argument, may exaggerate here and there when he speaks of the purpose of the sign which is not the purpose of the user. Elsewhere, he acknowledges
two purposes, the purpose of the sign and the purpose of the sign user, in a mutual
feedback control circuit. In an evolutionary perspective, he describes the loop in
which sign users using signs and signs using sign users interact in terms of a process of co-evolution:
Man makes the word, and the word means nothing which the man has not made it mean,
and that only to some man. But since man can think only by means of words or other external symbols, these might turn round and say: You mean nothing which we have not
taught you, and then only so far as you address some word as the interpretant of your
thought. In fact, therefore, men and words reciprocally educate each other. (CP 5.313,
1868)
There is a certain affinity between this Peircean account of the relationship between the signs and their users on the one hand and Saussures theories of the immutability of the verbal sign and the determination of each signifier by the system
of which it is part on the other. As Stetter (1989: 159) has pointed out in his Saussurean interpretation of the above quote: we are not the masters of our signs who
determine the interpretation of terms but it is rather determined by presuppositions
inherent in the system of language, but Peirces object of the sign is of course
more than the system of which the sign is a part.
The co-evolution of signs and their users can only take place if purposes are ascribed to both. But this does not mean that the sign user is the autonomous agent
external to the signs as which the intentionalists describe him (see 2.3). Peirce
solves the apparent paradox of the conflict between the purposes of the sign user
and the one of the sign with the argument that sign users are themselves objects of
signs (see 2.5) so that their intentions are, in turn, purposes of signs. By no means
do sign users pursue nonsemiotic purposes, for they themselves are and act as signs
in the flow of thought-signs and public signs which constitute them as signs. This is
the line of argument of the famous passage in which Peirce defines the human
being as a sign:
175
There is no element whatever of mans consciousness which has not something corresponding to it in the word; and the reason is obvious. It is that the word or sign which
man uses is the man himself. For, as the fact that every thought is a sign, taken in conjunction with the fact that life is a train of thought, proves that man is a sign; so, that
every thought is an external sign, proves that man is an external sign. That is to say, the
man and the external sign are identical, in the same sense in which the words homo and
man are identical. Thus my language is the sum total of myself; for the man is the
thought. (CP 2.314, 1868)
If the sign user is a sign, his or her intentions are signs, too, and in the speakers utterances the underlying intentions have to be interpreted as objects of signs. As the
object determines the sign so does the sign determine its interpretants. Words only
stand for the objects they do, and signify the qualities they do, because they will
determine, in the mind of the auditor, corresponding signs (CP 2.92, 1902). As
any object of a sign determines the sign which it represents so do the speakers intentions, since they are objects of signs. This is how the speakers intention participates in the creation of an interpretant.
2.3.
Intentionality of communication and the purpose of verbal signs
The classics of linguistic pragmatics agree in considering intentionality as a prerequisite of verbal communication. Levinson (1983: 16) summarizes as follows:
Communication consists of the sender intending to cause the receiver to think
or do something, just by getting the receiver to recognize that the sender is trying to cause that thought or action. So communication is a complex kind of intention that is achieved or satisfied just by being recognized.
Searle (1983: 16466) postulates two intentions. First, there must be an intention to represent on the speakers side, but since it is possible to represent without
the purpose to address or to influence anybody, a second intention is required: the
intention to communicate. Searle (1983: 166) explains that it is impossible to inform a hearer that it is raining without intending that this message represents
truly or falsely the weather condition expressed. Not only the speakers intention
but also the hearers awareness of this intention is a prerequisite of communication. Furthermore, the belief produced in the hearers mind as the result of the
message must have its cause in the speakers intention (cf. Sperber and Wilson
1986: 23).
In structural semiotics, intentionality is the criterion by which communication
is distinguished from signification. The distinction has its roots in the writings of
Buyssens (1943).8 All signs which humans can interpret, including the signs of
physical nature, have signification, but not all of them are intended; only those
signs constitute acts of communication which have an addresser whose intention it
is to convey a message to an addressee. In the tradition of analytical language philosophy, Grice (1957) interprets the difference in question in terms of the dichot-
176
Winfried Nth
omy of natural (i.e., unintentional) and nonnatural (i.e., intended) meanings.

The latter term is unfortunate, though. To call the signs of natural language nonnatural, as Grice does, fails to do justice to the continuity of semiosis in the evolution of life from nature to culture (cf. Nth 2008).
Prieto (1966: 3338) anticipates Searles theory of the two communicative intentions when he distinguishes between two indications which signs have in communication, a notificative and a significative indication. Significative indication is
the one by which a sign indicates whatever it means (i.e., its signification). Notificative indication, by contrast, indicates that the sign is being communicated with
the intention of a sender. The content of a notificative indication is so to speak: Attention! This sign is intended to convey a message to you! (cf. Hervey 1982: 71).
If signs can only be intended or not, intentionality cannot be a matter of degree.
However, intentions are volitions, and wishes and desires may vary in strength.
Addressers may be divided as to their intentions, they may have doubts and hesitate whether they should communicate or not, or they may merely pretend to communicate and act as if they communicated (cf. Parret 1994). Such factors of intentionality are rarely taken into consideration in linguistic pragmatics.
Psychoanalysis, by contrast, has shown how verbal and nonverbal messages
may also communicate hidden intentions which the utterer may want to conceal
but fails to do, e.g., when he or she, impatient with the hearer, loses control over his
or her voice, sounding angry despite the intention to appear calm. In such a scenario, the speakers loss of control over his or her voice is the sign of a conflict between two intentions, one determined by the unconscious mind, the other by the
conscious self. The conflict of intentions is thus the symptom of a divided self in
which a conscious mind aims at concealing what the unconscious mind intends to
express.
Linguistic pragmatics also disregards what the communication theorists
Watzlawick, Beavin, and Jackson (1967) say about the ubiquity of communication
in everyday life. In their study Pragmatics of Human Communication, the authors
postulate the metacommunicative axiom that it is impossible not to communicate,
be it verbally or nonverbally: Behavior has no opposite. There is no such thing as
nonbehavior []. One cannot not behave, and any form of behavior in social interaction is of semiotic relevance since one cannot not communicate (Watzlawick et al. 1967: 4849). Even silence and doing nothing are messages.
Peirces semiotic theory of intentionality (cf. Kappner 2004) and his theory of
the purpose of signs in sign processes are rather incompatible with the views of intentionality held in mainstream linguistic pragmatics.9 Peirce defines the sign as a
medium (MS10 339: 283r, 1906), but not as one which mediates between an addresser and an addressee; instead, the sign logically mediates between its object,
which it represents, and its interpretant, which is the interpretative effect it creates.
The object of the sign is the verbal or nonverbal knowledge, feeling, behavior, or
habit which the sign user associates with the sign. The knowledge of its object is a
177
prerequisite for the interpretation of the sign. While the object of the sign includes
aspects of what linguists define as meaning (see 3.3), the interpretant includes the
illocutionary force and the perlocutionary effect of a speech act, as defined by Austin (1962). In Peirces definition, the interpretant is the cognitive, emotional, or behavioral effect which results from the interpretation of the sign.
According to Peirce, the sign is a semiotic agent with its own intention (cf. 2.2).
Signs continue to exert their agency in the absence of those who first produced
them. This topic is also discussed by Derrida (1988: 8), who uses the metaphor of
the machine for the agency which the written sign exerts irrespective of its producer: To write is to produce a mark that will constitute a sort of machine which is
productive in turn, and which my future disappearance will not, in principle, hinder
in its functioning, offering things and itself to be read and to be rewritten. Peirces
view of the agency of the sign goes beyond the mere assumption of its interpretability in the absence of its writer. Signs have a life of their own, which can be
studied in processes of language change and in the way texts are interpreted in new
ways as time passes and signs change. In 1904, Peirce writes about the purpose of
the symbol: The symbol, by the very definition of it, has an interpretant in view.
Its very meaning is intended. Indeed, a purpose is precisely the interpretant of a
symbol (EP 2: 308).
With purpose, Peirce thus does not mean the sign users intention, but the
signs intention to represent its object and to create and interpretant, i.e., to be interpreted in another sign (MS 1476, 1904). Purpose is thus a semiotic teleology
inherent in the sign. Not only uttered or written signs have purposes, but also
thought-signs. Their purpose is to act in a mental dialogue in which one thoughtsign is translated or interpreted in a subsequent one (CP 5.284, 1868).
On the one hand, the sign represents its object to the mind of an interpreter, on
the other, the object represented by the sign is in a sense the cause, or determinant,
of the sign even if the sign represents its object falsely (CP 6.347, 1909; cf. Parmentier 1985). To justify the autonomy of the sign by its being determined sounds
contradictory, but it describes the process of an autopoietic self-determination of
the sign. Since the object of the sign is not external, but internal to the sign, to say
that the sign is determined by its object means that it is determined by a dynamics
inherent in itself and not by an agency outside of the sign (cf. Colapietro 1995).
The way in which a sign is determined by its object depends on its modes of relation to its object and its interpretants. Rhematic signs, such as words in isolation,
are determined to represent their objects as such, without affirming, questioning, or
negating their existence. Dicentic signs, such as affirmations, are determined by
the truth of what they represent, and arguments are determined by the laws of logic
and reasoning in general. Symbols are determined by the habits of their users, by
which they have become symbolic signs. Indices are determined by their objects
because they are related to them by causal, temporal, or spatial relations, whereas
icons represent their objects by means of qualities of their own. All these factors
178
Winfried Nth
determine the sign independently of their use. For example, a rhematic symbolic
legisign such as the word apple, uttered by a speaker, is determined by the habits of
the speakers of English to represent the fruit of the apple tree by means of this
word; the word represents its object as a general kind without affirming the existence of any particular fruit. The sign user may try to escape from his or her determination by the object of this sign and use this word to refer to a pear, but in the
long run, words cannot be used in this way because the speech acts will fail. A sign
user may also lie or argue against the laws of logic, but in the long run, to ignore the
truth and the laws of reasoning will not pay. Furthermore, verbal signs are determined by biolinguistic and evolutionary constraints, which have restricted the
speakers possibilities of articulation and perception (cf. Deacon, 1997: 116).
2.4.
Conversational cooperation, commonage, and the common ground
The model of communication underlying the classical theories of linguistic pragmatics is one of rational discourse.11 The aim of dialogic communication is guided
by the ideal of mutual consensus; the speakers purpose must be achieved. Communication is modeled as being cooperative; its goal is not only comprehension but
also the full acceptance of the intended message on the part of the hearer; speech
acts can only succeed or fail, and successful speech acts fulfill felicity conditions;
the speakers intentions are described as being sincere, and the messages are always relevant. Such accounts of verbal exchange between speakers and hearers
model communication according to the ideals of a desirable symmetry between the
agents interacting in dialogue. The underlying assumption is that a full congruence
between the speakers intended meaning and the hearers understanding of it is not
only possible but also desirable.
To reach consensus is indeed an ideal of dialogic interaction in rational discourse. The very concept of communication implies, in its root, the notion of making something common to those who communicate. Morris (1946: 195) expresses
this normative goal of communicative interaction in a definition according to
which communication is the use of signs to establish a commonage of signification. The assumption that commonage is an essential characteristic of communication is not undisputed. In everyday language, e.g., it is contradicted by the
military metaphors by which speakers represent their arguments as positions
which need be defended (Lakoff and Johnson 1980: 7). In the sociological theory
of communication, Luhmann (1984) starts from the basic assumption that the differences between the utterers and the interpreters frames of mind instead of their
common ground, is the source of all dialogic interaction, whereas Habermas
(1981) founds his Theory of Communicative Action on the assumption that consensus is constitutive of rational communicative interaction.
As a prerequisite for, and means of, reaching the normative ideal of dialogic
commonage, linguistic pragmatics recognizes such principles as Grices maxims
179
of conversational cooperation and its derivative, the principle of relevance advocated by Sperber and Wilson. A precursor in the study of these normative principles is Peirce (Pietarinen 2004, Short 2007: 21314). Although by no means restricted to the study of rational discourse (Oehler 1995: 26768), Peirces
speculative rhetoric describes conversational cooperation as follows:
Honest people, when not joking, intend to make the meaning of their words determinate,
so that there shall be no latitude of interpretation at all. That is to say, the character of
their meaning consists in the implications and non-implications of their words; and they
intend to fix what is implied and what is not implied. They believe that they succeed in
doing so, and if their chat is about the theory of numbers, perhaps they may. (CP 5.447,
1905)
However, immediately after this description of rational conversational cooperation, Peirce also points out that commonage can only serve as a regulative idea, an
ideal which can rarely be reached in ordinary conversation. The sentence after the
above passage goes on to say: But the further their topics are from such presciss12,
or abstract, subjects, the less possibility is there of such precision of speech. Insofar as the implication is not determinate, it is usually left vague (CP 5.447,
1905). Furthermore, Peirce also argues that, despite all efforts of cooperation, dialogues also evince a fundamental divergence of interest between the participants.
From this perspective, Peirce calls the participants of dialogue opponents (MS
515: 25, s. d.; cf. Hilpinen 1995: 293) and describes dialogic interaction in terms of
a metaphorical scenario of war: The utterer is essentially a defender of his own
proposition and wishes to interpret it so that it will be defensible. The interpreter,
not being so interested, and being unable to interpret it fully without considering to
what extreme it may reach, is relatively in a hostile attitude, and looks for the interpretation least defensible (MS 9: 34, c.1903).
Relevance, so much discussed in linguistic pragmatics since Grice, is another
principle about which Peirce has written (cf. Pietarinen 2004). Peirce discusses it
as follows: If the utterer says Fine day! he does not dream of any possibility of
the interpreters thinking of any mere desire for a fine day that a Finn at the North
Cape might have entertained on April 19, 1776. He means, of course, to refer to the
actual weather, then and there, where he and the interpreter have it near the surface
of their common consciousness (MS 318: 3233, 1907). Whereas Grice would
describe the elliptic utterance Fine day as one which requires filling the gap of
what remains unsaid by means of nonconventional conversational implicatures,
Peirce would consider most conversational implicatures unnecessary since communication, studied from the semiotic perspective, is not only verbal communication. What linguistic pragmatics describes as gaps between what is said and what
is left out may be a gap in the verbal message, but interpreters of a dialogue do not
only interpret words; they also interpret nonverbal signs and the situational circumstances of the utterance. What appears a gap in the verbal message may not be
a gap in the multi-channel message of the dialogic exchange. Furthermore, not
180
Winfried Nth
everything that is left open and undetermined in conversation needs to be filled by

implicatures either. It is wrong to assume that the speakers message is always
clearly determined and that the interpreters only task is to fill the gaps in order to
restore the intentions which seem unclear. Vagueness and the impossibility of resolving it may be the speakers purpose and the interpreters source of creative
thinking (see 3.3).
In Peirces speculative rhetoric, commonage is not only a normative goal; it is
also its prerequisite of the dialogue. As the prerequisite of communication, it belongs to the domain of the object of the sign; as its goal, it pertains to the interpretant created by the sign. The object of the sign can only be known if both the utterer
and the interpreter have collateral experience of the Real or Dynamical Object of a
sign. In order to know what a sign represents, we need to have previous acquaintance with what the sign denotes (CP 8.179, s. d.). The sign itself cannot express its
dynamical object, it can only indicate and leave the interpreter to find out by collateral experience what it represents (EP 2: 198, 1909). There is nothing nonconventional in this process of interpretation, as Grice assumes with his implicatures. Interpretation, according to Peirce is always a natural process of abductive,
inductive, and deductive inferencing.
An aspect of the object of the sign, i.e., that which must be presupposed for
communication to be successful, is the common ground of experience and knowledge which the utterer and the interpreter share. Both Grice and Searle have described it as a prerequisite of conversation. Pietarinen (2004: 302305) shows in
which respect Grices and Searles theories of the preconditions of successful communication were anticipated by Peirce. In 1908 Peirce formulates it as follows:
No man can communicate the smallest item of information to his brother-man unless they have [] common familiar knowledge [] such that each knower knows
that every other familiarly knows it, and familiarly knows that every other one of
the knowers has a familiar knowledge of all this (MS 614: 12, in Pietarinen
2006: 438). To make clear that no circulus vitiosus is involved in this endless reciprocity, Peirce underlined that this infinite regress is a logical and not a psychological one: Of course, two endless series of knowings are involved; but knowing
is not an action but a habit, which may remain passive for an indefinite time (MS
614: 12, in Pietarinen 2006: 438).
Insofar as commonage is the goal of communication, it pertains to the interpretant of the sign. Peirce calls the meaning effects common to the speaker and the
hearer the cominterpretant of the sign. At the same time, he recognizes that there
may be effects of meaning which are specific to the utterer and the interpreter, the
former being the intentional, the latter the effectual interpretant:
181
There is the Intentional Interpretant, which is a determination of the mind of the utterer;
the Effectual Interpretant, which is a determination of the mind of the interpreter; and
the Communicational Interpretant, or say the Cominterpretant, which is a determination
of that mind into which the minds of utterer and interpreter have to be fused in order that
any communication should take place. This mind may be called commens. It consists of
all that is, and must be, well understood between utterer and interpreter at the outset, in
order that the sign in question should fulfil its function. (EP 2: 478, 1906)
The commind, as Peirce calls the fusion of the two minds in one and the same
commens elsewhere, is a normative ideal. In practice, the ideal of a commind can
never be fully reached since mutual understanding is only possible in a fragmentary way. In 1907 (MS 318), Peirce points out that no interpreter can be said to have
access to the utterers mind. Knowledge about what a speaker means can only be
obtained in fragments, and such fragments are mere copies of a scrap torn out of
anothers life. By such scraps we can only supplement the ideas of [our own]
life (MS 318: 194, 1907; cf. Johansen 1993: 198202). As interpreters we must
match those fragments found in the signs of the utterance with our own discourse
universe and find out where they can be inserted or recopied in our own panorama of universal life (MS 318: 194, 1907). Interpretation is thus a semiotic
patchwork put together by abductive, inductive and deductive reasoning.
2.5.
Dialogue as the prototype of verbal communication, and the dialogical

nature of the sign
Peirce not only considered the dialogue as a wonderfully perfect kind of signfunctioning (EP 2: 391, c. 1906), he also had derived his first model of the sign together with the first version of his three categories, the I, the IT, and the THOU
from the three elementary constituents of a dialogue (W 113: xxvii, 174, 1865).14
Fisch explains how the model of the sign is derived from the model of the dialogue:
Peirce began where most of us begin, with a model, which, taken by itself, would
suggest too narrow a definition; namely the model of conversation between two
competent speakers of the same natural language say, English. [] It goes without saying that words are signs; and it goes almost without saying that phrases,
clauses, sentences, speeches, and extended conversations are signs (Fisch 1986:
357).
In Peirces fully developed semiotics, a sign, or representamen, is something
which stands to somebody for something in some respect or capacity. It addresses
somebody, that is, creates in the mind of that person an equivalent sign, or perhaps
a more developed sign. That sign which it creates I call the interpretant of the first
sign. The sign stands for something, its object (CP 2.228, c.1897). Sign, object,
and interpretant must be understood as following each other in time. The object
comes first insofar as it precedes the sign; it is that which must be known before the
sign can be understood. The sign comes next since it represents the object to an in-
182
Winfried Nth
terpretant, and the interpretant follows last since it is created by the sign. This temporal sequence makes clear how the three correlates of the sign correspond to the
three correlates of a dialogue, addresser (or sender), message, addressee (or receiver): the sign (or Representamen) is the message; it addresses the interpretant as
its receiver, and the sender of this message is its object, which determines the sign
(cf. Bergman 2003: 11, Pietarinen 2006: 426).
Despite the close association of the interpreter and the interpretant in the process of interpretation, the interpretant must not be confused with the mind of the
addressee of a message. In its full definition, the interpretant is any significant effect of the sign (CP 2.303, 1902), be it a feeling, action, or thought, and since significant effects of signs are not restricted to individuals and to the present moment
of sign interpretation, the interpretant can also be a memory, a habit, a fashion, an
innovation, a revolution, a war, or a peace treaty. In all cases, the interpretant is
itself a sign.
The object of the sign corresponds to its sender in an equally impersonal understanding (cf. 2.2 on the utterers intention as an object of the sign). Ransdell (1977)
explains why Peirce equates the two instances of conversation which linguists are
used to distinguishing as the referent of the sign on the one hand and the sender of a
message on the other: to the degree that the reality, the phenomena which we perceive, experience, or think about, are interpretable phenomena, they are objects of
signs. Reality is a sign producer since it is the cause of the signs which we have of
it; after all, we interpret reality by means of its signs, which in turn create interpretants. Both the sender of a message who, although not a creation of this message,
is nevertheless the source of its distribution, and the signs object precede the
moment and are the causes of the production of the sign.
Not only the sender but also the sign itself has an agency in the process of semiosis, as shown above (2.23). Peirce calls it the action of the sign (e.g.,
CP 5.47273, 1905), but again, this agency, which is due to the purpose of the sign,
is neither restricted to the purpose of a human being nor must it be the effect of a
verbal sign. There are even signs which have no sender at all. These are evidently
the natural signs (cf. Nth 2008). Since there can be no isolated sign, Peirce attributes a Quasi-utterer to (natural) signs without a real utterer (CP 4.551, 1906).
There are also signs without an interpreter but a Quasi-interpreter (CP 4.551,
1906), for example, if nobody takes the trouble to study the record (EP 2: 404,
1907). According to Peirces dialogic semiotics, it is essential that the utterer or
quasi-utterer as well as the interpreter or quasi-interpreter are not instances exterior to, but can be found within the sign, in which they must nevertheless be distinct although they are, so to say, welded in the sign (CP 4.551, 1906). A summary of the resulting simultaneity of the presence and absence of the utterer in the
dialogical sign is the following:
183
The action of a sign generally takes place between two parties, the utterer and the interpreter. They need not be persons; for a chameleon and many kinds of insects and even
plants make their livings by uttering signs, and lying signs at that. Who is the utterer of
signs of weather, which are not remarkably veracious, always? However, every sign certainly conveys something of the general nature of thought, if not from a mind, yet from
some repository of ideas, or significant forms, and if not to a person, yet to something
capable of some how catching on. (MS 318, S. 17, 1907)
In the dialogic process of conversation, the continuous alternation between the role
of the utterer and the one of the hearer involves a permanent transformation of
signs into interpretants and interpretants into the objects of new signs creating new
interpretants. This continuous process of alternations and transformations is a
source of creativity in semiosis. Since not only the utterers but also the objects of
the signs participate in the dialogue in the sense that they determine its interpretants, not only the voices of two participants but a plurality of voices is audible.
Dialogue is a Bakhtinian polyphonic chorus with a plurality of participants speaking the voices of many discourses (cf. Nth 2009).
3.
Pragmatism and Peirces semiotics of meaning and reference
Peirces semiotics is also relevant to issues of linguistic pragmatics such as the relationship between and the respective scopes of semantics and pragmatics, the difference between speaker, hearer, and utterance meaning, the nature of reference
and indexicality in language, and the theory of speech acts. Even the concept of
pragmatics has its origin in Peirces writings.
3.1.
Pragmatics, pragmatism, and pragmaticism
The term pragmatics, as introduced by Morris (see 1.), was clearly inspired by the
philosophy of pragmatism, whose main representatives in the first decades of the
20th century were William James and John Dewey, and whose founder was Peirce
(cf. Murphy 1990). In a letter of 1912, Peirce explained why and how he coined the
term in the 1870s (Houser 1998: xxi-xxii): When I gave the doctrine of pragmatism the name it bears [] I derived [ it] from , behavior, in order
that it should be understood that the doctrine is that the only real significance of a
general term lies in the general behavior which it implies (in: Eisele 1987: 95).
Morriss change of the concept of pragmatism to pragmatics made the term more
compatible with the names of the neighboring fields of pragmatics in linguistics
syntactics and semantics as well as with the term linguistics itself.
One of the main aims of pragmatism is to overcome the dualism between
thought and action (Colapietro 1992: 43031). Peirce formulated the essence of
pragmatism, as he conceived it, in several versions of his pragmatic maxim for-
184
Winfried Nth
mulated in several versions since 1878. In 1902, the maxim stated the following:
Consider what effects, that might conceivably have practical bearings, we conceive the object of our conception to have. Then, our conception of these effects is
the whole of our conception of the object (CP 5.9, c. 1905). Peirce gives the following account of how this maxim should be applied: Pragmatism is the principle
that every theoretical judgment expressible in a sentence in the indicative mood is
a confused form of thought whose only meaning, if it has any, lies in its tendency to
enforce a corresponding practical maxim expressible as a conditional sentence
having its apodosis in the imperative mood (CP 5.18, 1903). Colapietro (1992:
433) gives an example of how this principle can be followed in practice: To learn
most clearly the meaning of an assertion e.g., of Sugar is soluble requires us
to translate the assertion into a maxim such as If you place this substance into
water or similar liquids it will dissolve.
Disappointed with how his word coinage became abused in the merciless way
that words have to expect when they fall into literary clutches (CP 5.414, 1905) in
a trend that culminated in the deterioration of pragmatism into behaviorism
(Nadin 1993: 223), Peirce, in 1905, began to replace the term pragmatism with the
neologism pragmaticism, a word ugly enough to be safe from kidnappers, as its
author remarked (CP 5.414, 1905).
Parret (1983: 3) believes that there are three main lessons which the philosophy
of pragmatism has taught to linguist pragmatics: Meaning is context bound,
rationality is discourse bound, and signifying should be seen as understanding.
What linguistic pragmatics has learned from the tradition of pragmatism is certainly the importance of the study of language with respect to its consequences for
human action and behavior.
A common ground of Peirces semiotics and the classics of linguistic pragmatics is the normative approach to communication. Among the other insights
which Peirce has contributed to linguistic pragmatics is the pragmatic principle of
expressibility often attributed to Searle. Searles (1969: 220) definition is: For
any meaning X and any speaker S whenever S means (intends to convey, wishes to
communicate in an utterance, etc.) X then it is possible that there is some expression E such that E is an exact expression of or formulation of X. Peirce reduces it to the form: thought and expression are really one (CP 1.349, 1903).
3.2.
Reference and indexicality
Reference and deixis are key issues of linguistic pragmatics; in Peirces semiotics,
in which the term reference is hardly used (cf. Nth 2006), these topics are studied
as indexical signs. Searle (1969: 2629, 7296) studies reference as a speech act.
Expressions may be used to refer or they may not be used to refer. A man came is
used to refer to a man who came; John is a man is not used referentially, but
predicatively (Searle 1969: 27). Necessary conditions of a referring speech act are:
185
whatever is referred to must exist (1969: 77); There must exist one and only one
object to which the speakers utterance of the expression applies, and: The hearer
must be given sufficient means to identify the object from the speakers utterance
of the expression (Searle 1969: 82).
Existence, singularity, and referential selectivity are criteria of the genuine
index in Peirces semiotics. In fact, all of Searles conditions of reference are conditions of the genuinely indexical sign: A genuine Index and its Object must be
existent individuals (whether things or facts) (CP 2.283, c. 1902), and indexical
signs are directions of what to do to find the object (CP 2.289, c. 1893). In
contrast to Searle, Peirce also studies to what degree nonverbal signs can be genuine indices and to what degree language may be insufficient for the purpose of referring to an individual object. Furthermore, he also studies another kind of index,
which he calls degenerate index. Such indices do not refer directly to individual
existing object; instead, they refer to symbols or mental objects (cf. Nth 2000:
187). Their prototype is the relative pronoun, which is a sign referring to another
sign.
Since not all expressions are used to refer, according to Searle, not all of them
have a referent. In Peirces semiotics, too, only genuine indices have a referent in
the sense of referring indexically to an individual existent object, but although not
all signs have a referent, all signs have an object.15 Peirces semiotics is utterly incompatible with the dualistic theories of referential semantics which split verbal
reference into two mutually independent domains, one being external to the human
mind, the other being mental and hence internal to it. According to Peirce, not reference and sense are the correlates of the sign, but the object and the interpretant,
and all signs have an object as well as an interpretant. All words are signs which
represent an object and determine an interpretant, even the conjunction and. Savan
(1994: 189) describes its object as our experience of how of two or more entities
are combined to form a group, a set, or otherwise a whole, but since the object of
the sign need not be a material thing, we should add that the idea of a symmetrical
conjunction is among the objects of this sign.
A sign without an object would be a contradiction in terms in Peirces semiotics. In contrast to the referent of logical semantics, which is defined as an individual or class of existents, Peirces object of the sign may also be a feeling, an
experience, a cognition, a thought, an imagination, or even a fictional event. Unlike
the sense or meaning of a word, the interpretant is not necessarily a mental concept, a thought, or an idea corresponding to the sign; it can again be a feeling, an
action, or even be a material thing, provided it is a result of a sign. For example, the
cake baked according to a cooking recipe is among the interpretants of the recipe
practically interpreted by the cook. The object of the sign may also be something
believed formerly to have existed or expected to exist or something of a
general nature (CP 2.232, 1910). Only rarely is it a thing. It is rather that information, knowledge, or experience which an observer of the sign must have in
186
Winfried Nth
order to be able to interpret the sign. Peirce had adopted the term object from 13th
century scholastic terminology where objectum meant a creation of the mind in its
reaction with a more or less real something [] upon which cognition is directed
(MS 693A: 33; Pape 1996b: 115).
The index is defined in opposition to the symbol and to the icon as a class of
signs comprising natural but also conventional signs. An index is really affected
(CP 2.248, c. 1903) by, or physically connected with, its object; the sign and its
object make an organic pair, but the interpreting mind has nothing to do with
this connection except remarking it, after it is established (CP 2.299, c. 1895).
Among the features of Peirces index are that it directs the interpreters attention
towards the object and that it asserts nothing, but shows the object (cf. Goudge
1965: 5354). Examples of verbal and nonverbal indexical signs are: a weathercock, a yardstick, a photograph, a rap on the door, a pointing finger, any other gesture that indicates a present emotion, a cry for help, deictic words, proper names,
possessive, relative, personal, and selective pronouns.
All verbal utterances contain explicit or implicit indices referring to, or identifying, the situational circumstances, the time, place, or persons in whose context
they were uttered. The adequate interpretation of these indices is necessary to distinguish successfully between the relevant and irrelevant interpretation of a sign
(cf. 2.4):
If, for example, a man remarks, Why, it is raining! it is only by some such circumstances as that he is now standing here looking out at a window as he speaks, which
would serve as an Index (not, however, as a Symbol) that he is speaking of this place at
this time, whereby we can be assured that he cannot be speaking of the weather on the
satellite of Procyon, fifty centuries ago. (CP 4.544, 1906)
Morris did not adopt the term index, but his category of identifiors corresponds to
Peirces index (1946: 154, 362). But in contrast to Peirce, Morris restricted the
class of identifiors to spatio-temporal deixis. Identifiors signify locations in space
and time (locata) and direct behavior toward a certain region of the environment.
The identifior has a genuine, though minimal, sign status; it is a preparatorystimulus influencing the orientation of behavior with respect to the location of
something other than itself. Morris distinguished three kinds of identifiors: indicators, which are non-language signals, descriptors, which describe a spatial or
temporal location, and namors, which are language symbols, and hence, substitute signs synonymous with other identifiers.
3.3.
Meaning in dialogue and its vagueness
Several meanings are involved in the utterance of a speaker addressed to a hearer.

Searle (1969: 4548) distinguishes between intended and conventional meanings,
Leech (1983: 6) between pragmatic and semantic meanings, and Levinson (1983:
187
1617) between speaker meanings and utterance or sentence meanings. The various kinds of meaning distinguished by these authors partly overlap, but some of the
criteria according to which these distinctions are made differ significantly.
Leechs semantic meaning roughly corresponds to Levinsons sentence
meaning and Searles conventional meaning. All three kinds of meaning are opposed to what is studied in linguistic pragmatics, i.e., how meaning is intended by a
speaker and understood by a hearer. Leech defines semantic meaning as a meaning defined purely as a property of expressions in a given language, in abstraction
from particular situations, speakers, or hearers (1983: 6). Can such a meaning
abstracted from the circumstances of its use be called conventional, as Searle
suggests when he opposes conventional to intended meaning? The idea that
intended meaning is opposed to conventional meaning is rejected by Strawson
(1971), who points out that conventional meaning is not only semantic meaning
but pertains to pragmatic meaning, too, since speech acts, like words and sentences, are also determined by conventions.
The study of pragmatic meaning, according to Leech (1983: 14), begins with
the study of utterances: In fact we can correctly describe pragmatics as dealing
with utterance meaning and semantics as dealing with sentence meaning. With
this definition, Leechs dichotomy of semantic vs. pragmatic meaning is no longer
compatible with Levinsons dichotomy of utterance meaning and speaker meaning, which suggests that the study of speaker meaning is the domain of pragmatics
and not the study of utterance meaning.
Furthermore, Levinsons distinction between speaker meaning and utterance
meaning raises the question of how speaker meaning can differ from utterance
meaning at all. Is not the utterance meaning the meaning conveyed by the utterer,
i.e., speaker? Can an utterer, by means of an utterance, convey any other meaning
than his own or her own meaning?
The distinctions drawn by the three authors thus do not only overlap, they also
contradict each other. If speaker meaning is the utterers and thus also the intended
meaning, one can no longer oppose it to utterance meaning (as Levinson does).
Speaker meaning can only be intended meaning since whatever a speaker utters
must be intended, but if utterance meaning is speaker meaning and speaker meaning is intended meaning, the congruence of theses meanings makes Leechs distinction between semantic and pragmatic meanings invalid. On the other hand, if
speaker and utterance meanings are intended meanings, and if utterances are determined by conventions, too, Searles distinction between intended and conventional
meanings is invalid.
If there is speaker meaning there should also be hearer meaning. Grice (1989)
takes it into consideration but the dichotomies set up by Leech, Searle and Levinson neglect it. Sbis and Fabbri (1980: 314) criticize this neglect for reducing the
interpretive work of the addressee [] to the mere recognition of a set of coherent
intentions on the part of the agent/speaker. The neglect of the category of a hearer
188
Winfried Nth
meaning in linguistic pragmatics is due to its normative approach to the study of

conversation. Like Chomsky, who pursues a normative approach to syntax when
he postulates the idealized competent speaker as one who commits no mistakes in
uttering sentences, Searle (1969: 20) explicitly adopts a normative approach to the
description of the speakers meanings in conversation when he declares that cases
where the speaker does not say exactly what he means the principal cases of
which are nonliteralness, vagueness, ambiguity, and incompleteness are not theoretically essential to linguistic communication.
Speakers can mean more than they say (Leech 1983: 9). According to Searle
(1969: 45) the difference between what a speaker means and what he or she says is
a difference between the meanings intended by the speaker and those conventionally inherent in his or her utterances. Meaning is more than a matter of intention; it
is also at least sometimes a matter of convention, says Searle (1969). The theory
of implicatures is based on the assumption of such differences between meaning
and saying.
Peirces theory of meaning is pragmatic throughout (cf. Hilpinen 1995: 297).
Meaning is the idea which the sign attaches to its object (CP 5.6, 1905), but
meanings are not a priori inherent in signs; they can only reveal themselves in processes of semiosis (Pape 1996a: 308). Words only have meaning in so far as we
are able to make use of it in communicating our knowledge to others and in getting
at the knowledge that these others seek to communicate to us (CP 8.176, 1903).
Thus, no semantic or conventional meaning can be found in Peirces assumptions. In accordance with the pragmatic maxim, the meaning of an utterance reveals itself in its interpretations as well as in the actions and habits which are its
consequences. In one of his later definitions, Peirce concludes that the meaning of
any [assertion] is the meaning of the composite of all the propositions which that
[assertion] would under all circumstances empower the interpreter to scribe (MS
280, ca. 1905; cf. Pietarinen 2005: 1769). Since utterances have no inherent meaning independent of their use in speech acts, Peirce does not distinguish between utterance and speaker meaning. Furthermore, he distinguishes between many more
kinds of meaning. Some aspects of meaning are associated with the object of the
sign insofar as the sign presupposes collateral knowledge of its object and a common ground of knowledge shared by the speaker and the hearer. Other meanings
pertain to the interpretant, the effect of meaning created by the sign (cf. Bergman
2000: 234 for both). In the theory of the interpretant discussed above, we find the
intentional interpretant as a counterpart to speaker meaning and the effectual interpretant as the counterpart to Grices hearer meaning, but Peirce distinguishes
other kinds of interpretants (cf. Johansen 2002: 48). His immediate interpretant,
the meaning which the sign indicates before it is interpreted, comes close to
Searles conventional meaning, but the interpretant is not always the meaning of a
sign in the linguistic sense. For example Peirce distinguishes between an emotional, and energetic, and a logical interpretant. The first is an emotional, the sec-
189
ond a bodily reaction to a sign. Only the third comes close to the kind of meaning
which linguists usually describe since logical meaning is the meaning of a concept
but since not all words are concepts, words also have meanings which are not logical interpretants.
Peirce also acknowledges the possibility of differences between saying and
meaning so much discussed in linguistic pragmatics, but his account of how meaning is conveyed in the dialogue of an utterer and an interpreter is based on other
premises. Meaning is not determined by rules, conventions, and speakers intentions, as Searle (1969: 4345) sees it, but by habits, common grounds, and the collateral experience of the object of the sign. Meaning is not only vague by exception, but it is essentially vague (Nth and Santaella 2009). If meaning is fuzzy in
language because of the generality and vagueness of words, it is even vaguer if the
feelings, desires, and inner conflicts are considered which speakers and hearers
have when they communicate. Emotions, e.g., are never certain and clear; they are
necessarily vague, merely expressible and interpretable in the form of allusions,
conjectures, surmises, guesses, intuitions, and communicative negotiations. Therefore, no communication of one person to another can be entirely definite, i.e., nonvague. [] Wherever degree or any other possibility of continuous variation subsists, absolute precision is impossible. Much else must be vague, because no mans
interpretation of words is based on exactly the same experience as any other mans.
Even in our most intellectual conceptions, the more we strive to be precise, the
more unattainable precision seems (CP 5.506, c. 1905).
Vagueness and generality are the two major causes of indeterminacy in conversation, but this indeterminacy leads not only to misunderstandings or even failures
of communication. Instead, indeterminate signs are also a source of creativity to
both utterers and interpreters. The interpreter can and must exert his or her own
imagination in determining what the vague or general sign means. He or she can
specify what is left undetermined by associating it with meanings from his or her
own textual universe. The utterer of a vague sign has the advantage that he or she
can suggest meanings without being held responsible for them. Hence vagueness is
not only a risk that endangers communication; it also offers a useful potential of
creativity to those who communicate. Peirce develops these ideas of the advantages and risks of taking and giving freedom of semantic choice with an example of
a generality and one of referential vagueness:
A sign [] is objectively general in so far as it extends to the interpreter the privilege of
carrying its determination further. Example: Man is mortal. To the question, What
man? the reply is that the proposition explicitly leaves it to you to apply its assertion to
what man or men you will. A sign that is objectively indeterminate in any respect is objectively vague in so far as it reserves further determination to be made in some other
conceivable sign, or at least does not appoint the interpreter as its deputy in this office.
Example: A man whom I could mention seems to be a little conceited. The suggestion
here is that the man in view is the person addressed; but the utterer does not authorize
190
Winfried Nth
such an interpretation or any other application of what she says. She can still say, if she
likes, that she does not mean the person addressed. Every utterance naturally leaves the
right of further exposition in the utterer; and therefore, in so far as a sign is indeterminate, it is vague, unless it is expressly or by a well-understood convention rendered general. (CP 5.447, 1905)
Elsewhere, Peirce gives indefinite pronouns as the example of the risks and
chances which generality and vagueness bring to the semantic choices made by the
utterer and the interpreter: Some means that the speaker is to select an instance,
while Every or Any means that a second person is to perform the selection. Of
course, it is easier to satisfy the conditions of a statement if one can select ones
own examples, except for this, that he who undertakes to find an example guarantees that there is one, while if he leaves the selection to another, and there is none,
his statement is not broken down (CP 2.523, fn 1, 1893).
4.
Semiotics of speech acts: Peircean and Greimasian contributions
The theory of speech acts is the branch of linguistic pragmatics to which semiotics
has both contributed as a precursor and to which it has offered alternative approaches. The semiotic precursor of speech act theory is again Peirce; the alternative approach to be presented in the following is an approach based on the semiotics of A. J. Greimas. Only passing reference can be made to the semiotic
elements of a pragmatic theory of speech acts which may be found in the typology
of discourse of Charles Morris (1946: 203205; cf. Nth 2000: 94) and to the rather
different elements of a semiotics of some speech acts developed in the framework
of M. A. K. Hallidays systemic semiotics by Ventola (1987).
4.1.
Peirces contributions to speech act theory
Peirces contributions to speech act theory have been the focus of interest of several studies.16 In wider contexts they have also been discussed by Hilpinen (1995,
1998) and Pietarinen (2004, 2006). Actually, Peirce has even anticipated key terms
and issues of speech act theory, among them Austins thesis that by speaking we
do things and that this doing is an act. In a fragment of 1908, Peirce describes
the speech act of taking an oath in these terms: Taking an oath is not mainly an
event of the nature of a [] Vorstellung, or representing. It is not mere saying, but
is doing. The law, I believe, calls it an act (CP 5.546).
In anticipation of the distinction between an illocutionary act and its propositional content of an utterance later drawn by Searle (1969: 2329), Peirce defines a
proposition as the logical content of a sentence, irrespective of the particular
mood in which it is uttered (cf. Brock 1981a: 281). Sentences differ from utterances as types differ from tokens:
191
A sentence, in the sense here used, is a single object. Every time it is copied or pronounced, a new sentence is made. But a proposition is not a single thing and cannot
properly be said to have any existence. Its mode of being consists in its possibility. A
proposition which might be expressed has all the being that belongs to propositions although nobody ever expresses it or thinks it. It is the same proposition every time it is
thought, spoken, or written, whether in English, German, Spanish, Taglog, or how. A
proposition consists in a meaning, whether adopted or not, and however expressed. (MS
599: 56, ca. 1902)
Among the speech acts which Peirce describes as moods of propositions are assertion, judgment, question, denial, doubt, command, teaching, taking an oath, laying
a wager, lying, poetic language, and forms of politeness.17 Expressing a proposition in one of these moods involves an act, an exertion of energy, and is liable to
real consequences, or effects (CP 5.547, ca. 1908). The exertion of energy associated with these moods is evidently equivalent to Searles illocutionary force of
speech acts. Examples and illustrations of the consequences which such speech
acts have for their utterers and interpreter are: An assertion is an act by which a
person makes himself responsible for the truth of a proposition. Nobody ever asserted that the moon is made of green cheese; yet this is a familiar proposition
(MS 599: 5, ca. 1902). The person who took an oath has only made an assertion if
the denotation of the proposition is a fact (CP 4.500, ca. 1903):
[An oath] would be followed by very real effects, in case the substance of what is asserted should be proved untrue. This ingredient, the assuming of responsibility, which is
so prominent in solemn assertion, must be present in every genuine assertion. For
clearly, every assertion involves an effort to make the intended interpreter believe what
is asserted, to which end a reason for believing it must be furnished. But if a lie would
not endanger the esteem in which the utterer was held, nor otherwise be apt to entail
such real effects as he would avoid, the interpreter would have no reason to believe the
assertion. (CP 5.546, 1908)
By contrast, a conversation solely determined by the conventions of politeness has

no consequences as to its truth: Nobody takes any positive stock in those conventional utterances, such as I am perfectly delighted to see you, upon whose falsehood no punishment at all is visited (CP 5.546, 1908). Among the speech acts
whose consequences depend on the truth of the proposition expressed are not only
assertions and oaths, but also laying a wager. Peirce describes the consequences of
these speech acts as follows:
What is the difference between making an assertion and laying a wager? Both are acts
whereby the agent deliberately subjects himself to evil consequences if a certain proposition is not true. Only when he offers to bet he hopes the other man will make himself
responsible in the same way for the truth of the contrary proposition; while when he
makes an assertion he always (or almost always) wishes the man to whom he makes it to
be led to do what he does. Accordingly in our vernacular I will bet so and so, is the
phrase expressive of a private opinion which one does not expect others to share, while
You bet is a form of assertion intended to cause another to follow suit. (CP 5.31, 1902)
192
Winfried Nth
Peirce considers not only the conditions of publicly expressed speech acts but also
such mental acts, as a judgment or a doubt. A judgment is a mental act by which
one makes a resolution to adhere to a proposition as true, with all its logical consequences (CP 5.31, 1902). Among the mental acts expressing propositions which
are not asserted is the doubt: I may state it to myself and worry as to whether I
shall embrace it or reject it, being dissatisfied with the idea of doing either
(CP 5.31, 1902). Peirce also specifies several of those conditions of speech acts
which Searle later called preparatory conditions (cf. Brock 1981a: 282). For
example, Peirces formulation of what Searle calls the preparatory condition of a
speech act is: The only way to get a sufficient understanding of a proposition is to
imagine one person communicates in some language to a second. The utterer does
not, properly speaking, communicate it to the interpreter unless the latter is at the
time ignorant (or oblivious) of its truth (MS 284: 41, 1905; Brock 1981a).
The distinction between the illocutionary and the perlocutionary effects of a
speech act has almost no counterpart in Peirces theory of speech acts. A consequence of the pragmatic maxim is that speech acts do not only sometimes have perlocutionary effects; they have always such effects. The study of these effects is the
study of the interpretant, which is a necessary constituent of all signs (see 2.4).
4.2.
Speech acts as narrative scenarios: The Greimasian approach
Elements of the structural semiotics of A. J. Greimas18 are the foundation of a theory of speech acts developed by Sbis and Cooren.19 Sbis (1983: 100101) criticizes Searles speech act theory for being based on a causal theory of action (cf.
Davidson 1980) which results in a too strictly speaker centered approach to the
study of dialogues. According to Sbis, the causal approach to speech acts means a
regrettable return of the subject, the triumph of the empirical speaker and the resuscitation of the use of introspection as a criterion of sense (1983: 100101). To
liberate pragmatics from its mentalist bias, she proposes a semiotic conception of
acts and actions. In accordance with the definition proposed by Greimas and
Courts (1979: 5), Sbis defines an act as a fair-tre, a make-be, i.e., the producing of a state (or of a change of state). She argues that the agent of an act thus
defined is not the cause of the speech act but simply its presupposition. Instead of
the study of speech acts, Sbis proposes the study of interactions between speakers
and hearers. The focus should be on the hearer as an active participant of the dialogue; it should be taken into consideration that the hearer has the choice (i) to select an acceptable interpretation of the speech act, and (ii) to either accept the
speech act, under such an interpretation, as a successful act, or to completely or
partly reject it as more or less inappropriate and unhappy (Sbis and Fabbri
1980: 305).
Instead of investigating the utterers agency in dialogic interaction, Sbis
(1983: 102, 1987: 259) deals with action as an ascribed concept: the interpreter
193
ascribes a certain change-in-the-world to the responsibility of the [utterer as the]

agent. In this perspective, agency shifts from the utterer to the interpreter, whose
decision it is to ascribe intentions and meanings to an utterer or not. With this shift
of perspective, the analysis can take into consideration ascriptions of agency to
quasi-utterers in Peirces sense (see 2.5). According to Sbis (1987: 259), the utterer of a sign, whose agency is now merely an agency ascribed by the hearer, can
thus be a human individual, an infra-individual instance (e.g., the Unconscious), a
super-individual construction (class, party, society, church, and the like), and even
a personified natural agent. Hence, in a semiotic speech act theory adopting these
premises, not only speakers are agents, but even beings other than humans can be
said to do things with words (Cooren 2006: 8). Two common examples of assertive and commissive speech acts in which agency is ascribed to nonhuman agents
are: This decision contests his result and This signature commits you to payment (Cooren 2006: 17).
Based on her assumption that acts result in transformations of states, Sbis
(1985: 529) arrives at a redefinition of basic concepts of speech act theory: (1) A
speech act is the production of an utterance insofar as it brings about a change for
which the utterer is taken to be responsible []. (2) An illocutionary act is the
production of a change of context at the conventional level of the addressers and
the addressees modal competences. (3) A perlocutionary act is the production of
a change of context at the material (psychological or praxeological) level of the addressees reactions. An example of her Greimasian approach to the illocutionary
force of speech acts is Sbiss (1985: 530) study of the discursive figure of manipulation, which she defines as the agency of a manipulating addresser who puts the
manipulated addressee into a position of not being able not to do. The application
of Greimass theory of modalities leads Sbis (1984: 100) to a typology of modal
changes of the speakers and hearers states in speech acts, e.g.: in exercitives (such
as granting or permitting), the speaker addresses the addressee in the deontic mood
of being allowed (can), and the effect in the addressee is one of the modality of
obligation (ought); in commissives, these two modal ascriptions are inverse.
Taking up Sbiss and Greimass elements of a semiotic pragmatics, Cooren
(2000: 81169) proposes a model based on the premise that speech acts involve a
narrative scenario typically involving three actants: the speaker as the agent, the
hearer as the recipient, and the utterance as the object of transfer (Cooren 2000:
86). Uttering is seen as an act of production in which the object produced by the
agent, i.e., the utterance, is irremediably severed or dissociated from its origin
so that the producer no longer has control over the object produced and acquires
an agency of its own (Cooren 2000: 8283). A speech act is always embedded in a
plurality of verbal and nonverbal texts; it does not only refer to, it also produces a
situation. Continuing to act in absence of its utterer, the speech act becomes the
product of multiple texts that speak on behalf of their producer, creating one or
several different situations (Cooren 2000: 87).
194
Winfried Nth
Following Greimass discourse grammar, which describes narrative roles by

means of the actantial model of verb valency of Tesnire (1959), Cooren develops
a typology of illocutionary acts based on verb valencies. According to this model,
assertives are bivalent and performatives are trivalent speech acts (2000: 89). The
latter involve an agent, an object, and a recipient. The tetravalent actantial model
of the factive verbs, which involve four actants, then allows opening up the traditional speaker/hearer schema by acknowledging action from a distance, which
could be called tele-action (Cooren 2006: 2). The utterance John asked Mary to
send his report to Bill is an example of a tetravalent teleactive speech act. To telecommunicate, thus means to delegate a speech act to another agent who communicates it on behalf of the delegating agent: one discursive agent mobilizes other
agents that serve as representative or delegates of the delegating agent (Cooren
2000: 6).
In view of the fundamental differences between Peirces pragmaticist semiotics
and Greimass structural discourse semiotics, it is remarkable that both schools of
semiotics, for very different reasons, agree in their criticism of the speaker-centered approaches to speech acts. It is equally remarkable that, quite independently
of each other, both schools of semiotics postulate models of semiotic agency in
verbal interaction radically distinct from those proposed by the classics of linguistic pragmatics. Nevertheless, the fundamental difference between the two semiotic approaches in this latter respect must not be ignored. Whereas Greimass
model attributes agency to actants in discourse, Peirce attributes agency and autonomy to the sign itself.
Notes
1. Cf. Posner (1991) and Nth (2000: 433438).
2. See Parret (1983), Posner (1997), Nadin (1987, 1993), Deledalle, ed. (1989), Driel, ed.
(1991), Hilpinen (1995, 1998), Pankow (1995), Pietarinen (2004, 2005, 2006), Rellstab
(2008), and Andersen (2009).
3. Cf. Liszka (1996: 315), Pietarinen (2004), and Houser (2009: 91).
4. CP refers to Peirce (193158).
5. Cf. Fisch (1986: 339), Liszka (1996: 109), and Bergman (2007).
6. EP 2 refers to Peirce (1998).
7. Oehler (1990, 1995), Colapietro (1995), Bergman (2000, 2004), Santaella and Nth
(2004), and Pietarinen (2006).
8. See also Prieto (1966, 1975; cf. Nth 2000: 228).
9. See 2.2 and Burkhardt (1990).
10. MS refers to Peirce (1979).
11. Cf. Allwood (1978: 4) and Levinson (1983: 103).
12. For Peirces distinction between presciss (from prescind) and precise see Spinks (1991:
1923) and CP 5.449, 1905. Roughly, prescission is giving attention to one element
and neglect of the other (CP 1.549, 1867).

13.
14.
15.
16.
17.
18.
19.
195
W 1 refers to Peirce (1981).

Cf. Oehler (1995: 268) and Johansen (1993).
Cf. Santaella (1988, 1990, 1994), Joswick (1996), and Pape (1996b).
Brock (1975, 1981a, b), Martens (1981), and Thibaud (1997).
Cf. Brock (1981b) and Hilpinen (1998).
For Greimas, see Nth (2000: 11219).
Sbis and Fabbri (1980), Sbis (1983, 1984, 1985, 1987, 1994, 2002), and Cooren
(2000, 2006).
References
Allwood, Jens
1978
On the Analysis of Communicative Action (Gothenburg Papers in Theoretical
Linguistics 38). Gothenburg: University, Department of Linguistics.
Andersen, Christiane
2009
Semiotics. In: Jef Verschueren and Jan-Ola stman (eds.), Key Notions for
Pragmatics, 217228. Amsterdam: Benjamins.
Austin, John
1962
How to Do Things with Words. London: Oxford University Press.
1954
Bergman, Mats
2000
Reflections on the role of the communicative sign in semeiotic. Transactions of
the Charles S. Peirce Society 36.2: 225254.
Bergman, Mats
2003
Peirces derivations of the interpretant. Semiotica 144: 117.
Bergman, Mats
2004
Fields of Signification. Explorations in Charles S. Peirces Theory of Signs.
(Philosophical Studies from the University of Helsinki 6), Helsinki, PhD dissertation.
Bergman, Mats
2007
The secret of rendering signs effective: the import of C. S. Peirces semiotic
rhetoric. The Public Journal of Semiotics 1.2: 211.
Brock, Jarrett
1975
Peirces conception of semiotics. Semiotica 14: 124141.
Brock, Jarrett
1979
Principal themes in Peirces logic of vagueness. Peirce Studies (Lubbock,
Texas) 1: 4150.
Brock, Jarrett
1981a
An introduction to Peirces theory of speech acts. Transactions of the Charles
S. Peirce Society 17: 319326.
Brock, Jarrett
1981b
Peirce and Searle on assertion. In: Kenneth Ketner, Joseph M. Ransdell, Carolyn Eisele, Max H. Fisch and Charles S. Hardwick (eds.), Proceedings of the
C.S. Peirce Bicentennial International Congress. 281287. Lubbock, TX:
Texas Tech Press (Graduate Studies Texas Tech University 23).
196
Winfried Nth
Bhler, Karl
1933
Die Axiomatik der Sprachwissenschaft. Frankfurt am Main: Klostermann.
Burkhardt, Armin, (ed.)
1990
Speech Acts, Meaning and Intentions. Berlin: Walter de Gruyter.
Buyssens, Eric
1943
La communication et larticulation linguistique. Bruxelles / Paris: Presses
Universitaires (revised edition 1964).
Carnap, Rudolf
1975
Introduction to Semantics and Formalization of Logic, 2 vols. in 1. Cambridge:
Harvard University Press [1942/43].
Colapietro, Vincent
1992
Purpose, power, and agency. The Monist 75: 423444.
Colapietro, Vincent
1995
Immediacy, opposition, and mediation: Peirce on irreducible aspects of
the communicative process. In: Leonore Langsdorf and Andrew R. Smith
(eds.), Recovering Pragmatisms Voice: The Classical Tradition, Rorty, and
the Philosophy of Communication, 2348. Albany, NY: State University
Press.
Cooren, Franois
2000
The Organizing Property of Communication. Amsterdam: Benjamins.
Cooren, Franois
2006
Between semiotics and pragmatics: Opening language studies to textual
agency. Paper presented at the annual meeting of the International Communication Association, Dresden. <http://www.allacademic.com/meta/p87081_
index.html, access 05-25-2009>.
Davidson, Donald
1980
Essays on Actions and Events. Oxford: Oxford University Press.
Deacon, Terrence
1997
The Symbolic Species. New York, NY: Norton.
Deledalle, Grard (ed.)
1989
Semiotics and Pragmatics. Amsterdam: Benjamins.
Derrida, Jacques
1988
Limited Inc. Evanston, IL: Northwestern University Press.
Driel, Herbert van (ed.)
1991
Het semiotisch pragmatisme van Charles S. Peirce. Amsterdam: Benjamins.
Eisele, Carolyn
1987
Peirces pragmatics. In: Herbert Stachowiak (ed.), Der Aufstieg des pragmatischen Denkens im 19. und 20. Jahrhundert (Pragmatik: Handbuch des pragmatischen Denkens, vol. 2.), 8397 Hamburg: Meiner.
Fisch, Max H.
1986
Peirce, Semeiotic, and Pragmatism. Ed. Kenneth L. Ketner and Christian
J. W. Kloesel. Bloomington: Indiana University Press.
Gorl, Dinda
1993
Evolving through time: Peirces pragmatic maxims. Semiosis 71/72: 313.
Goudge, Thomas A.
1965
Peirces index. Transactions of the Charles S. Peirce Society 1: 5270.
197
Greimas, Algirdas J. and Joseph Courts

1979
Smiotique: Dictionnaire raisonn de la thorie du langage. Paris: Hachette.
Trad. 1982. Semiotics and Language: An Analytical Dictionary. Bloomington,
IN: Indiana University Press.
Grice, Paul
1957
Meaning. Philosophical Review 64: 377388.
Grice, Paul
1989
Habermas, Jrgen
1981
Theorie des kommunikativen Handelns, 2 vols. Frankfurt am Main: Suhrkamp.
Habermas, Jrgen
1995
Peirce and communication. In: Kenneth Laine Ketner (ed.), Peirce and Contemporary Thought. Philosophical Inquiries, 243266. New York: Fordham
University Press.
Hervey, Sndor G. J.
1982
Semiotic Perspectives. London: Allen and Unwin.
Hilpinen, Risto
1995
Peirce on language and reference. In: Kenneth Laine Ketner (ed.), Peirce and
Contemporary Thought. Philosophical Inquiries, 272303. New York: Fordham University Press.
Hilpinen, Risto
1998
Pragmatics and pragmatism: On C.S. Peirces pragmatic theory of meaning. In:
Paul Weingartner, Gerhard Schurz, and Georg Dorn (eds.), Die Rolle der Pragmatik in der Gegenwartsphilosophie The Role of Pragmatism in Philosophy,
151164 (=Akten des 20. Internationalen Wittgenstein-Symposiums. Schriftenreihe der Wittgenstein-Gesellschaft, vol. 26) Wien: Hlder-Pichler-Tempsky.
Houser, Nathan
1998
Introduction. In: Nathan Houser and Christian J. W. Kloesel (eds.), The Essential Peirce. Vol. 1, xvii-xxxviii. Bloomington: Indiana University Press.
Houser, Nathan
2009
Peirce, phenomenology and semiotics. In: Paul Cobley (ed.), The Routledge
Companion to Semiotics, 89100. London: Routledge.
Jakobson, Roman
1960
Linguistics and poetics. In: Thomas A. Sebeok (ed.), Style in Language,
350377. Cambridge, MA: MIT Press.
Johansen, Jrgen Dines
1993
Dialogic Semiosis. Bloomington: Indiana University Press.
2002
Literary Discourse: A Semiotic-Pragmatic Approach to Literature. Toronto:
University of Toronto Press.
Joswick, Hugh
1996
The object of semeiotic. In: Vincent M. Colapietro and Thomas M. Olshewsky
(eds.), Peirces Doctrine of Signs, 93102. Berlin: de Gruyter.
Kappner, Stefan
2004
Intentionalitt aus semiotischer Sicht: Peirceanische Perspektiven. Berlin: de
Gruyter.
198
Winfried Nth

1980
Metaphors We Live By. Chicago: University Press.
Leech, Geoffrey N.
1983
1983
Liszka, J. Jakb
1996
A General Introduction to the Semeiotic of Charles S. Peirce. Bloomington: Indiana University Press.
Liszka, J. Jakb
2009
Rethinking the pragmatic theory of meaning. Cognitio (So Paulo) 10.1:
6180.
Luhmann, Niklas
1984
Soziale Systeme. Frankfurt am Main: Suhrkamp.
Martens, Ekkehard
1981
C. S. Peirce on speech acts. In: Kenneth Ketner, Joseph M. Ransdell, Carolyn
Eisele, Max H. Fisch and Charles S. Hardwick (eds.), Proceedings of the
C.S. Peirce Bicentennial International Congress, 289292. Lubbock, TX:
Texas Tech Press (=Graduate Studies Texas Tech University 23).
Montague, Richard
1970
Pragmatics and intensional logic. Synthese 22: 6894.
Morris, Charles W.
1938
Foundations of the Theory of Signs (Foundations of the Unity of Science:
Towards an International Encyclopedia of Unified Science, vol. 1.2.) Chicago,
IL: University Press [1970].
Morris, Charles W.
1946
Signs, Language, and Behavior. New York: Braziller. (Quoted from the reprint
in C. W. Morris. 1971. 73398.)
Morris, Charles W.
1964
Signification and Significance: A Study of the Relations of Signs and Values.
Cambridge, MA: MIT Press.
Morris, Charles W.
1971
Murphy, John P.
1990
Pragmatism from Peirce to Davidson. Boulder, CO: Westview Press.
Nadin, Mihai
1987
Pragmatics in the semiotic frame. In: Herbert Stachowiak (ed.), Der Aufstieg
des pragmatischen Denkens im 19. und 20. Jahrhundert, 148170. (Pragmatik: Handbuch des pragmatischen Denkens, vol. 2.) Hamburg: Meiner.
Nadin, Mihai
1993
Semiotics in action: The pragmatic level. In: Herbert Stachowiak (ed.), Spachphilosophie, Sprachpragmatik und formative Pragmatik, 219250. (Pragmatik: Handbuch des pragmatischen Denkens, vol. 4.) Hamburg: Meiner.
Nth, Winfried
2000
Handbuch der Semiotik, 2nd rev. ed. Stuttgart: Metzler.
Nth, Winfried
2002a
Charles Sanders Peirce, pathfinder in linguistics. Interdisciplinary Journal for
Germanic Linguistics and Semiotic Analysis 7.1: 114.
199
Nth, Winfried
2002b
Wrter als Zeichen: Einige semiotische Aspekte der Sprache. In: Jrgen
Dittmann and Claudia Schmidt (eds.), ber Wrter, 932. Freiburg: Rombach.
Nth, Winfried
2006
Representation and reference according to Peirce. Caderno. 9a Jornada do
Centro de Estudos Peirceanos / IIIrd Advanced Seminar on Peirces Philosophy
and Semiotics, 7183. (So Paulo, PUC-SP), So Paulo.
Nth, Winfried
2008
Natural signs and the natural ingredients of cultural signs. In: Martin Dring,
Hermine Penz and Wilhelm Trampe (eds.), Language, Signs and Nature: Ecolinguistic Dimensions of Environmental Discourse, 2137. Tbingen: Stauffenburg.
Nth, Winfried
2009
On the instrumentality and semiotic agency of signs, tools, and intelligent machines. Cybernetics and Human Knowing 16.12: 1136.
Nth, Winfried and Lucia Santaella
2009
Die Relevanz der Peirceschen Semiotik des Vagen fr die Theorie der Kommunikation. In: Siegfried J. Schmidt (ed.), Die (Un)mglichkeit der Kommunikation. Frankfurt am Main: Peter Lang (Special Issue SPIEL: Siegener Periodicum zur Internationalen Empirischen Literaturwissenschaft 26.1 [2007],
7396).
Oehler, Klaus
1990
Peirce oder Habermas? In: Elisabeth Walther and Udo Bayer (eds.), zeichen
von zeichen fr zeichen: festschrift fr max bense, 4247. Baden-Baden: Agis.
Oehler, Klaus
1995
A reponse to Habermas. In: Kenneth Laine Ketner (ed.), Peirce and Contemporary Thought, 267271. New York: Fordham University Press.
Pankow, Christiane
1995
Semiotics. In: Jef Verschueren, Jan-Ola stman and Jan Blommaert (eds.),
Handbook of Pragmatics Manual, 469476. Amsterdam: Benjamins.
Pape, Helmut
1996a
Charles Sanders Peirce. In: Tilman Borsche (ed.), Klassiker der Sprachphilosophie, 307324. Mnchen: Beck.
Pape, Helmut
1996b
Object and final cause in Peirces semeiotic. In: Vincent M. Colapietro and
Thomas M. Olshewsky (eds.), Peirces Doctrine of Signs, 103117. Berlin: de
Gruyter.
Parmentier, Richard J.
1985
Signs place in medias res: Peirces concept of semiotic mediation. In: Elizabeth Mertz and Richard J. Parmentier (eds.), Semiotic Mediation, 2348. Orlando, FL: Academic Press.
Parret, Herman
1983
Parret, Herman (ed.),
1994
Pretending to Communicate. Berlin: de Gruyter.
200
Winfried Nth
Peirce, Charles Sanders

193158 Collected Papers, Vols. 16. ed. Charles Hartshorne and Paul Weiss, Vols. 78,
ed. Arthur W. Burks. Cambridge, MA: Harvard University Press (quoted as
CP).
1979
The Charles S. Peirce Papers, 30 reels, 3rd microfilm edition. Cambridge, MA:
The Houghton Library, Harvard University, Microreproduction Service
(quoted as MS).
1981
Writings of Charles S. Peirce. Vol. 1: 18571866. Bloomington, IN: Indiana
Univ. Press (quoted as W 1).
1998
The Essential Peirce. Selected Philosophical Writings. Vol. 2, 18931913).
Ed. by the Peirce Edition Project. Bloomington, IN: Indiana University Press
(quoted as EP 2).
Pietarinen, Ahti-Veikko
2004
Grice in the wake of Peirce. Pragmatics and Cognition 12.2: 295315.
2005
Relevance theory through pragmatic theories of meaning. In: Bruno G. Bara,
Lawrence Barsalou and Monica Bucciarelli (eds.), Proceedings of the XXVII
Annual Conference of the Cognitive Science Society, 17671772. Mahwah, NJ:
Lawrence Erlbaum,
2006
Signs of Logic. Peircean Themes on the Philosophy of Language, Games, and
Communication. Dordrecht: Springer.
Plato
1953
The Dialogues of Plato, transl. Benjamin Jowett, 4th ed. (1st ed. 1871). Oxford:
Clarendon. Also: http://www.gutenberg.org/dirs/1/6/1/1616/1616.txt.
Posner, Roland
1985
La syntactique, sa relation la smantique et la pragmatique, la morphologie
et la syntaxe, et la syntagmatique et la paradigmatique. In: Herman Parret and
Hans-Georg Ruprecht (eds.), Exigences et perspectives de la smiotique. Recueil dhommages pour A.J. Greimas, vol. 1, 7396. Amsterdam: Benjamins.
Posner, Roland
1991
Research in pragmatics after Morris. Dedalus (Lisbon) 1: 115156.
Prieto, Luis J.
1966
Messages et signaux. Paris: Presses Universitaires.
Prieto, Luis J.
1975
Etudes de linguistique et de smiologie gnrales. Genve: Droz.
Ransdell, Joseph
1977
Some leading ideas of Peirces semiotic. Semiotica 19: 157178.
Rellstab, Daniel
2008
Peirce for linguistic pragmatics. Transactions of the Charles Sanders Peirce
Society 44.2: 312346.
Santaella, Lucia
1988
Charles Sanders Peirces object (of the sign). Versus 49: 5358.
201
Santaella, Lucia
1990
An introduction to Peirces concept of object. Cruzeiro semitico 13: 7582.
Santaella, Lucia
1994
Why things are objects of the sign. Cruzeiro semitico 18: 3746.
Santaella, Lucia
1999
Methodeutics, the liveliest branch of semiotics. Semiotica 124: 377395.
Santaella, Lucia and Winfried Nth
2004
Comunicao e semitica. So Paulo: Hacker.
Savan, David
1994
C. S. Peirce and American semiotics. In: Michael Shapiro and Michael
C. Haley (eds.), The Peirce Seminar Papers 2, 179208. Providence, RI: Berghahn.
Sbis, Marina
1983
Actes de langage et (acte d)nonciation. Langages 18.70: 99106.
Sbis, Marina
1984
On illocutionary types. Journal of Pragmatics 8: 93112.
Sbis, Marina
1985
Manipulation et sanction dans la dynamique des actes de langage. In: Herman
Parret and Hans-Georg Ruprecht (eds.), Exigences et perspectives de la smiotique. Recueil dhommages pour A.J. Greimas, Vol. 1, 529538. Amsterdam:
Benjamins.
Sbis, Marina
1987
Speech acts and context change. In: Thomas T. Ballmer and Wolfgang Wildgen
(eds.), Process Linguistics, 254280. Tbingen: Niemeyer.
Sbis, Marina
1994
Language and dialogue in the framework of the analytical philosophy of ordinary language. In: Edda Weigand (ed.), Concepts of Dialogue, 159170. Tbingen: Niemeyer.
Sbis, Marina
2002
Cognition and narrativity in speech act sequences. In: Anita Fetzer and Christiane Meierkord (eds.), Rethinking Sequentiality, 7197. Amsterdam: Benjamins.
Sbis, Marina and Paolo Fabbri
1980
Models ( ? ) for a pragmatic analysis. Journal of Pragmatics 4: 301319.
Searle, John R.
1969
Speech Acts: An Essay in the Philosophy of Language. Cambridge: University
Press.
Searle, John R.
1983
Intentionality. Cambridge: University Press.
Shannon, Claude E. and Warren Weaver
1949
The Mathematical Theory of Communication. Urbana: University of Illinois
Press.
Short, Thomas L.
2007
Peirces Theory of Signs. Cambridge: University Press.
1986
202
Winfried Nth
Spinks, Cary W.
1991
Peirce and Triadomania. Berlin: de Gruyter.
Stetter, Christian
1989
Some remarks concerning the theory of speech acts and the idea of pragmatics. In: Gerard Deledalle (ed.), Semiotics and Pragmatics, 155165.
Amsterdam: Benjamins.
Strawson, Peter
1971
Intention and convention in speech acts. In: John R. Searle (ed.), The Philosophy of Language, 2338. Oxford: University Press.
Tesnire, Lucien
1959
Elments de syntaxe structurale. Paris: Klincksieck.
Thibaud, Pierre
1997
Between saying and doing: Peirces propositional space. Transactions of the
Charles Sanders Peirce Society 33: 270327.
Ventola, Eija
1987
The Structure of Social Interaction: A Systemic Approach to the Semiotics of
Service Encounters. London: Frances Pinter.
Watzlawick, Paul, Janet H. Beavin, and Don D. Jackson
1967
Pragmatics of Human Communication. New York: Norton.
1953
Philosophische Untersuchungen Philosophical Investigations. Oxford:
Blackwell.
7.

Nikola Kompa and Georg Meggle
Introduction
According to what might be called the traditional paradigm in the philosophy of
language, linguistic interpretation is a linear and serial process. In particular, pragmatic interpretation (such as the working out of Gricean implicatures) is claimed
to commence only after the semantic interpretation of a given sentence has been
worked out. This traditional paradigm has come under attack from various directions. Some philosophers contend, for instance, that there is no such thing as a
purely semantic content or a semantically expressed proposition (or truth-condition) but that pragmatic inferences are necessary at every level of linguistic interpretation: this claim will be discussed in the first part (chapter 1) of this article. One
may even go one step further and maintain that semantic notions need to be spelled
out pragmatically in the first place, that linguistic interpretation requires comprehension of a particular type of action and that semantics is thus simply a part of
pragmatics: this will be explored in the second part (chapter 2) of our article.
1.
Pragmatic inferences and semantic rules
1.1.
Two approaches to language and meaning
In 1938, Charles Morris distinguished three different dimensions of semiosis in

his theory of signs:1
Syntax studies the relations of signs to other signs, semantics is concerned with the
relation between signs and objects, and pragmatics takes the relation between signs
and their interpreters as its subject matter. To partition the domain this way has
been common ever since. Morriss characterization, however, leaves the three
areas severely underdetermined. For example, pragmatics is commonly supposed
204
to cover such diverse phenomena as speech acts, deixis, presuppositions or implicatures, their unifying feature being only that they have something to do with the
relation between signs and interpreters (and that they cannot easily be accommodated within a semantic theory). Also, the characterization may be not as clear-cut
as Morris and others hoped it to be. In particular (and as a consequence), where
exactly to draw the line between semantics and pragmatics is a hotly contested
issue in current philosophical and linguistic debate.2 Moreover, the difficulties one
encounters when trying to sharpen the distinction between semantics and pragmatics point to a deeper problem about how to theoretically approach language in
general. It is common in the philosophy of language to contrast two different approaches to questions concerning language, meaning and interpretation. Robert
Brandom, for example, puts the difference thus: some philosophers view language
as a kind of practice or activity, others prefer a semantic Tarskian approach to language (cf. Brandom 2002: 4041). The former focus on how language is used; they
study language as a kind of practice, as something speakers do in order to achieve
certain aims. The latter, on the other hand, strive to develop a theory of content that
is closely tied to Alfred Tarskis truth-theoretical considerations. Underlying the
two different approaches to language are different views of what a language is,
which David Lewis aptly characterizes thus:
What is a language? Something which assigns meanings to certain strings of types of
sounds or of marks. It could therefore be a function, a set of ordered pairs of strings and
meanings. [] (Lewis 1975/1983: 163)
Those who favor this view of what a language is think of language as an abstract
system that systematically pairs certain strings of sounds or marks with certain
other entities, called meanings. The mapping can be done, it seems, without taking
language users into account. In the contrasting view, on the other hand, the language user takes center stage:
What is a language? A social phenomenon which is part of the natural history of human
beings; a sphere of human action, wherein people utter strings of vocal sounds, or inscribe strings of marks, and wherein people respond by thought or action to the sounds
or marks which they observe to have been so produced. (Lewis 1975/1983: 164)
Which of these two views of language will prove theoretically fruitful is the point
at issue between semantically and pragmatically oriented accounts of language and
linguistic interpretation. Very roughly, the pragmaticist will advocate a usagebased approach to language. S/he will, accordingly, think of language as a human
activity, a social phenomenon that helps us interact and communicate with our fellow human beings. His/her opponent, the semanticist, takes a language to be a
formal system that can be studied fairly independently of how human beings actually use language. His/her aim is to come up with a formal semantic theory that
allows us to assign every sentence of language L its meaning in L. More specifically, s/he will try to pair syntactically analyzed sentences (of a language) with
205
their truth-conditions, as s/he typically takes the meaning of a sentence to be its

truth-condition (cf., e.g., Portner and Partee 2002: 3).
Semantic interpretation was commonly taken to be independent of pragmatics.
Some even claimed that semantics comes first (cf. Manor 2001: 67f). Pragmatic interpretation (such as the derivation of Gricean implicatures) comes in only after the
semantic interpretation of the expressions in question has been worked out. But
new work on the context sensitivity of natural languages questions not only the semantics first-view but the idea of semantic independence as such. More specifically, in recent years the project of a formal semantic theory has come under attack
from what came to be called contextualist approaches to language and meaning.
Very roughly, the contextualist can be characterized as someone who takes context sensitivity to be a ubiquitous feature of natural languages, a feature much more
pervasive than deixis or anaphora and much less well-behaved than implicatures,
for example. S/he favors a pragmatically oriented, usage-based approach to meaning, arguing that linguistic (even truth-conditional) interpretation involves pragmatic inferences all the way down. Her/his opponent, on the other hand, advocates
a formal semantic approach to meaning according to which truth-conditional interpretation is a purely syntactic-semantic endeavor. S/he would, presumably, agree
with the following characterization:
[] in general, formal theories can be characterized as fundamentally syntax-driven
theories, which claim that it is possible to deliver an account of the propositional or
truth-conditional content of a sentence in natural language simply via formal operations
over the syntactic features of that sentence, that is, over the lexical items it contains and
their mode of composition. (Borg 2004b: 3)
The pragmatically oriented contextualist and his/her semantically oriented opponent negotiate what was traditionally taken to be key semantic notions: the notion
of what is said in the utterance of a sentence, the notion of the proposition expressed by the utterance, the notion of a sentences truth-conditions, and at the
most basic level the notion of meaning itself. Semanticists try to save one or another of these notions from pragmatic intrusion.
1.2.
Formal semantics
Let us first try to get a better grasp of what form a formal semantic theory might
take before we consider, in a second step, the objections the contextualist raises
against this kind of approach. Very roughly, formal semantic theories try to develop, for as large a fragment of natural language as possible, a formal representation that contains all syntactically and semantically relevant information. In a
first step, the logical form of each sentence in the relevant fragment of the language
in question has to be worked out. In a second step, the elements of logical form
have to be interpreted, i.e., each element has to be assigned a semantic value in ac-
206
cordance with the respective interpretation rules. More specifically, the idea is that
there are interpretation rules that assign semantic values to the basic expressions in
the language. Composition rules then tell us which value to assign to complex expressions, the semantic value of a complex expression being a function of the semantic values of its parts. In this manner we work our way up until, after finitely
many steps, we are able to assign semantic values (i.e., truth-conditions) to whole
sentences. Compositionality obviously plays an important role here, for if language were not compositional, no recursive procedure for assigning semantic
values to expressions of arbitrary complexity would be available. Yet such a recursive procedure is exactly what the semanticist is looking for: a procedure that
yields for every well-formed sentence (of finite length) its meaning, i.e., the proposition expressed or its truth-conditions, after finitely many steps. And to interpret
an expression is to follow such a recursive procedure. Here is a somewhat lengthy
quote from Jason Stanley, a (moderate) proponent of formal semantics, to illustrate
the point:
Successful interpretation involves assigning denotations to the constituents of the logical form, and combining them in accord with composition rules that do not vary with
extra-linguistic context. The denotations that successful interpreters will assign to constituents of a logical form will be constrained by the linguistic conventions governing
those elements. [] What results from a successful application of this first stage of
interpretation is a unique proposition, a fully truth-evaluable entity. (Stanley 2002:
149150)
One of the attractive features of a formal semantic approach is that it seems to provide a good explanation of the productivity of language. Obviously, competent
speakers are able to produce novel sentences and to understand sentences they
have not encountered before. This fact stands in need of explanation. The formal
semanticist seems able to provide an explanation by pointing out the recursive
character of linguistic interpretation.
Yet what semantic value do interpretation rules assign to the basic expressions
of a language? What is the semantic value, i.e. the meaning, of a linguistic expression? It is the answer to this question on which philosophical theories of meaning differ. According to the reference theory of meaning it is the expressions reference, often called its extension: what the expression is used to talk about or refer
to. Unfortunately, the reference theory of meaning faces a couple of serious and
well-known objections, some of which already attracted Gottlob Freges attention
and brought him to distinguish between an expressions reference (what he called
Bedeutung) and its sense (Sinn, cf. Frege 1891, 1892a and 1892b). Freges insights, in turn, loomed large in the development of so called intensional semantics.
Rudolf Carnap, who attended some of Freges lectures in Jena, strived to find a
suitable method for the semantic analysis of meaning (Carnap 21956: 2) and proposed the concept of extensions and intentions of an expression as a suitable tool.
This distinction between an expressions extension and intension soon caught on.
207
Very roughly, the idea is this: It may not be adequate to equate an expressions
meaning with its actual extension because the latter may depend on the way the
world is, which may be subject to change. Yet isnt the task of an expressions
meaning to pursue this kind of dependence? In other words, isnt it the case that
meaning determines the expressions extension in any possible world? Such functions from possible worlds to extensions are called intensions. The intension of an
expression is supposed to determine its extension (i.e. its truth-value, in the case of
whole sentences) in any possible world. In other words, intensions take possible
worlds as arguments and provide extensions as truth-values. Hence, intensions
seem to be better candidates than extensions for what we intuitively think of as an
expressions meaning.
One of the most influential later developments in intensional semantics is Montague Grammar. It was originally developed by Robert Montague who elaborated
on ideas introduced by Rudolf Carnap, Alfred Tarski and Alonzo Curch and subsequently modified and expanded by other logicians, philosophers and linguists
(cf. Montague 1970a, 1970b, 1973, and also Partee 1973, 1975; Porter and Partee
2002).
In the course of these considerations it became more and more obvious that the
extension of an expression not only depends on the way the world is but also, in
some cases at least, on quite specific features of the context in which the expression
is used. The most prominent examples are indexical expressions such as I,
here or now. What these expressions refer to depends on the context of utterance; similarly for demonstrative pronouns such as this or that. Insight into
this double-dependence of extension on context and world respectively has led to
the development of what came to be called two-dimensional semantics. According
to a renowned view held by David Kaplan (1989), for example, the overall import
of a linguistic expression breaks down into two components: its character and its
content. Both can be construed as functions, the character of an expression being a
function from the context of utterance into the expressions content, the content
being a function from possible circumstances of evaluation (very roughly, something like a possible world or an index, terminology differs here) into the extension
of the expression in the respective circumstances.3 This opens up a new way of accommodating context sensitive expressions within a formal semantic theory: just
associate with each context sensitive expression a function or rule (a character)
that determines, for any possible context of its use, the expressions content (its extension, in the case of indexical or demonstrative expressions) in that context.
Thought of in this way, linguistic (truth-conditional) interpretation is (context-sensitivity notwithstanding) still a purely syntactic-semantic enterprise. No pragmatic
processes seem to be required; a result that nicely fits in with the formal semantic
project as sketched above.
208
1.3.
Context-sensitivity and pragmatic inferences
Still, as mentioned above, the formal semantic project came under attack when it
became obvious that there is much more context-sensitivity in natural language
than the formal semanticist is willing to acknowledge. Moreover, it is far from
clear that she could accommodate all the context-sensitivity within her formal
semantic approach even if she wanted to. The main problem is that pragmatic inferences seem to be required after all in order to truth-conditionally interpret utterances containing context-sensitive expressions and to assign these expressions a
semantic value in context. Yet the nature of pragmatic inferences does not square
well with the allegedly deductive, algorithmic nature of semantic interpretation
(cf. Borg 2004a: 218219).
1.3.1.
The non-monotonic nature of pragmatic inferences
The need for pragmatic inferences in linguistic interpretation has long been acknowledged in philosophical and linguistic theorizing. Ever since Paul Grice developed his theory of conventional and conversational implicatures the need for
pragmatic inferences had become obvious. According to Grice (1989), what a
speaker means by an utterance divides up into what she says and what she implicates. What she says is (subject to certain qualifications) closely tied to the conventional meaning of the words used. The implicatures (at least the conversational ones), on the other hand, have to be pragmatically worked out. Usually, in
generating an implicature, a speaker exploits certain maxims her hearers expect
her to comply with, maxims to the effect that she should try to be relevant, succinct, truthful and informative. By violating or flouting one of these maxims, a
speaker can communicate something other or even more than what she explicitly
said.
To use one of Grices own examples, if you ask me whether our common friend
Smith has a girlfriend at the moment, and I reply by saying that he has been paying
a lot of visits to Zurich lately, you might reason as follows. What she said was not
directly relevant for my question. But then, I have no reasons to assume that she is
not cooperative. So she must have meant to communicate something other than
what she said. What could that be? Maybe she meant to suggest that Smith has a
girlfriend in Zurich. The conclusion is defeasible, though. It might need revision
in the light of new information. And this line of reasoning is an example of a pragmatic inference. Since Grice, it has been common to think of pragmatic inferences
as being restricted to cases such as these. More specifically, pragmatic inferences
are commonly taken to come into play in linguistic interpretation only after the
semantic, truth-conditional content of the sentences has been worked out. Yet that
is exactly what the contextualist denies. She claims that pragmatic inferences are
required all the way down. And the problem is that the defeasible nature of prag-
209
matic inferences clashes with the deductive nature of semantic interpretation. This
is problematic for various reasons:
Firstly, because in these cases it is not enough to just follow any allegedly semantic rule or algorithm, because there is no semantic rule or algorithm to follow.
(Though that is not to say that there might not be any syntactic rule, i.e. a rule of
grammar).
Secondly, a different conception of context is called for than the one invoked in
syntactic-semantic interpretation if our aim is to accommodate all different forms
of context-sensitivity. According to a very common conception of context (e.g.
Lewis 1980; Kaplan 1989), an utterance context is taken to comprise a neat list of
contextual features such as a world, a speaker, a time and a place of utterance, and
thus to determine at least the values of indexical expressions.4 Yet such a conception of context will not do for present purposes. For one thing, because the idea of
context determining anything stands in dire need of explanation. More importantly,
what is needed is a conception of context according to which a context is seen as
something that provides information: context has to provide all the information
that speaker and interpreters have to draw on in order for communication and interpretation to be successful. That seems to be the idea behind Kent Bachs notion
of wide context:
Wide context does not literally determine anything. It is the body of mutually evident information that the speaker exploits to make his communicative intention evident and
that his audience relies upon, taking him to intend them to do so, to identify that intention. (Bach and Bezuidenhout 2002: 285)
Bach emphasizes that wide context only enables the hearer to figure out the semantic value of a given expression but does not determine it. Content is ascertained
on the basis of information provided by wide context. This epistemic notion of
context as providing speaker and hearer with information about how to interpret a
given utterance seems to be better suited for modelling all different forms of context-sensitivity.5
Thirdly, and most importantly, pragmatic inferences often take the form of an
inference to the best explanation or rather to the best interpretation.6 As a result,
they are liable to be defeated by new information. And there is no limit to the
amount of information that may turn out to be relevant (cf. Borg 2007: 356 and
Borg 2004b: 79; Recanati 2004: 54). More specifically, pragmatic inferences are
defeasible, non-monotonic, in the following sense: It may be the case that although
conclusion C follows (not in the deductive sense, of course) from a set P of premises, it no longer follows from a set M, such that M P, i.e. such that P is a proper
subset of M. In other words, adding new premises might turn a good inference
into a bad one. (Defeasible reasoning is, of course, not confined to cases of linguistic interpretation. There is, by now, a huge literature on this mode of reasoning;
see for an overview Koons 2009). But if that is so, then there may not be an algo-
210
rithmic procedure for pragmatic interpretation as there is no guarantee that the task
can be completed after finitely many steps. Inferences concerning speaker intentions are examples of pragmatic inferences. Here is an example of Borgs to illustrate the defeasible nature of intentional interpretation:
So, imagine that you see Sally filling a glass of water from the tap. Then you might reason as follows: Sally is getting a glass of water form the tap. The best explanation for
this action is that Sally is thirsty and wants a drink; therefore Sally is thirsty and wants a
drink. Clearly this is a non-demonstrative piece of reasoning and it is susceptible to the
influence of an open-ended range of contextual factors. For instance, say that you know
that Sally has just come in with Sourav and that Sourav is wearing running gear and
looks out of breath, then the best explanation for Sallys action might be that Sourav is
thirsty and Sally is getting a drink for him. Or imagine that Sally has just glanced at her
potted plant, then the best explanation might be that she wants to water her plant. (Borg
2004b: 7879)
Interpreting a persons actions in the way described clearly requires pragmatic inferences. And the interpretation of linguistic utterances is similar in that respect to
the interpretation of actions. For many linguistic expressions are such that in order
to figure out what has been said, speaker intentions at least to the extent that
speakers manage to make them manifest need to be taken into account. Actually,
even indexical and demonstrative expressions seem to require pragmatic inferences for their interpretation; what place or time exactly a speaker refers to with an
indexical expression seems to depend, partially at least, on which region in space
or what stretch of time he intended to refer to. (But let us put these problems aside
for the time being.) Yet intentional interpretation is not the only case where pragmatic inferences come into play. In interpreting utterances, hearers not only draw
on contextually provided information about (manifest) speaker intentions; other
contextual information has to be taken into account, too.
1.3.2.
Context-sensitivity some examples
For the purpose of illustration, let us consider a few more examples of context-sensitivity in natural language. Suppose a speaker says
The suitcase is too heavy.
at one time while trying to put a suitcase into the trunk of a car, at another time
while packing his hand luggage for a flight to Paris (for a similar example, cf. Bach
1994: 268). Or suppose a speaker says:
Everyone came to the party. (Stanley 2005: 223224)
at one time while discussing whether all the important people in his department
came to the party, at another time when commenting on how crowded the party
was.
As opposed to sentences such as triangles have three sides or bachelors are
unmarried, our example sentences are context-sensitive in that they can be used to
211
say different things relative to different contexts of utterance; and their truth-conditions differ accordingly. Here is another example.
Jill cant continue. (Borg 2004b: 228)
This sentence can be used to say that Jill cannot continue school, or that she cannot
continue dance classes, or university education, and so on. What it is used to say on
a particular occasion of utterance usually depends on context and can, accordingly,
be figured out by taking context into account. Moreover, whether what is thereby
said is true or false will depend on context, too. Similarly if a speaker says
Tom is tall.
The speaker might have used the sentence to say that Tom is tall for a fifth grader or
that he is tall for an NBA player, etc. Moreover, his utterance may well be true in
the first case but false in the second.7
Accordingly, the contextualist will claim that in order to figure out what has
been said (not what has been implicated in the Gricean sense!) on a particular occasion of their use, most natural language sentences require pragmatic inferences
on the part of the interpreter. Moreover, the contextual clues the interpreter has to
draw on include not only information about the speakers intentions but also information about the purpose or point of the conversation, the participants shared
background assumptions and interests, etc.
Even more interesting from a contextualist point of view are still subtler forms
of context sensitivity. Many natural-language expressions can be used to depict
slightly different actions or entities in different contexts of their use (owing to the
fact that language is used for classification, and classification itself is context-sensitive). Accordingly, they can be interpreted differently, again depending on the
purpose or point of the conversation, the participants shared background assumptions, etc. That is just to say that there is ample polysemy in natural language,
where [p]olysemy is understood here in a broad sense as variation in the construal
of a word on different occasions of use (Croft and Cruse 2004: 109). Compare the
following utterances:
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
Jane opened the window

Bill opened his mouth.
Sally opened her book to page 56.
Mike opened his briefcase.
Pat opened the curtains.
The child opened the package.
The carpenter opened the wall.
The surgeon opened the wound. (Carston 2002: 325)
To open a wound is not the same thing as to open a book. And a wall is usually
opened not in the way one opens ones mouth. The word open can be used to depict a variety of different actions; and an interpreter has to contextually work out,
i.e., pragmatically infer the appropriate understanding. There may be default inter-
212
pretations, though, depending on the object of the verb. But these preferred interpretations are not inevitable. Also, one might think that in all these cases there is
something like a semantic core meaning. In the above example, manipulate such
that the inside/interior/thing underneath can be seen might be a promising candidate. But then the question arises of whether open is supposed to be synonymous
with manipulate such that the inside/interior/thing underneath can be seen. If it is
not supposed to be synonymous, the meaning of open differs from its core meaning, a puzzling result. Yet if it is supposed to be synonymous then one might
wonder what the point of exact synonyms could possibly be in natural language?
(And let us put aside the problem that giving a synonym is not the same as giving
the meaning of an expression, for one might know that a and b are synonymous, without knowing the meaning of either a or b.) Moreover, new work on
language acquisition seems to show that children, when learning a language, operate on the assumption that every difference in form (of linguistic expressions)
marks a difference in meaning (cf., e.g., Clark 1993: 64; Tomasello 2003: 7275ff;
Bloom 2000: 65). Otherwise language acquisition would be a mystery. So the idea
of exact synonyms and, consequently, of core meanings in the case of polysemy is
rather problematic.
Whether a sentence containing a polysemic expression is correctly applicable
to a particular situation depends on the participants interests and purposes, their
background assumptions and interests, etc. Given such and such purposes, interests, etc, it may be correctly applied, given different purposes, interests, etc., it may
be not correctly applied. Yet couldnt we specify further and try to make some of
the purposes, interests or common assumptions explicit? The problem is, as Anne
Bezuidenhout has pointed out, that
[t]here is no sentence that we can produce that can settle all questions about how some
original sentence is to be understood, since language doesnt function that way. It is not
self-interpreting. (Bezuidenhout 2002: 113)
In performing pragmatic inferences, one may have to draw on contextual information concerning the circumstances of utterance, speaker intentions (to the extent that they are manifest), the participants interests and background assumptions, the purpose or point of the conversation, and so on. More generally, we
interpret someones utterance against the background of a shared system of knowledge and a common view of how nature works and how our culture works
(Searle 1980: 226227).
This idea has been elaborated, among others, by Charles Travis (e.g., 1985,
1996, 1997), Julius Moravcsik (1998), and Anne Bezuidenhout (e.g, 2002), to
name just a few. Moravcsik, for example, takes polysemy to affect almost all natural language expressions (cf. Moravcsik 1998: 33).8 Here is an example of his to illustrate the point:
213
The verb walk admits of polysemy. On the one hand, there is a common meaning core,
locomotion with legs in appropriate position. But one has to add: appropriate distance,
covered in an appropriate time. For what counts as a walk for a toddlers first attempts
does not count as a walk for a normal, healthy adult, and the walk of a recovering patient
in a hospital is still a different matter. What counts as a walk depends on different senses
of the word. (Moravcsik 1998: 3536)
How the verb walk is to be understood, i.e., what counts as a walk, depends on
context, more specifically on the various ways we can and do interact with the
world, on our purposes, intentions, etc.: Our understanding of words is influenced
by the variety of ways in which we relate to the world. (Moravcsik 1998: 91) And
there is no limit to the different applications a given word may enjoy; there is no
way of telling in advance what the different senses of a word are that we need it to
have in order to be able to talk about everything we want to talk about (cf. Moravcsik 1998: 37). There are as many different uses to which we may put the expressions of our language as there are different purposes to which we may interact
with the world. The meanings of our words provide rough and ready guidelines for
their own future applications, in the light of new interests, purposes, assumptions
etc. Yet, as Moravcsik emphasizes, we cannot predict or exactly specify all past
and future human interactions with nature that have or will affect our uses of the
words of our language.
To echo John Austin, the question is: Why do we call different things by the
same name? (Austin 31979a: 69) A sensible answer seems to be: It allows us to use
a finite vocabulary to talk about or describe an (in principle) infinite array of situations. Jonathan Cohen puts the point thus:
Rather, it is an enormous convenience that the same word can often be uttered in one or
other of several different though related senses. Instead of having to learn a very much
larger number of words, each with fixed and context-independent meanings, we can
learn a relatively small number of words with variable meanings and then exploit their
verbal or situational contexts of utterance in order to disambiguate their actual occurrences. (Cohen 1985: 132)
Our expressions are flexible enough to fit all the various situations we encounter almost perfectly. Jon Barwise and John Perry speak of the efficiency of language
(Barwise and Perry 1983: 5; cf. also Goodman 21976: 80). The expressions of our
language answer to our human needs, they serve our practical purposes. They can be
contextually adjusted, if need be. And need there is as we never encounter exactly
the same situation twice over. And when we encounter a situation we have not encountered before, our expressions allow us to conceptually structure it by using familiar vocabulary. Sometimes, though, we have to unduly stretch the meaning of a
given expression so that it will become applicable to the case at hand. That yields
polysemy or metaphor, depending on the extent to which the new situation differs
from old ones. Yet metaphor, just as polysemy, seems to defy formal treatment. Both
phenomena seem to require pragmatic inferences for their interpretation.
214
Metaphor in particular is commonly used to conceptualize abstract and mental

phenomena, often with the help of vocabulary familiar from the realm of sense experience. Metaphor is, as Nelson Goodman nicely put it, a matter of teaching an
old word new tricks. (Goodman 21976: 69) That is why we are feeling blue, or
complain about someone being cold, and so on. We also use metaphors in the explanation of human behavior. We say that he broke down under pressure thereby
exploiting the metaphor the mind is a brittle object (cf. Lakoff and Johnson 1980:
28). We say that she is feeling up, that her spirits rose, or that she sank into a coma.
With the help of these orientational metaphors (as Lakoff and Johnson 1980: ch.
4 call them), we conceptualize mental phenomena by giving them a spatial orientation. But abstract phenomena too are commonly conceptualized with the help of
metaphorical expressions. We construe theories and arguments as buildings: The
theory needs more support, it lacks a foundation; the argument collapsed, it was not
solid enough (cf. Lakoff and Johnson 1980: ch. 10). Or we think of love as if it
were a journey: look how far we have come; yet now we are at a crossroad; from
here on we have to go separate ways as this isnt going anywhere; we are in a deadend street. We commonly use expressions from the concrete realm of sense experience to conceptualize abstract and mental phenomena (cf. also Keller and Kirschbaum 2003: 36 and 99). So if one were to say, as formal semanticists with their
fondness for a clear-cut distinction between literal and figurative meaning would
be inclined to do, that metaphorical utterances are always strictly speaking false,
that would render most of our talk about abstract or mental phenomena false. Polysemy and metaphor are central features of natural languages without which we
would not be able to structure the multifariousness of our experience and to conceptually copy with mental and abstract phenomena.
Some philosophers take these and similar considerations to show that word
meanings (or concepts) are not ready-made entities but have to be constructed in
context. According to Robyn Carston, for instance, often in linguistic interpretation an ad hoc concept is constructed and functions as a constituent of what is
explicitly communicated. (Carston 2002: 357) And, presumably, pragmatic inferences are required in the construction of these ad hoc concepts. But one might even
go one step further and try to do without something like linguistic meaning (as traditionally understood) altogether (cf., e.g., Recanati 2004: 146ff; Croft and Cruse
2004: 97ff). The idea would be that although words encode semantic information,
the information itself does not amount to something like a full-fledged meaning.
Rather, it provides the starting point for pragmatic processes, which result in a contextually specified interpretation of the expressions in question. Linguistic meaning would provide information that constrains interpretation, while pragmatic
processes are required to yield the contextually relevant contribution the expression makes on a particular occasion of its use. According to this conception of
linguistic interpretation, the notion of linguistic meaning itself needs to be pragmatically spelled out.
1.4.
215
Conclusion
The preceding considerations cast doubt on the idea of a functional, systematic pairing of sentences with meanings (truth-conditions) as propagated by formal semantic
approaches to language. This way of approaching language seems to ignore the
highly context-sensitive nature of linguistic interpretation. Many natural language
expressions require for their proper interpretation that context be taken into account.
This kind of context-sensitivity rests on a conception of context according to which
context provides information about the participants shared interests, intentions, purposes, background assumptions and so on. The information is extracted by pragmatically inferring the contextually appropriate interpretation of a given context-sensitive utterance. Pragmatic inferences are indispensable in guring out what a given
expression contributes to the truth-conditions of a sentence containing it (i.e., to the
proposition expressed). Accordingly, the value assigned to the expression hardly deserves to be called a (purely) semantic value. And the process of interpretation performed hardly deserves to be called (purely) semantic interpretation. Rather, pragmatic processes seem to intervene at (almost) all stages of linguistic interpretation.
And maybe even the notion of linguistic meaning itself is a pragmatic notion.
Pragmatic semantics
2.1.
The center of pragmatic semantics
Just cancel the second word (maybe) of the last sentence in the preceding paragraph, and you have already reached the very center of Pragmatic Semantics (PS).
Whereas traditional semantics (including formal truth-conditional semantics as its
most developed form) starts with linguistic expressions and then asks how to define the various kinds of their respective linguistic meanings, PS starts just the
other way round, namely with the understanding (i.e., the knowing of the meanings) of actions. From the point of view of PS, semantics is a special part of pragmatics (Action Theory), in the ideal case therefore a part of Action Logic.
Throughout, we will have to make use of the following two-fold distinction
Type
Token
Action
(3)
(4)
Action-Product
(1)
(2)
or of its already language-related specification, respectively:

Utterance
Type
Token
Utterance-Product
Expression (form)
(Concrete) Expression
216
Whereas traditional semantics starts from the right side, (1) above, PS starts from
and is primarily focussed on the left one.
Correspondingly, pragmatics is covering quite differing programs. Whereas,
as exemplified in chapter 1 above, according to traditional semantics it may mean
anything which goes beyond field (1), the program of Pragmatic Semantics as
sketched here is to define in the end even (1)-meanings themselves in terms of action-theoretical concepts.
In PS actions (and their meanings) come first. For this semantic paradigm,
there even need not be any expressions to start from. Actions, think of gestures, for
example, do not need to be product-producing actions (utterances of expressions)
in order to have meanings. Thus, in this approach expressions themselves may
be recognized or introduced as what they are: constructions on the basis of (some
continuity in) the meanings of the corresponding actions. Thus, PS is primarily a
Semantics of Actions.
2.2.
Rule-based versus intention-based pragmatic semantics
In Pragmatic Semantics we also find the division between nominalists and Platonists, depending on whether type-meanings (meanings of action-types, field (3) in
the schema above) can themselves be defined in terms of token-meanings ((4) in
our schema) or not. Since type-meanings (of the PS-relevant sort) are social facts
(i.e., state of affairs which, relative to population P, are facts iff it is common believe in P that they are facts), whereas the respective meanings of action-tokens are
just meanings for some particular subject, this (Platonist versus nominalist) division is a special case of the more general division within Social Philosophy between collectivist versus individualist approaches, according to whether or not
meanings as social facts are taken to be conceptually inexplicable with reference to
individualistic terms.
This distinction is very often identified with a distinction between the two possible readings of Wittgensteins basic equation of meaning = use, 9 which one may
refer to as the rule-based reading versus the intention-based reading. Whereas according to the rule-based reading we have, in order to understand (i.e. to know the
meaning of) a sign A, just to know the rules we have to follow in order to use A
correctly, the intention-based reading is referring to what in a particular situation A
is being used for. Which alternative will be the most fruitful one? This has been the
most debated question in the whole field of pragmatic semantics up to now (see
2.52.7 below).
2.3.
Intention-based semantics
This overview will start from Intention-Based Semantics, which is definitely following a strictly nominalistic / individualistic strategy by embedding PS into gen-
217
eral individualistic Action Theory including Rational Decision Theory as its most
important part (see Jeffrey 1983). The positions of the main collectivist alternatives can then easily be located by referring to their denials of and their alternatives to the respective aspects of Intention-Based Semantics.
For Intention-Based Semantics, as sketched by Herbert P. Grice (1957) and
further developed mainly by Stephen Schiffer (1972), Jonathan Bennett (1976),
David Lewis (1969), Franz von Kutschera (1975) and Georg Meggle (2010),
human languages are primarily used for interpersonal communication. Accordingly, linguistic meanings as social facts are primarily defined in terms of the communicative regularities or conventions holding in the respective social groups
(populations). But as the meanings of communicative action tokens may and often
do diverge from the meanings of the exemplified action type, communicative actions have first to be defined more generally, i.e. on the token level. Thus, the Intention-Based Semantics-program (even when restricted to the action side, i.e., not
yet including some expressions as their action-products) has to involve two steps:
(i) explication of general concepts of communicative acts (as tokens); and (ii) explication of (meanings of) communicative action types by means of (i). Whereas
step (ii) is at the very center of Intention-Based Semantics, (i) is its necessary preliminary stage defining communicative acts in terms of their respective communicative intentions.10
2.3.1.
General communication theory
Communicative acts are instrumental acts, with the most basic types of communicative acts being either directives or informatives, depending on whether Ss (the
speakers) primary communicative aim is to get H (the hearer) either to do or to believe something. But notice that this way of conveniently speaking of speaker
and hearer does not imply that the respective communicative acts are restricted
to acoustic utterances. The differentia specifica of communicative acts versus instrumental acts in general is that communicative acts are acts which, at least from
the point of view of S, will be successful if and only if they will be understood by
their addressee H. This entails the so called reflexivity-condition of communication: For an act to be a communicative one, its actor S has to intend it to be
understood by H (as being a communicative act). A General Communication Theory has to explicate these postulates in a non-circular way.
For this explication, one best starts from the more general communicative
acts: Instrumental acts are acts done for some end (or, to put it differently, done
with specific intentions). More precisely (but already strongly adapted to the
special cases of decisions under certainty, i.e., cases where the deciding person
expects the results of her actions with a maximal degree of probability), person X
is doing f with the intention of bringing it about that the state of affairs A will be
the case: (1) X does f, (2) X wants that A, and (3) X believes (expects) that A
218
will be brought about as and only as a consequence of her doing f. Instrumental

acts will thus be successful iff the belief (3), let us call it the expectation of success, turns out to be true, i.e., if A will in fact come about in the way expected
by X, i.e., as and only as a consequence of Xs doing f. And you understand this
act iff you know its ends, i.e., iff you know the intentions with which f is or was
being done by X.11
As there are no restrictions as to what ends one may try to achieve by means
of communication, we have to distinguish between the intentions with which S is
performing her communicative act (i.e. her communication related intentions) in
general versus her acts communicative intentions in particular, this latter term
referring only to those intentions that are necessary for an act to be a (respective)
communicative act.
Unlike instrumental acts, communicative acts do not only involve the expectation of success, but also the expectation of understanding (i.e., the expectation
that H will understand Ss communicative act), both expectations connected in the
differentia specifica way already mentioned above: that S expects that success will
be brought about as and only as a consequence of understanding. Thus, for the
communicative act to be successful this expectation must turn out to be true, i.e.
communicative success must in fact come about in the way expected by S, i.e., as
and only as a consequence of Hs understanding.
The most crucial and contested question to be settled is: What do we have to
know in order to understand Ss communicative act? To answer this question, one
has to notice that the Reflexivity-Condition (RC) above already implies what we
might call the absolute openness of communication. Whatever condition has to be
satisfied in order for an act to be a communicative one (lets say condition ), it follows from RC that this very condition itself has to be intended to be recognized
by H as being satisfied as well, which is itself a necessary condition for the act to be
a communicative one, etc. Logically, this openness implied by RC will have no
upper end. For such an unrestricted (and in this sense absolute) openness lets
write openness* for short. From RC we get straight to O*:
O*
Communicative intentions must be open* ones.
On the grounds of this condition, Grices first proposal that from Ss point of view
only Hs recognition of Ss primary communicative intention is necessary, can immediately be seen to be inadequate. Grices communications are not open enough.
But Grices Basic Model (BM) can easily be improved:12
BM
Ss doing is an attempt at communication, addressed to H, and of the content that H is to do r, iff (i) S intends, by doing the act, to make H do r; and
(ii) S believes (expects) that she will achieve this aim iff she is letting H
openly* know that S has this intention (i)
219
Communication is thus regarded as proposed by Grice plus the factor of openness*. This kind of openness*, however, relates only to communicative intentions,
i.e., to those intentions which are necessary conditions for an act to be a communicative act. Sincerity is not a necessary condition for an act to be a communicative
one; communicative openness (being a case of openness*) is therefore absolutely
compatible with the insincerity of the communicative act in question (including
downright lies).
This concept of communicative acts (as defined in BM) is a very broad one.
Though it involves the expectations of understanding and of success, it implies nothing about the reasons these expectations may or will have to be based on in order
to be (rationally to be expected by S to be) satisfied. This is fine as far as communicative acts (and their subjective meanings) in general are concerned. But things
change when we come to the special cases of communication by means of (types
of) actions with some intersubjective communicative meaning.
2.3.2.
Meanings of action types
Meaning and understanding (i.e. knowledge and recognition of meaning) are two
sides of one and the same coin. This holds both for subjective and intersubjective
meanings. (Inter-) subjective meanings of (the type f of) an action are relative to
the (class of) persons who understand the actions (of type f) as having these (inter-)
subjective meanings. In addition, meanings of both kinds are relative to (types of)
situations.
The fact, that some gesture (of type) f, lets say thumbs up, as used in the
population P in situations of type , means p (e.g.: you are fine) may be
equated roughly with the following regular meaning condition of f in P relative to
: Whenever (or when in sufficiently many cases) in a -situation a S (as a member
of P) addresses H (as another member of P), making a gesture (of type) f, it holds
that by doing f, S is trying to communicate to H that p (You are fine). Since
(type-) meanings (as social facts) have to be known to be (intersubjective) meanings, the above regular-meaning-definiens will have to be postulated as being common knowledge in P (regular meaning in a narrow sense of f in P relative to ).13
Regular meanings of f (in P for ) are, in -situations and for all members of P, the
best possible reasons for the respective Ss expectation of understanding.
Conventional meanings result from the likewise rationally founded expectation
of success: H will believe what S is trying to make her believe by this very action f
(namely that p) iff H will understand Ss communicative action f. Normally this
will hold only if (S expects that) H believes that, as far as p is concerned, neither S
will be deceptive to her (H) nor that S himself will be at fault.
And these expectations will be stable (within P and relative to -situations) in
the long run only if they will always (or in most or at least in sufficiently many
cases) turn out to be satisfied. And this again will be true only if both S and H will
220
be following a joint strategy, which in turn will be stable only if this strategy is
based on joint interests, as for example, that H is to believe p iff it is in fact the case
that p. Iff, in P and relative to -situations, the communicative expectations of success are based in this way, then the regular meanings of the respective communicative actions will also be conventional ones. This explication is roughly the same
as David Lewiss explication of signalling-conventions (in Lewis 1969).
Now, and this is the decisive step from actions to expressions as their products,
the meanings of (complete or unstructured) signs or expressions (you may think of
the gesture thumbs up again) can be equated with the conventional meanings of
the communicative acts of producing these expressions.
2.3.3.
Linguistic meaning
Unfortunately, when turning to languages and linguistic meanings of structured

signs, Lewis opted out of the whole Intention-Based Semantics-program by putting his proposals on the shoulders of traditional realistic semantics. How this predicament may be avoided in relation to a simple language (of first order predicate
logic) is shown in Kutschera (1975), who provides us with an illustration of the
schema according to which a grammar can be constructed. For a discussion of
such a pragmatic reconstruction see Bennett (1976) and Meggle (2010), the former
trying to couch it in more behavioristic terms, the latter in more formal terms. Following this approach, the meanings of so called realistic semantics (even in its
most developed form of intensional semantics) can be shown to be successfully explicable in terms of Intention-Based Semantics. But, as in many other places, explicability does not imply reduction.
2.4.
Classical speech act theory
Though rightly looked at as being a paradigm case of pragmatics, (classical)

speech act theory was definitely not (intended to be) a special case of PS. This was
clear from the start because John L. Austin (1962), in his basic distinction between
locutionary, illocutionary and perlocutionary (aspects of) acts, was referring to the
use of a sentence in its literal meaning in the sense of Frege. From Austins point
of view, speech act theory should only be interested in the illocutionary functions
of speech acts. This view is shared by his most influential interpreter, John R.
Searle. In this sense, Searles slogan that a theory of language is part of a theory of
action (Searle 1969: 17) can only be misleading.
As to the class of illocutionary acts, Austin did not give us an explicit definition, only an open list of examples. But there are two elements in his characterizations of illocutions which obviously contradict the explication of communicative
acts (in the wide sense of directives and informatives) as sketched in 2.3.1 above.
According to Austin, illocutionary acts are (i) essentially conventional acts; and, as
221
such, (ii) they have to exclude any perlocutionarity.14 Whereas (i) may be
weakened (cf. Bach and Harnish 1979) by distinguishing between essentially conventional illocutionary acts (including Austins favorites of marrying or baptizing
someone) on the one side, and mere communicative actions on the other side, the
non-perlocutionarity of illocutions remained the central dogma of classical
speech-act-theory (even with regard to communicative acts).
There is some truth in this dogma, but this truth can be easily reduced to the distinction between communication related intentions in general (i.e. perlocutionary
intentions in a broad sense) versus communicative intentions in particular (i.e. perlocutionary intentions in the narrow sense) (cf. 2.3.1 above). With regard to the
broad sense of perlocutionary, the non-perlocutionary dogma is (by definition)
true; but in this sense it is (as implied in 2.3.1) no objection to the presented definition of (the communicative variant of) illocution at all.
2.5.
Speech act based PS
It was William P. Alston (1964) who first tried to build a PS on speech act theory.
His basic idea was that two (unstructured) expressions and can be said to have
the same meaning iff they have the same illocutionary potential, i.e., could be used
to perform the same illocutionary acts, where the same illocutionary acts was
meant to refer not only to acts of the same illocutionary types (assertions, directives, questions etc.) but also to these acts as having also the same contents (assertions that p, directives to bring about q, questions as to whether it is the case etc.).
In other words: Illocutionary potential in Alstons sense is already a content-determined illocutionary act potential (CDIAP). And though his basic idea was presented by him only in the form of an identity criterion for the meanings of and , it
was clear that Alston wanted to identify and thus also define the meanings of expressions by means of their CDIAPs. That was the essential difference: Whereas in
Austins theory meanings of expressions are presupposed as being given, Alstons
idea was to define these expression meanings themselves by means of the meanings of illocutionary actions. If this is the line along which meaning should be
analyzed, then the concept of an illocutionary act is the most fundamental concept
in semantics and, hence in the philosophy of language (Alston 1964: 37 [emphasis
mine]).
If we substitute a communicative act for an illocutionary act, Alstons
speech act approach would seem to coincide with the Intention-Bases Semantics as
sketched in 2.3 above. But notice that even Alston in 1964 subscribed to the two
dogmas of classical speech act theory. However, nearly 40 years later, he has extended his basic idea of meanings as CDIAPs into a full blown PS of its own (Alston 2000). Beside Alston, the most detailed and best developed rule-oriented PS is
Eike von Savignys theory (1983). Whereas the intention-based PS as sketched
above starts with the communicative meaning of concrete utterances (action-
222
tokens, i.e. field (4) of the schema in 2.1 above), Savignys rule-based PS starts
with the conventional meaning of utterance types (action-types, i.e.field (3) in the
same schema). According to this approach, differences in meaning are correlated
with differences in the conventional commitments of the persons involved in the
respective utterance-situation. And these commitments in turn are defined by
means of rules as explicated by Hart (1961) with reference to regularities, sanctions and acceptance of these sanctions. Finally, sentence meanings are to be introduced as theoretical constructs in order to systematically explain the salient structures to be found in these regularities.
To systematically compare the IB- versus these and other Rule-Based Approaches to PS with regards to their various pros and cons is one of the biggest desiderata in the whole field of modern philosophy of language.
2.6.
Variety of language games
As far as Wittgensteins meaning = use equation is concerned (see 2.2 above),

the usual distinction between the rule-reading versus the intention-reading of it
should not be the last word. At least not in the sense of an either-or distinction because there is some chance that both readings need not be really exclusive. First, as
noticed in 2.6, the two readings may refer to different explication-levels, the intention-reading to the token-one, the rule-reading to the type-one. Next, as noticed in
2.4, only a special class of perlocutionary intentions is relevant for the definition of
communicative/illocutionary acts: communicative intentions (see 2.3.1). As these
intentions are necessary conditions for some action to be a communicative action
(a directive, for example), they may be looked at as being part of the internal relation, which, according to Wittgenstein ( 21), is constitutive of the kind of
meaning his various language games are meant to be exemplifications of. According to this reconstruction of Wittgensteins meaning as use theory,15 the relevant intention-reading is part of the rule-reading itself.
But there are many other language games (many other illocutions) whose internal relations do not include any reference to perlocutionary intentions. One
might therefore say that meanings (illocutions) do not per se entail perlocutions.
But from this it does not follow that perlocutions (as part of meanings) are per se
excluded. There is more than just one kind of meaning.
2.7.
Evaluations
This difference is of the utmost importance for any evaluation of the various PSapproaches.
What kinds of illocution are the most important ones? The IB-semanticists (as
sketched in 2.3 above) take it that interpersonal communication is at the very basis.
But they do not deny, of course, that languages may have many other functions as
223
well. Some neo-Griceans, interpreting Grices original starting-point of speakers

meaning as covering much more than communicative intentions, do accordingly
extend this basis and include, e.g., the intention to express ones thoughts or feelings (Davis 2003). Others, like Austin, think it best to start with fully institutionalized or even ritualized kinds of performances. Finally, there are scholars like Brandom (1994), who take the whole PS-project (of starting with primitive acts or
intentions and then constructing something like linguistic meanings out of them) to
be on the wrong track because they claim that language has to be presupposed before attributing even these most primitive elements.
Before we postulate the one PS which embraces all sorts of things language
may be used for, we should keep in mind that we just cannot come up with the complete theory of language and still hope for its consistency.
Notes
1. For an overview of the development of semiotics, see e.g., Scholz 2003, for semiotics
and pragmatics Nth, this volume, and for a general historiographic overview of pragmatics Koyama, this volume.
2. For an account of the relationship between semantics and pragmatics cf. Saeed, this volume.
3. Kaplan also speaks of the character of indexical expressions as of a rule that fully determine[s] the referent for each context. (1989: 491)
4. For this specific conception of context cf. Lewis A context is a location time, place,
and possible world where a sentence is said. (1980: 21), and Kaplan, who adds the
speaker or acting participant as a further constituent, argues that an index representing a
context is a 4-tuple <w, x, p, t>, consisting of a world (w), an actor (x), a place (p) and a
time (t) such that x is in w at t in p (Kaplan 1989: 509).
5. Stalnakers notion of context might be promising here: So I propose to identify a context (at a particular point in a discourse) with the body of information that is presumed,
at that point, to be common to the participants in the discourse. (1998: 98)
6. For more on inferences to the best explanation in general, see, e.g., Lipton (1991).
7. Some context-sensitive sentences may have preferred or default completions, though. In
the absence of any special stage setting, we will interpret the sentence: Sams daughter
got married and became pregnant, for example, as meaning that Sams daughter first
married and then became pregnant. For more on the notion of a preferred or default interpretation see Levinson (2000).
8. See also Searle (1978: 220221) and (1980: 221).
9. Which may be treated as something like the cornerstone of the whole of PS.
10. For formal explications of step (i) see Meggle (1997) and for step (ii) Meggle (2010).
11. Following Max Weber one could also say that Action Understanding in this sense is to
know the actions subjective meaning / subjektiver Sinn.
12. With (BM) being already a reconstructed version of Grices own version (1957): S
meant something by x is (roughly) equivalent to S intended the utterance of x to produce some effect in an audience by means of the recognition of this intention.
224
13. Where Common Knowledge in P that p may be easily defined this way: (i) All members
of P know that p; (ii) All members of P know that (i) etc., with this etc to be defined
in formal terms as usual by recursion.
14. See Austin (1962: ch. 9), where perlocutionary effects are by definition restricted to effects going beyond (conventional) illocutionary effects.
15. Very forcefully proposed for the first time by Feyerabend (1955) and heavily relying on
Wittgensteins famous language game, cf. 2 in his Philosophical Investigations.
References
Alston, William P.
1964
Philosophy of Language. Englewood Cliffs, NJ: Prentice-Hall.
Alston, William P.
2000
Illocutionary Acts and Sentence Meaning. Ithaca and London: Cornell University Press.
Austin, John L.
1962
How to do things with Words. Oxford: Oxford University Press.
Austin, John L.
31979a
The Meaning of a Word. In: John Austin, Philosophical Papers, 5575. Oxford: Oxford University Press.
Bach, Kent and Robert M. Harnish
1979
Linguistic Communication and Speech Acts. Cambridge, MA: Harvard University Press.
Bach, Kent
1994
Semantic slack. In: Savas L. Tsohatzidis (ed.), Foundations of speech act
theory: Philosophical and linguistic perspectives, 267291. London/New
York: Routledge.
Bach, Kent and Anne Bezuidenhout
2002
Distinguishing Semantics and Pragmatics. In: Joseph Keim Campbell, Michael
ORourke and David Shier (eds.), Meaning and Truth: Investigations in Philosophical Semantics, 284309. New York/London: Seven Bridges Press.
Barwise, Jon and John Perry
1983
Situations and Attitudes. Cambridge, MA: MIT Press.
Bennett, Jonathan
1976
Linguistic Behaviour. Cambridge: Cambridge University Press.
Bezuidenhout, Anne
2002
Truth-Conditional Pragmatics. Philosophical Perspectives 16: 105134.
Bloom, Paul
2000
How Children Learn the Meanings of Words. Cambridge, MA: MIT Press.
Borg, Emma
2004a
Formal Semantics and Intentional States. Analysis 64: 215223.
Borg, Emma
2004b
Minimal Semantics. Oxford: Oxford University Press.
Borg, Emma
2007
Minimalism versus Contextualism in Semantics. In: Gerhard Preyer and Georg
Peter (eds.), Context-Sensitivity and Semantic Minimalism: New Essays on
Semantics and Pragmatics, 339359. Oxford: Oxford University Press.
225
Brandom, Robert B.
1994
Making it explicit. Reasoning, Representing and Discursive Commitment.
Brandom, Robert, B.
2002
Pragmatics and Pragmatisms. In: James Conant and Urszula M. Zeglen (eds.),
Hilary Putnam: Pragmatism and Realism, 4059. London: Routledge.
Clark, Eve V.
1993
The lexicon in acquisition. Cambridge Textbooks in Linguistics, Cambridge:
Carnap, Rudolf
1956
Meaning and Necessity: A Study in Semantics and Modal Logic. 2nd ed. Chicago: The University of Chicago Press.
Carston, Robyn
2002
Cohen, Jonathan L.
1985
A Problem about Ambiguity in Truth-Theoretical Semantics, Analysis,
Vol. 45/3: 129134.
Croft, William and Cruse, Alan D.
2004
Cognitive Linguistics. Cambridge Textbooks in Linguistics, Cambridge: Cambridge University Press.
Davis, Wayne A.
2003
Meaning, Expression, and Thought. Cambridge: Cambridge University Press.
Feyerabend, Paul
1955
Wittgensteins Philosophical Investigations. Philosophical Review: 449483.
Frege, Gottlob
1891
Funktion und Begriff. Lecture held 1891 at a meeting of the Jenaischen Gesellschaft fr Medizin und Naturwissenschaft. Reprinted in Gottlob Frege
1986, 6th ed., 1839.
Frege, Gottlob
1892a
ber Sinn und Bedeutung. Zeitschrift fr Philosophie und philosophische Kritik 100, 2550. Reprinted in Gottlob Frege 1986, 6th ed., 4065.
Frege, Gottlob
1892b
ber Begriff und Gegenstand. Vierteljahrsschrift fr wissenschaftliche Philosophie XVI, 192205. Reprinted in Gottlob Frege 1986,6th ed., 6680.
Frege Gottlob
1986
Funktion, Begriff, Bedeutung. Edited by Gnther Patzig, 6th ed. Vandenhoeck
und Ruprecht, Gttingen.
Goodman, Nelson
1976
Languages of Art. An Approach to a Theory of Symbols. 2nd ed. Indianapolis:
Hackett.
Grice, H. Paul
1957
Meaning. The Philosophical Review 66: 377388. Reprinted in Grice 1989.
Grice, H. Paul
1989
Hart, Herbert L.A.
1961
The Concept of Law. Oxford: Clarendon Press.
226
Jeffrey, Richard C.
1983
The Logic of Decision. 2nd ed. Chicago: University of Chicago Press.
Kaplan, David 1989
Demonstratives. An Essay on the Semantics, Logic, Metaphysics, and Epistemology of
Demonstratives and Other Indexicals. In: Joseph Almog, John Perry, and Howard Wettstein (eds.), Themes from Kaplan, 481563. Oxford: Oxford University Press.
Keller, Rudi and Ilja Kirschbaum
2003
Bedeutungswandel: Eine Einfhrung. Berlin: de Gruyter.
Koons, Robert
2009
Defeasible Reasoning. In: Edward N. Zalta (ed.), The Stanford Encyclopedia
of Philosophy (Winter 2009 Edition). <http://plato.stanford.edu/archives/
win2009/entries/reasoning-defeasible/>.
Kutschera, Franz von
1975
Conventions of Language and Intensional Semantics. Theoretical Linguistics
Vol 2/3: 255283.
1976
Einfhrung in die intensionale Semantik. Berlin: de Gruyter.
1980
Metaphors We Live By. Chicago: University of Chicago Press.
2000
Lewis, David
1969
Convention. Cambridge, MA: Harvard University Press.
Lewis, David
1975
Languages and Language. In: Keith Gunderson (ed.), Minnesota Studies in the
Philosophy of Science, Vol. VII, 335. University of Minnesota Press. Reprinted in David Lewis 1983: Philosophical Papers. Volume 1, 163188. Oxford: Oxford University Press.
Lewis, David
1980
Index, Context, and Content, in Stig Kangar and Sven hman (eds.), Philosophy and Grammar, 79100. Dordrecht: Reidel.
Lipton, Peter
1991
Inference to the Best Explanation. London: Routledge.
Meggle, Georg
1997
Grundbegriffe der Kommunikation. Berlin: de Gruyter.
Meggle, Georg
2010
Handlungstheoretische Semantik. (Habilitation thesis 1984). Berlin: de
Gruyter.
Montague, Richard
1970
English as a Formal Language. In: Bruno Vicentini et al. (eds.), Linguaggi
nella societ e nella tecnica, 189224. Mailand: Edizioni di Comunit. Reprinted in Montague 1974, 188221.
Montague, Richard
1970b
Universal grammar. Theoria 36: 373398. Reprinted in Montague 1974,
222246.
227
Montague, Richard
1973
The proper treatment of quantification in ordinary English. In: Jaakko J. Hintikka, Julius M. Moravcsik and Patrick Suppes (eds.), Approaches to natural Language, 221242. Dordrecht: Reidel. Reprinted in Montague 1974,
1734.
Montague, Richard
1974
Formal Philosophy: Selected Papers of Richard Montague. Edited by Richmond H. Thomason. New Haven/London: Yale University Press.
Moravcsik, Julius M.
1998
Meaning, Creativity, and the Partial Inscrutabilty of the Human Mind. CSLI
Publications, Stanford.
Manor, Ruth
2001
On the Overlap of Pragmatics and Semantics. Synthese 128: 6373.
Morris, Charles
1938
Charles Morris (eds.), International Encyclopedia of Unified Science. Vol. 1,
Nr.2, 77138. Chicago: University of Chicago Press. Reprinted in Charles
Morris, 1971, Writings on the General Theory of Signs, 1771. Mouton: The
Hague/Paris, quotations are to Morris 1971.
Partee, Barbara
1973
Some transformational extensions of Montague grammar. Journal of Philosophical Logic 2: 509534.
Partee, Barbara
1975
Montague grammar and transformational grammar. Linguistic Inquiry 6/2:
203300.
Portner, Paul and Barbara H. Partee
2002
Introduction. In: Paul Portner and Barbara H. Partee (eds.), Formal Semantics:
The Essential Reading, 116. Oxford: Blackwell.
Recanati, Franois
2004
Literal Meaning. Cambridge; Cambridge University Press.
Savigny, Eike von
1983
Zum Begriff der Sprache. Konvention, Bedeutung, Zeichen. Stuttgart: Reclam.
Schiffer, Stephen
1972
Meaning. Oxford: Oxford University Press.
Scholz, Oliver R.
2003
Semiotik und Hermeneutik. In: Roland Posner, Klaus Robering and Thomas A.
Sebeok (eds.), Semiotik/Semiotics: Ein Handbuch zu den zeichentheoretischen
Grundlagen von Natur und Kultur / A Handbook on the Sign-Theoretic Foundations of Nature and Culture, Vol. 3, 25112561. Berlin: de Gruyter.
Searle, John R.
1969
Searle, John R.
1978
Literal meaning. Erkenntnis 13: 20724.
Searle, John R.
1980
Background of Meaning. In: John Searle, Ferenc Kiefer, Manfred Bierwisch
(eds.), Speech Act Theory and Pragmatics, 221232. Dordrecht: Reidel.
228
Stalnaker, Robert C.
1998
On the Representation of Context. Journal of Logic, Language, and Information 7. Reprinted in Stalnaker 1999 Context and Content. 96113. Oxford: Oxford University Press.
Stanley, Jason
2002
Making it Articulated. Mind and Language 17: 149168.
Stanley, Jason
2005
Semantics in Context. In: Gerhard Preyer and Georg Peter (eds.), Contextualism in Philosophy: Knowledge, Meaning, and Truth, 221253. Oxford: Oxford
University Press,
Tomasello, Michael
2003
Constructing a Language: A Usage-Based Theory of Language Acquisition.
Travis, Charles
1985
On what is strictly speaking true. Canadian Journal of Philosophy 15:
187229.
Travis, Charles
1996
Meanings Role in Truth. Mind 105: 451466.
Travis, Charles
1997
Pragmatics. In: Bob Hale and Crispin Wright (eds.), A Companion to the Philosophy of Language, 87107. Oxford: Blackwell.
1953
Philosophische Untersuchungen. Ed. by G.E.M. Anscombe and Rush Rhees.
Oxford: Basil Blackwell.
8.
Foundations of pragmatics in functional

linguistics
Saskia Daalder and Andreas Musolff
1.
Introduction
The relatively young discipline of linguistic pragmatics has had forbears in several
older branches of linguistics. This chapter focuses on important prefigurations of
parts of linguistic pragmatics from the late nineteenth century to around 1970
which were informed by a functional view of language. We view functional linguistics as the study of language that starts its research with a focus on the functions of language in social life: the effects of language use, differentiated as to
types of communicators, types of contexts and types of language uses. This includes the analysis of the primarily intended goals of the speech participants in
their speaking and reacting as well as the study of the long-time effects of language
use in the life of the individual and in the history of the language.
One could name quite a few functionally inclined linguists of that period whose
theories are to some extent congenial to modern pragmatics, starting with the
theorists Dwight Whitney (18271894), Michel Bral (18321915), Georg von
der Gabelentz (18401893) and Henry Sweet (18451912) and ending with, e.g.,
Andr Martinet (19081999) and Michael Halliday (*1925). However, in this article we will confine ourselves to results of functional linguistics that have actually
been transmitted to linguistic pragmatics and may be counted among its foundations. Such results are found foremost in the works of four theorists: the German
psychologist Karl Bhler (18791963), the Czech scholar of English Vilm Mathesius (18821945), the Russian general linguist and scholar of Slavonic languages
Roman Jakobson (18961982) and the British general linguist Michael Halliday.
Of these, we will discuss the relevant theories of the first three. Bhler proposed an
influential model of three fundamental language functions, which Jakobson
came to extend to the model with six such functions that has had a strong impact
on pragmatics, the ethnography of communication and social semiotics.1 Furthermore, both Bhler and Jakobson analysed the functioning of the deictic elements of language, emphasising their role in the situated nature of language
use. Mathesius, who was more specifically concerned with the comparison of living languages as to their possibilities of expression, provided the foundations of
the functional analysis of utterances in terms of the notions theme and rheme
(topiccomment), which has become one of the stock items of pragmatic
analysis. For their ideas in this domain, both Bhler and Mathesius were indebted
to the work of the older and less well-known psychologically oriented German lin-
230
guist Philipp Wegener (18481916), who held views that also show a strong affinity to those of modern linguistic pragmatics. In our sections 25, we will in turn
discuss the contributions by Wegener, Bhler, Mathesius and Jakobson. In the concluding section 6, we consider the questions of whether and how their insights into
the functionality of language can still be of value for linguistic pragmatics. In our
view, these insights still provide valid reasons to integrate the description of elements of the situational context into linguistic analysis itself and thus to question
the wisdom of completely separating research into the system of language from
pragmatics, the study of its use.
2.
Philipp Wegener
Philipp Wegener (18481916), a classics teacher, grammar school headmaster and

psychological linguist from Northern Germany, developed in a relatively small
body of work a remarkably modern, communication-oriented analysis of language.2 It resulted from his opposition against a particular claim of contemporary
historical linguistics. The Neogrammarian theorists had acknowledged the relevance of psychological factors for processes such as the borrowing of linguistic
elements by one language or dialect from another, but not for so-called internal,
primarily phonetic language change, which was thought to be caused purely by
physiological mechanisms. Motivated by his study of developments in the local
dialects, Wegener wished to demonstrate the concrete psychological and communicative background also of internal changes, and this forms the first topic in
his 1885 publication Investigations into the Fundamental Questions on the Use of
Language (Untersuchungen ber die Grundfragen des Sprachlebens).3 Wegener
uses the key concept of automatisation: changed phonetic realisations may appear in hindsight not to fulfil any special goals, but they are in fact the results of
conscious choices by speakers, made at one point in time, of alternative possibilities with special communicative effects, which subsequently have become automatised through habit. A communicative basis can also be seen in the ontogeny of
language. In time, the child learns to use sound patterns with meanings of a general
nature, but this develops only after a stage in which a particular vocal expression is
a purposive action, essentially connected to the childs momentary needs and
meant to appeal to the carers for help (Wegener [1885] 1991: 419).
Wegener goes on to claim that the general purpose of speaking is indeed to influence the actions, volition or thoughts of another person in a way the speaker
thinks worthwhile (Wegener 1991: 6467, 180). He has no use for the well-known
subjectivist picture of the individual speaker who, in speaking, merely mirrors
his own thoughts; such a view would make the idea of communication as accommodation between different minds all but void (Wegener 1902: 402). Focusing
rather on the role of the hearer, Wegener emphasises that the hearer is the true locus
231
of meaning, because he constructs meaning first and foremost not from linguistic
cues but according to his knowledge of the world and of the purposes of social action, helped further by his inferential abilities. Within this dialogical theory of
meaning, the function of the words and grammatical elements the speaker presents
is to help organise the hearers expectations (Wegener 1991: 157158, 180182).
The importance of the asymmetrical interaction between speaker and hearer is
highlighted in the detailed analysis of many grammatical constructions that Wegener provides in his book. As the general functional structure of utterances, Wegener proposes a bipartition consisting of exposition (Exposition) and predicate (Prdicat).4 The predicate carries the core function of the utterance, it
indicates the new and important material for the situation of utterance, and whatever effect is achieved in the hearer is connected primarily to this part. The exposition is everything in the utterance that the speaker has deemed necessary for
the hearer to situate the information contained in the predicate in the way it is
meant to be situated (Wegener 1911; 1991: 1922).
Elements of the situation itself may provide enough information to let the
predicate do its work, as Wegener explains. Any aspects of the hearers momentary
attention, such as a context of actions that have just been carried out (or are in the
process of being carried out), a topic from the social or institutional situation at
hand or an element from the general cultural situation, can be sufficient for him to
understand an unaccompanied predicate Well done as uttered by the speaker.
Language thus need not always be congruous to what the mind understands (Wegener 1991: 2127).5 However, an exposition of linguistic material is often
added to help achieve clarity as to the situated fact in relation to which the predicate has to be understood. Note that according to Wegener, the exposition need not
always precede the predicate; there may be many reasons for a speaker to choose
the reverse ordering. Intonation, but also grammar, can fulfil the task of indicating
to the hearer just what part of the utterance has which role.
Wegener (1991: 4760) also relates important phenomena of word semantics to
his communicative bipartition of utterances.6 It is not the same for a word to appear
in the exposition of an utterance or to appear in the predicate. The words of the exposition are used by the hearer to determine the real-world referent, and their
meanings should therefore be relatively congruous to the relevant thought. By
contrast, words chosen for the predicate may well have a variable, only loosely-fitting relation to their real-world referent and often have a metaphorical character.
For his precise understanding, the hearer can rely on inferences that are drawn
from elements of the situation and from expositional parts of the utterance. When
such inferences become automatised through repeated use, the word meaning loses
its flexibility, increases in its congruousness to thought and thereby becomes appropriate to fulfil a role also in the expositional parts of utterances. To give a modern example: it is often said, figuratively, that the family is the cornerstone of society. The phrase has become so well-known that the original predicate part (the
232
cornerstone of society) can now also be used in a role of exposition and still be
understood to have the family as referent: the cornerstone of society is in trouble.
But this does not yet hold unrestrictedly for cornerstone on its own; of course,
this could well be a further development. The process as sketched amounts to semantic change and, more specifically, the well-attested, continuous fading (abblassen) of metaphors (Wegener 1991: 5157).
It is an interesting fact that in this picture, pragmatic analysis can explain linguistic phenomena that are usually considered to be semantic or grammatical in nature. Pragmatics could thus be conceived of as an overarching discipline encompassing linguistics the full potential of such a conception does not seem to be
exhausted by modern pragmaticists.7
Although Wegeners approach was appreciated and taken into account by such
functionalist theorists as the important Polish anthropologist Bronisaw Malinowski (18841942) and the British linguists Alan Gardiner (18791963) and Michael Halliday,8 it never gained real notoriety in general linguistics and was not
used by the founders of modern linguistic pragmatics. Nor did the revival of Wegeners pioneering ideas in works by the German communication theorists Gerold
Ungeheuer (19301982) and Johann Juchem (19392003)9 succeed in making
them known beyond a rather small circle of pragmaticists. However, some of Wegeners proposals did reach modern pragmatics through an intermediary. Karl
Bhler helped spreading Wegeners analysis of the interweaving, in each utterance, of the speakers expression of thoughts, the intended effect on the hearer and
the representation of a part of reality. This tripartite analysis had a major influence
on the development of Bhlers theory of the functioning of language, as can
be seen already in one of his earliest linguistic publications (Bhler 1909b:
119123).10 It eventually led to Bhlers model of three language functions, the
organon model, to be discussed in section 3.1, which in turn was further differentiated by Roman Jakobson and became part of linguistic pragmatics in that extended version (see section 5.1).
3.
Karl Bhler
3.1.
Bhlers axiomatic view of language functions
In 1934, Karl Bhler (18791963), Ordinarius professor of psychology at the

University of Vienna, published a monograph entitled Theory of Language. The
Representational Function of Language (Sprachtheorie. Die Darstellungsfunktion
der Sprache). Its introductory First Part deals with The Principles of Language
Studies, which include four basic axioms. The first of these axioms, the organon-model, is introduced as the maxim of the three semantic functions (Leitsatz
von den drei Sinnfunktionen) of all linguistic structures (Bhler 1990: 28; 1934:
233
22).11 Bhler cites Platos definition in Kratylos of language as an organum

(means/tool) for a speaker to inform the other of something about the things
(Bhler 1990: 30; 1934: 24) to motivate the term organon-model. It was this
1934 version of the model of language functions that inspired a number of later
scholars to develop the notion of Functional Linguistics with a strong emphasis
on pragmatic as well as sociolinguistic aspects, and it is quoted as such in textbooks to this day.12 Before we look at its systematic theoretical status and further
adaptation, however, we need to analyse its development in Bhlers own work.
1934 did not mark the starting point of Bhlers work on meaning functions
by any means; rather, it was its climax and almost also its ending due to the interruption of Bhlers work in 1938 when Nazi Germany annexed Austria.13 Building on Edmund Husserls (18591938), Anton Martys (18471914) and Wilhelm Wundts (18321920) theories of meaning,14 as well as on his own work on
thought psychology in the context of the Wrzburg school around Oswald Klpe
(18621915),15 Bhler had published an article on the Critical assessment of new
sentence theories (Kritische Musterung der neueren Theorien des Satzes) as early
as 1918, in which he proposed a triadic model of language functions (Leistungen). It is based on the three main foundations of any meaning-constitutive
interaction, i.e.,
1) the speaker whose feelings and attitudes are given expression (Kundgabe),
2) the hearer for whom it provides a stimulus that elicits reactions (Auslsung),
3) objects or states of affairs that form the referents for representation (Darstellung) (Bhler 1918: 16).
After moving to the Chair of Psychology at Vienna University in 1922, Bhler
went on to develop this triadic model further in two major psychological studies. In
The Crisis of Psychology (Die Krise der Psychologie) of 1927, the model serves as
the basis for a critical assessment of contemporary psychology. The two basic semantic functions of expression (Kundgabe, which, as Bhler emphasises, also includes its receptive counterpart, Kundnahme) and guidance (which also involves
mutuality) are present even in animal communication; the ability to produce symbolic representations (of objects/states of affairs), on the other hand, is characteristic of human communication and helps to adapt the first two functions to a higher
level, so that all three of them are only fully activated in human communication
(Bhler 1927: 3233, 37, 5051). The three functions model thus forms a framework for conceiving of psychology as the integrated study of subjective experience, social behaviour and structures of objective sense (Bhler 1927: 29,
5762). The enormous conceptual range of this function-concept is evident also
in the further monograph on Theory of expression (Ausdruckstheorie) of 1933,
where Bhler refers to it as the basis to analyse the relationship between miming,
gesture and language (Bhler 1933a: iiiiv).
234
With specific respect to linguistics, Bhler explicitly claimed axiomatic

status for the functions model in his two articles The whole of language theory, its
system and its parts (Das Ganze der Sprachtheorie, ihr Aufbau und ihre Teile) and
The axiomatics of the language sciences (Die Axiomatik der Sprachwissenschaften), published in 1932 and 1933, respectively. He now called the functions
expression (Ausdruck), appeal (Appell), and representation (Darstellung)
(Bhler 1932: 106; 1933a: 7490) and incorporated them into a set of four principles, which prefigure the axiomatic introductory part of Sprachtheorie (although they appear in different order).16 Compared with the earlier versions of the
three functions-model, both articles share with the Sprachtheorie the detailed discussion and acknowledgement of Ferdinand de Saussures (18571913) Cours de
linguistique gnrale (published posthumously in 1916; German translation published in 1931).
Saussures famous lectures had motivated Bhler to reconceptualise language
theory from a semiological (Saussure), or as Bhler calls it, sematological
viewpoint (1990: 33, 42; 1934: 27, 3435). However, Bhlers interpretation of the
Cours differed in significant ways from contemporary structuralist readings.17
Whilst acknowledging the epoch-making impact of the Cours (Bhler 1990: 78;
1934: 7), he criticises its presentation of the speech circuit as a relapse into 19th
century psycho-physics and he insists that linguistic communication comprises
different types of signs (1990: 3134, 4243; 1934: 2528, 3437). Most important
in our context is his discussion of Saussures distinction of a linguistique de la
langue from a linguistique de la parole. Bhler puts the emphasis not on the langueaspect, but instead praises Saussure for having shown what would have to be discovered in order to be really able to initiate a linguistique de la parole (1990: 8;
1934: 7). This foregrounding of the speech/action aspect of language was at odds
with the predominant reading of the Cours as an endorsement of a system-oriented
core linguistics. It thus provides a first indication of Bhlers interest in the pragmatic orientation as an essential methodological perspective for linguistic research.
Besides the pragmatically interpreted Cours, Bhler cites three further works
as coming closest to providing a functionalist grounding of linguistics as an urgently needed complement to the old grammar (1990: 27; 1934: 23):
a) Philipp Wegeners Investigations into the Fundamental Questions of the
Use of Language ([1885] 1991) (discussed in our section 2) had taught Bhler
about the listeners role in the social action dimension of speech;18
b) Karl Brugmanns (18491919) essay on Indo-European pronoun systems
(cf. 1904) had provided him with a theory of demonstrative deixis as a framework
to analyse the speech situation;
c) Alan Gardiners Theory of Speech and Language (1932), which owed its
publication partly to Bhlers encouragement.19 Bhler praised it as the most interesting attempt to carry out a project similar to his own but adds the reservation
that Gardiner strives exclusively for a situational theory of language, whereas
235
he, Bhler, wants to account also for types of language use that are removed from
the situation (Bhler 1990: 28; 1934: 3435).20
Supported by these three authorities (as well as by the parole-oriented reading
of Saussures Cours), the organon-model appears in the Sprachtheorie as the
first of four axioms; it postulates three semantic functions (semantische Funktionen) for every instance of a linguistic sign (Sprachzeichen):
Es ist Symbol kraft seiner Zuordnung zu Gegenstnden und Sachverhalten, Symptom
(Anzeichen, Indicium) kraft seiner Abhngigkeit vom Sender, dessen Innerlichkeit es
ausdrckt, und Signal kraft seines Appells an den Hrer, dessen ueres und inneres
Verhalten es steuert wie andere Verkehrszeichen. (Bhler 1934: 28).
It is a symbol by virtue of its co-ordination to objects and states of affairs, a symptom
(Anzeichen, indicium: index) by virtue of its dependence on the sender, whose inner
state it expresses, and a signal by virtue of its appeal to the hearer, whose inner and outer
behaviour it directs as do other communicative signs. (Bhler 1990: 35).
It is in this version that the organon-model has entered pragmatic-functionalist

textbooks as a well-known historical reference (see above). However, its presentation in the Sprachtheorie as the first axiom (possibly meant to underline its preeminence) is less convincing than its introduction in the Axiomatics essay from a
year earlier. In that essay it appears as the last axiom and forms the conclusion of
the previous argumentation.21 In the Sprachtheorie, the model is followed first by
illustrative comments on various styles and registers in which either of the three
functional dimensions appears dominant (e.g. poetry, rhetoric, scientific discourse)
and a hint that its main thesis will be verified when all three books [on language]
that the organon model requires have been written (Bhler 1990: 39; 1934: 33).
This hint underlines the point of the subtitle of Sprachtheorie, The Representational Function of Language (Die Darstellungsfunktion der Sprache): this work
does not contain the complete Theory of Language whose necessity is implied
by the organon model; it still focuses mainly on the one function that had been the
object of traditional language philosophy and semantics.
This discrepancy, although openly announced, is regrettable from a pragmatic
point of view, for historical reasons (as the three promised books were never completed: The Crisis of Psychology and Expression Theory contain valuable linguistic
insights but have a much wider, psychologically defined remit) as well as in theoretical respects. As a largely programmatic formulation, the organon model lacks an
explication by Bhler that would have clarified its methodological implications. In
the Sprachtheorie itself, the model is only intermittently and cursorily referred to
and does not provide a systematic structuring principle of the book. The following
remarks therefore highlight aspects of the Sprachtheorie that can be most closely
linked to the three functions model, without claiming to give a systematic overview of Bhlers opus magnum.
Within the Axiomatics-part of the Sprachtheorie, the second axiom is the signcharacter of language, which Bhler takes over not just from Saussure but from
236
the sign-theoretical tradition going back to the medieval scholastic debates on representation and which he relates in a bold move to the contemporary phonemephone distinction. Bhlers formula for the principle of the semantic character of
signs is abstractive relevance (Bhler 1990: 4056; 1934: 3348). Building on
the organon- and sign-axioms, he attempts in axiom III to achieve an integration of
functional and systemic language aspects by way of a cross-classification based on
the parole-langue and ergon-energeia distinctions by Saussure and Humboldt.
This yields four perspectives on language as an object of scientific investigation,
i.e. language as (1) speech action (Sprechhandlung), (2) language work
(Sprachwerk), (3) speech act (Sprechakt) and language structure (Sprachgebilde) (Bhler 1990: 5780; 1934: 4869).
This schema has been criticised for being terminologically and methodologically inconsistent.22 Its English translation in the 1990 edition adds further problems as it suggests false friends such as Sprechakt and speech act. Bhlers
understanding of this aspect is far removed from Austinian, Searlean or Gricean
concepts; rather, it rests on the philosophical notion of sense-conferring acts
(sinnverleihende Akte), which Bhler had adopted from Husserl. Bhler praises
Husserls Logische Untersuchungen for highlighting sense-conferring acts as a
necessary object of language theory but he criticises the implication of an exclusively subject-centred perspective on meaning constitution, against which he insists
on the inter-subjective side of all use of signs (1990: 7374; 1934: 6365; see also
Nerlich and Clarke 1996: 194198, 211215).
The most interesting aspect of Bhlers schema from a pragmatic perspective is
the category of the speech action, which encapsulates the insight that all concrete speech is in vital union with the rest of a persons meaningful behaviour; it is
among actions and is itself an action (Bhler 1990: 61; 1934: 52). In the discussion in the Sprachtheorie Bhler gives only a relatively brief illustration by way
of famous quotations that embody a whole situation (e.g. Caesars alea jacta est)
and refers to developmental psychology; he also highlights the field-character of
every speech action and the importance of the act(ion) history, which includes the
language acquisition process (1990: 6566; 1934: 56). In the Axiomatics essay,
however, Bhler gives a more detailed explanation, drawing strongly on the action
theory developed by the fellow Wrzburg School psychologist Abraham Anton
Grnbaum (18851932), who had given a paper on Language as action (Sprache
als Handlung) at a workshop organised by Bhler at a psychology congress in
Hamburg in 1931 (Bhler 1933a: 5051). For Grnbaum, every action is a dynamic system of organism and environment (System: Organismus Umwelt); its
constitutive aspects are the agentive organisms need (Bedrfnis) and intention (Intention), the active completion (Erledigung) of the action and the agents reaction
to the completion (Erledigungsreaktion) (Grnbaum 1932: 165173). For the application to the speech action, Bhler re-labels the reaction to the completion as
fulfilment (Erfllung) and describes it further as consisting of the speakers con-
237
scious realisation that the process of meaning constitution has been concluded
(Bhler 1933a: 5051).23
In the Sprachtheorie, Grnbaums theory is not explicitly mentioned but
Bhler uses Grnbaums terminology in some parts and refers for further reading
to his own Axiomatics essay as well as to disputes among the schools of Gestalt
theory, where he takes a polemical stance against the supposed monism of the
Berlin School of Gestalt psychology around Max Wertheimer, Wolfgang Khler
and Kurt Koffka (Bhler 1990: 66; 1934: 56). In this allusion-laden discussion the
contours of Bhlers concept of speech as action are sometimes barely discernible.
There is just one hint that the topic is taken up again in chapter 10, which deals
with the integration of speaking into other meaningful behaviour, and more specifically into different types of surrounding fields (Umfelder) (Bhler 1990: 61;
1934: 52). This distance or detour of six chapters is motivated by Bhlers more
immediate focus on the concept of language fields, which overlaps with the notion of surrounding fields, thus leading to a potentially confusing double-terminology.
Within the Axiomatics-part of Sprachtheorie, the field concept comes into
focus in the fourth axiom, which deals with the syntax-semantics interface under
the heading Word and Sentence: the S[ymbol]-F[ield]-system of the type language (Bhler 1990: 8190; 1934: 6978), but the concept also dominates the remainder of the monograph. It serves Bhler to gain an overview of language phenomena that ranges from phonology and morphology via syntax and case theory to
text theory and stylistics. The principle unifying this theoretical and methodological tour de force is the distinction between, on the one hand, those aspects of linguistic signs that receive their precise meaning in the deictic field (Zeigfeld), i.e.
the spatio-temporal surroundings, the relationship of speaker and listener, their
shared knowledge and action goals, and on the other hand, those aspects of linguistic signs that are fully integrated in syntagms and are detachable from the immediate situation, thus forming a symbol field (Symbolfeld).24 As a programmatic outline, this two-field theory provides, as Garvin (1994) has pointed out,
an alternative to any rigid syntax-semantics-pragmatics distinction and has
proved to be a constant source of fresh insights. In terms of a theoretical focus,
however, Bhlers wide-ranging discussion covers too many data and theoretical
aspects and the important insights from the organon model and the speech action concept are taken up only intermittently in the field-theoretical chapters. In
the following section, we highlight these pragmatically relevant theory aspects
without claiming to provide a comprehensive reconstruction of Bhlers general
field-theoretical framework.
238
3.2.
Linguistic deixis and action
The two main concepts in the Sprachtheorie that are of relevance for linguistic
pragmatics are those of the deictic field and the surrounding fields, i.e. the context in which speech is meaningful as social action. Terminologically, both categories are evidently strongly integrated in Bhlers overarching field-theory,
but conceptually they also depend on the three-functions model. In both cases, the
multi-functionality of language and the fundamental importance of the speakerhearer relationship, which derive from the organon model, are constitutive of
Bhlers argumentation.
The origin of linguistic indexicality/deixis lies in the I-now-here-position of
the speaker, which serves as the starting point for spatial, temporal and social
orientation and co-ordination between the communication partners (Bhler 1990:
117136; 1934: 102120). The interlocutors shared horizon can then be extended
through the quasi-figurative use of the co-ordinates of the demonstratio ad oculos
(visual [or other sensory] demonstration) for the purpose of imaginationoriented deixis. When referring to objects that are not immediately given in the
discourse situation, speaker and hearer must either project them linguistically into
that situational context (e.g. by putting a problem before themselves) or, vice
versa, project their own situation co-ordinates onto an imagined context, or use a
mixture of both. Such imagination-oriented deixis is present already in everyday
language use and marks the first step in the gradual emancipation of language from
the immediate situational context (Bhler 1990: 158166; 1934: 140148). The
highest developmental stage of deictic language use is reached when indexical
language signs are used anaphorically and cataphorically as joints of speech, e.g.
as relative pronouns, text-deictic prepositions, prepositional adverbs or conjunctions (Bhler 1990: 438452; 1934: 385397). Here, the symbol-field is reconstructed along deictic co-ordinates; at the same time, deictic signs are used to
demonstrate relationships within a representational context.25
The second main pragmatically interesting discussion in the Sprachtheorie,
which Bhler had already flagged up in the context of the Axiomatics section when
dealing with the speech action, is chapter 10. Its title provides an additional typology of surrounding fields: The sympractical, the symphysical and the synsemantic field of Language Signs (1990: 175; 1934: 154). Again, as in the case of
the deixis-symbolising continuum, Bhler conceptualises the field-series as a cline
of abstraction. Linguistic signs can be tied closely to a specific situation or specific
objects (= the sympractical and symphysical fields) and so remain situation-bound.
The more strongly they are integrated into the linguistic con-text (= the synsemantic field) however, the more liberated they are from the specific circumstances
of the speech situation. In some genres, e.g. algebraic formula systems, the liberation from concrete, situation-related content is almost taken to the maximum (cf.
chapters 11, 25 of Sprachtheorie). The overlap between the synsemantic field
239
and the symbol field (Bhler 1990: 179189; 1934: 159168) blurs the boundaries
between the two field-theoretical concepts to some extent,26 which may be indicative of Bhlers difficulty to combine his focus on functional aspects of language with the perceived necessity of integrating these insights into the predominant contemporary tendencies of explaining language mainly as a system of
representation.
In the context of the surrounding fields, Bhler comes closest to a pragmatic
interpretation of language data in the passages dealing with the sympractical
field, which presuppose his three-functions axiom and Grnbaums action field
theory (see above). The sympractical field includes what would be called in modern terminology the situational context (as opposed to co-text), which communication partners need to know (and share) in order to be able to disambiguate
situation-bound utterances such as short commands, requests, greetings etc. Such
utterances had been viewed traditionally as deficient, on account of not being
syntactically well-formed sentences (Bhler 1990: 176179, 187189; 1934:
156159, 166167).27 Through examples such as that of a patron in a coffee house
asking for a particular type of coffee by way of uttering something like an espresso, please28 and of a passenger in a tram car choosing between types of
tickets, Bhler explains that an ellipsis interpretation is mistaken. In such cases,
the speaker does not have to produce semantically explicit or grammatically wellformed utterances, because the interlocutor, on account of shared knowledge of the
situation (the Umfeld), understands motives and plans for the action (in modern
terminology: the action schema or script) and uses the language sign only as a
diacritic. The situational context provides a sufficient fulfilment of all meaning
functions in the action field: the utterance is not elliptical but rather, in Bhlers
terminology, empractical (1990: 177179; 1934: 155158).
Generalizing from Bhlers hints, one can argue that language use is empractical in principle, insofar as it is always aimed at resolving ambiguities of intersubjective co-operation and based on some degree of shared background knowledge
(or at least, background assumptions). Whilst Bhler never formulated any communicative co-operation or relevance principles in the Sprachtheorie, his notion of the sympractical field as the basic action-structure in which language functions operate can be viewed at least as compatible with more recent pragmatic and
relevance-theoretical approaches.29
4.
Vilm Mathesius and the Prague Linguistic Circle
4.1.
Introduction
From 1930 onward, Bhler had scattered connections with members of the Prague
Linguistic Circle, the avant-garde group of linguists that propagated a view of language as a functional system with structural laws (Prague Linguistic Circle
240
[1929] 1983: 7779). Founded by the Czech linguist and scholar of English Vilm
Mathesius (18821945) in 1926, the Circle united under Mathesiuss presidency
some of his Czechoslovak colleagues and linguists of other nationalities who lived
in Prague or nearby, notably a number of Russian and Ukrainian migrs. Whereas
the functional point of view was the original contribution of Mathesius himself,
Roman Jakobson, the Russian vice-president of the Circle, added a structuralist,
much more system-oriented approach. Not surprisingly, Mathesiuss and Jakobsons legacies to pragmatics vary. The following section will detail Mathesiuss
functionalist findings that are relevant for pragmatics; in section 5, we will turn to
foundations of pragmatics deriving from Jakobsons work.
4.2.
Mathesiuss functional view of language
Since his earliest linguistic writings in the first decade of the century, Mathesius
had argued for a view of language as a means to an end.30 In the years of his university education in Prague, his general methodological outlook was formed by the
call for concreteness from the Czech philosopher Toms Masaryk (18501937),
who coupled this appeal with the desirability of democratising scientific and philosophical education (Masaryk 1887). Science had to give priority to the study of
close and accessible phenomena, including their comparative analysis; this
would prepare for historical research at a later stage.31 Accordingly, Mathesius
sought his linguistic inspirations in the works of opponents of Neogrammarian
dogmatism, such as Sweet, Otto Jespersen (18601943), von der Gabelentz and
Wegener (Mathesius [1936] 1966: 137), and he became an advocate of synchronic
language study.
The ideal of concreteness took shape in the emphasis which Mathesius laid
on the inherent synchronic variability of linguistic phenomena among individuals
of a language community as well as in the speech of a single individual; he spoke
of the static oscillation in speech and called this aspect of language its potentiality (Mathesius [1911] 1983a: 3). He stressed that the seeming homogeneity of
language is a product of the method of analysis and that without the phonetic phenomenon of synchronic oscillation, languages could not change (Mathesius
[1911] 1983a: 4, 32). These claims sound very modern and they were indeed approvingly cited in the influential paper on variability and language change by
Weinreich, Labov and Herzog (1968: 167169), although these authors also criticised Mathesius for not having pointed to the sociolinguistic regularities that are
inherent in variation.
In his ideas on synchronic oscillation in the domain of semantics, Mathesius
was at first strongly influenced by the stylistics of the Geneva linguist Charles
Bally (18651947) as well as by the idealistic approach of the German scholar of
Romance languages Karl Vossler (18721949). This stance implied an emphatic
interest in the expression of the speaking individual rather than in the social char-
241
acter of language, which he saw as having been overstated by nineteenth-century

linguistics (Mathesius [1911] 1983a: 2931, 3536; [1926] 1983b: 5253). In his
later work, Mathesius redressed the balance to some extent by way of an acknowledgement of the dominance, at least for language change, of the communicative
function of language vis--vis the original expressive function (Mathesius
[1926] 1983b: 5455). However, on account of his continued interest in stylistics
he kept a predilection for the expressive function and defined his functional
point of view as one that takes the meaning or function as its starting point and
tries to find out by which means it is expressed (Mathesius [1926] 1983b: 5658;
also [1929] 1983c: 122125). Mathesius thus came to neglect some of the necessary focus on the hearer as the key player in the constitution of meaning, something
which Wegener and Bhler had already reached in their work. This has to be
viewed as a general restriction of Praguian functional linguistics in its foundational
role for pragmatics.
Mathesiuss emphasis on the expressive function is understandable in the light
of his long-standing interest in the comparative study of word order patterns,
which was first kindled by the work of the French scholar of Ancient Greek Henri
Weil (18181909) (Weil [1844] 1887). Confronted with differences in word order
between Czech and English, Mathesius saw languages of different character.
Where a speaker of Czech usually filled the slot of the grammatical subject with the
agent of the action expressed by the verb, an English speaker appeared to prefer the
theme (Thema) of the utterance for this role. The Czech speaker therefore was
relatively free to order his sentence parts according to their contextual status of
theme or nucleus (rheme). For the English speaker to have a similar possibility readily at hand, the language had developed grammatical means for alternative choices of the grammatical subject, i.e. several types of passive constructions, which could be observed to make out a considerable proportion of English
verbal constructions (Mathesius 1929: 202203; [1929] 1983c: 126128, 133
134; 1930). The character of a language was thought to explain the links between
such descriptive facts of different types.
Mathesiuss characterology (or: language typology) thus relied on what he
called the actual articulation32 of utterances, which was to be distinguished from
its logico-grammatical articulation. The notion of actual articulation pointed to
the inherent situatedness of the structure in a specified context, and to its concrete quality; in short: it was a pragmatic structuring. Specifically, it consisted of
a basis or theme on the one hand and a nucleus, enunciation, or rheme
on the other.33 In a formulation from Mathesius (1975a: 81): The element about
which something is stated may be said to be the basis of the utterance or the theme,
and what is stated about the basis is the nucleus of the utterance or the rheme.
To an extent, Mathesiuss proposal of a functional sentence perspective resembles Wegeners ideas on the bipartition of utterances in exposition and
predicate that we discussed in section 2. Still, we note the less pronounced com-
242
municative emphasis with Mathesius: the hearer is absent in his formulation, and
even the speaker is no more than implied. Furthermore, Mathesiuss insight in the
role of the situation is much less developed. In any case, there appear to be no references by Mathesius to Wegeners theories on this subject, at least not from the
period in which Mathesius developed his notion of functional sentence structure
(the late 1920s). We have to decide that the relation is probably not a direct one.
It is certain however that Mathesiuss notions found favour in many quarters of
(pragmatic) linguistics and have become, under some of their original names or as
topic and comment, classics of pragmatic analysis (see, e.g., Lyons 1977:
501512; Levinson 1983: 8889).34
5.
Roman Jakobson
5.1.
Jakobsons model of linguistic functions
The Russian general linguist and scholar of Slavonic languages Roman Jakobson
(18961982) was intrigued by the seemingly autonomous functionality (teleology, self-movement) of language change all his life (Liberman 1987). For a
long time, Jakobson hoped to be able to account for it with the help of a Hegelian
logic of system change (Toman 1995: 171176). But in the milieu of the Prague
Linguistic Circle, he gradually came to recognise that the stages of a change in
progress had in fact to be conceived as different simultaneous facts, or styles,
within the language system. Especially after Jakobsons emigration to the United
States in 1941, his teleological point of view shifted to a focus on the less mysterious functions in a means-ends model of a synchronic system of systems (Jakobson 1963: 107).
In his new homeland, Jakobson found the study of language to have been, as he
remarked at a conference of anthropologists and linguists in 1952, strongly bulwarked by the impressive achievement of two conjoined disciplines the mathematical theory of communication and information theory (Jakobson 1953: 12).
This led him to discuss the factors of the communication process that were treated,
e.g., in the well-known publication by Shannon and Weaver (1949). In his paper,
Jakobson enumerates the four factors connected to the message in the information
theorists terminology (sender, receiver, topic of the message, code) and points to
the variability of their relative importance for a message. Apart from being associated with the factor of topic (information), a message may also be oriented toward the sender or the receiver, leading to a function of the message different
from a purely informative one, or, as the case may be, a hierarchical bundle of several functions (Jakobson 1953: 13). These three basic function varieties can be
considered equivalent to Bhlers (1990, 1934) functions of representation, expression and appeal, in the three-functions model that Jakobson subsequently
243
refers to as the traditional model of language (Jakobson 1960: 355; [1976] 1985:
115).35 It should be kept in mind however that Jakobson views the language functions as systemic facts throughout (facts of a system of systems). This distinguishes his theory from Bhlers approach (recall Bhlers championing of parole)
and even more from Wegeners; indeed, we are not aware of any references to Wegener in Jakobsons work.
Jakobson had come to linguistics via the study of Russian verse and of poetics
in general (cf. Rudy 1978; Toman 1995; Waugh and Rudy 1998: 22622265). Together with his compatriot, the literary scholar Jurij Tynjanov (18941943), he had
made a case for a parallel development of language, literature and other cultural
systems by structural necessity; there was to be no separation of the interconnected
codes of language and literature (Tynjanov and Jakobson ([1928] 1972). In a section of the 1929 Theses of the Prague Linguistic Circle, Jakobson and the Czech
literary scholar Jan Mukarovsky (18911975)36 had sought to contrast poetic language with communicative language. They had stated that poetic expression is
directed towards the way of expression itself and that the linguistic means of different levels, which in communicative speech tend to become automatised, in
poetic utterances tend to become foregrounded themselves (Prague Linguistic
Circle [1929] 1983: 94; cf. also 89). It is this idea that Jakobson takes up when he,
in his paper from 1953, adds a fourth, poetic function of language to Bhlers set
of three. For a message that is partly loosened from its referential meaning by
being oriented towards its own self, towards the palpability of its signs, the
poetic function of language is at work, be it as the predominant function of a literary fragment or as a subordinate, additional function of an otherwise prosaic
message (Jakobson 1953: 14).
Jakobson mentions also a fifth possibility: an orientation of the message toward
the factor of code, interesting because here the code (i.e., the langue) is not just an
abstraction but interacts with language use. The orientational set (explained as
Einstellung) for the code is witnessed in non-specialist discourse about such linguistic subjects as clarification of code elements, code switching, bilingualism,
areal diffusion, language variation and language change (Jakobson 1953: 1418).
This function is described with more clarity and termed the metalingual function
in a publication based on Jakobsons presidential address at a meeting of the Linguistic Society of America in 1956 (cf. Jakobson 1985: 116121). A sixth function
is mentioned in that address as well: Jakobson analyses anthropologist Malinowskis phatic communion function37 as being attached to messages with a set for the
factor of contact, or channel, as when interlocutors, without much information
transfer, check or revive the physical communication channel, or keep it going
(Jakobson 1985: 115).38
Jakobson (1960: 353357) sums up the resulting model of six language functions (with some terminological changes): the referential (denotative, cognitive)
function, the emotive (expressive) function, the conative39 function, the poetic
244
function, the metalingual (or glossing) function and the phatic function. They
are explained by the orientation of the message toward one of the six constitutive
factors of speech events, to wit the context referred to (the referent), the addresser,
the addressee, the message, the code, and the contact channel. It is in this version
that Jakobsons model of language functions has been widely cited by linguists of
several persuasions, including theoreticians of linguistic pragmatics (see, e.g.,
Hymes [1962] 1968: 99, 110124; Hymes 1983: 334; Lyons 1977: 5256; Levinson 1983: 41; Schiffrin 1994: 3233; Duranti 1997: 284287). It is clear, however,
that the model is focused on language as an undivided phenomenon, in which, just
as in Bhlers approach, the many facets of a pragmatic nature remain inclusive,
and therefore partly implicit. In order to gain an unimpeded view of pragmatic
facts, a systematic linguistic pragmatics had still to be drawn out.
5.2.
Jakobsons treatment of deixis
Among the further aspects of Jakobsons functionalism that are relevant for the development of linguistic pragmatics, his view of deictic phenomena is the most important. In a publication of 1957, which goes back to a paper delivered in Geneva
in 1950 (cf. Jakobson 1950), he treats the differentia specifica of indexical items
such as the personal pronouns I and you (Jakobson [1957] 1971). Like Bhler, Jakobson points to the fact that their precise meaning in context needs to take into account aspects of the situation, always unique, in which they happen to be used;
adopting Otto Jespersens (18601943) (1922: 123) term, he calls them shifters
(Jakobson 1971: 131). Jakobsons main interest however is focused on the question
of the status of these shifters as elements of the linguistic code, and he insists that
their symbolic meaning (for I, the addresser of the pronoun instance, and for you,
the addressee) should not be neglected either. Using Charles Sanders Peirces
(18391914) classification of sign modes and specifically following its interpretation by the American mathematician and philosopher Arthur Burks (19152008),
Jakobson argues in favour of according them the status of indexical symbols.40
First and second person pronouns in use are associated with the object they represent by an existential relation (i.e., pronoun token and object exist in the same
situation), therefore they indeed function as indices, just like acts of pointing
with the index finger would. However, on account of their conventional relation
with the role of speaker or hearer, which in different language codes holds for different forms (Latin ego, German ich, Russian ja), they partake just as well of the
symbolic mode of signs (Jakobson 1971: 132).41
Whereas the sober-minded Jespersen (1922: 123), in reaction to a remark by
the German idealist philosopher of subjectivity Johann Gottlieb Fichte
(17621814), had belittled attempts to derive deep philosophical conclusions from
the little linguistic trick involved in using the pronoun I correctly, Jakobson goes
on to develop its theoretical significance further. Considering the constitutive fac-
245
tors of speech situations, he notices for the shifters an overlapping of the factors
Code and Message: their general meaning, which is part of the Code, contains a
reference to the very Message of which the shifter in an instance of its use happens
to form a part (Jakobson 1971: 132). Jakobson explains this possibility by asserting
that both Code and Message can in principle refer as well as be an object of reference; in the specific case of the shifters, the Code has the function of referring,
while the Message is being referred to.42
This explanation takes up a similar remark by the French linguist mile Benveniste (19021976) in an article about pronouns that was published a year earlier
in the Festschrift For Roman Jakobson (Benveniste [1956] 1966a: 252). Benveniste, likewise anxious to find an aspect of the Saussurean langue (the Code) within
the typical discourse (Message) reality of first and second person pronouns, goes
further than Jakobson in postulating that it is precisely the use of elements symbolizing the roles of utterer-conceptualiser (nonciateur) and allocutee of a discourse event which leads to language users consciousness of their role as subjects. Those symbolic features of the language system then get parcelled out into
psychological reality. Benveniste also mentions that the process is strengthened by
various verbal forms that align with the perspective of the utterer of the discourse
(Benveniste 1966a: 254255; also [1958] 1966b: 258263). In his 1957 article,
Jakobson indeed uses the theory of the shifters to set up a new classification of verbal categories, in particular those of Russian. Like first and second person pronouns, verbal shifter categories (several forms in the paradigms of person, mood,
tense and evidentiality) have an encoded meaning in which the speech event (the
message) and/or one of its participants play a decisive role (Jakobson 1971:
133147).43
The relevance of the idea of the interweaving of Code (langue) and Message
(instance of discourse) to linguistic pragmatics becomes evident when the phenomenon of the shifters is described as a fact about speech events rather than the
linguistic code. It can be said then that by means of various indexical elements
of the language, speech events often encode, or index, important components of
the communitys social life. It was the American anthropologist Dell Hymes
(19272009) who realised that on this basis, Jakobsons speech functions could be
extended and seen to be differently organised for different communities, opening
up an additional realm of structure (Hymes 1983: 335). This was to result in the
ethnography of speaking and more broadly, the ethnography of communication
(Hymes [1962] 1968: 110129; 1964; 1972).
6.
Conclusion
Philipp Wegener, Karl Bhler, Vilm Mathesius and Roman Jakobson all of them
characteristically modernist theorists were fully aware of the pervasive relevance of language in countless domains of human life and culture. They projected
246
the pragmatic aspects which they had identified, such as the multifunctionality of
language, the functional structure of the sentence and the nature of deixis, onto a
rich picture of language that did not separate semantic from pragmatic phenomena
and often contained no fixed answer to the question as to what is language itself
and what is its use. Seen from a present-day perspective, their works may appear to
be lacking in systematic treatment of pragmatic facts. By the same token, however,
they have had and continue to have potential not for just one but many branches
of linguistic pragmatics. Wegeners ideas for the linguistic analysis of utterances
together with their embedding situations may still be used to further loosen the grip
of the received wisdom about the structure of linguistic utterances as phenomenal
entities per se. Similarly, Bhlers and Jakobsons theories of the functions of language, the associated models of the communication situation and the deictic signs/
shifters that derive their specific meaning from it offer the chance to review the
immunity of the theory of the language system to results of the analysis of language use.
Notes
1. Outside the discipline of linguistics, Bhlers model has been taken up by Jrgen Habermas, who in his theory of communicative action (cf. 1981) uses the three functional dimensions as reference points for a model of social communication. According to Habermas, every speaker, when making a meaningful utterance, intends it to be understood by
the hearer as legitimate (within the social context), truthful (i.e. as an expression of the
speakers own convictions) and true (in relation to objective states of affairs). Even
though these assumptions are often disappointed in actual communication, they act as
regulating norms for meaningful social exchange (Habermas 1981, vol. 1: 372377,
412). See further the article by Cooke in this volume.
2. For information on Wegener and his work, see especially Knobloch (1991); also Juchem
(1984, 1986), Knobloch (1988: 292297), Nerlich (1990: 153192; 1992: 8187), Graffi
(1991: 8789), Elffers (1993), Nerlich and Clarke (1996: 177183), Morpurgo Davies
(1998: 318320), Grimm-Vogel (1998), Tenchini (2008).
3. An English translation of Wegeners 1885 book appeared as part of Abse (1971); note
that Knobloch (1991: xliv) expresses a caveat about details of the translation.
4. In his publication from 1885, Wegener uses the term logical predicate; in later works he
switches to psychological predicate and statement (Mitteilung).
5. Knobloch (1991: xvi, xxxii) mentions Lazarus ([18561857] 1884) as a possible source
of Wegeners ideas in this area. About Moritz Lazarus (18241903), see Knobloch (1988:
268273, 411422) and Nerlich and Clarke (1996: 165168).
6. For details, see Hlzer (1987), Knobloch (1988: 296297; 1991: xviii-xix), Nerlich
(1990: 181185; 1992: 8687).
7. This is the opinion also of Knobloch (1988: 297) and Nerlich (1992: 87).
8. See the references to Wegeners work in Malinowski ([1923] 1946: 297), Gardiner (1932:
passim) and Halliday (1978: 109), particularly in relation to the notion of situation.
247
9. See Ungeheuer (1987, 1990) and Juchem (1984, 1986).

10. For details see Knobloch (1991: xxxix-xli).
11. Quotations from Bhlers works are referenced, where possible, to both the original
German publications and their English translation, i.e., in the case of the Sprachtheorie,
1934 for the German volume, 1990 for its translation.
12. See e.g. Schlieben-Lange (1979: 13); Levinson (1983: 41, 61); Brown and Yule (1983:
1): Coupland (2007: 12).
13. After the Sprachtheorie (1934), Bhler published several other linguistic articles up
until 1938 (Bhler 1935, 1936a, 1936b, 1938). Immediately after the occupation of
Vienna by German troops, Bhler was imprisoned by the Gestapo. His wife and academic collaborator, Charlotte Bhler, managed to effect his release and to organise their
emigration to the USA, but the continuation of their careers proved difficult: they only
gained posts as visiting or assistant professors and did not return to Europe after the war
(see Garvin 1966, Ch. Bhler 1984, Musolff 1997). The 1960 article Von den Sinnfunktionen der Sprachgebilde (Of the Semantic functions of linguistic structures) is an
excerpt from Sprachtheorie and of dubious editorial status (e.g. it cites the Organonmodell as Organmodell).
14 Husserl (19001901); Wundt (1900); Marty (1908). Bhler reviewed Martys work with
specific emphasis on the functional typology, see Bhler (1909a: 964946).
15. For Bhlers experiments on sentence comprehension in the context of Thought Psychology, see Bhler (1908), Wundt (1908), Wettersten (1988).
16. For the changes in the functions-terminology and the status and ordering of the
axioms, see Wunderlich (1969); Kamp (1977: passim); Beck (1980: 169192); Busse
(1975: 213215, 222229); Camhy (1980a: 9294); Swiggers (1981: 5455); Innis
(1982: 310); Eschbach (1984b: 9193); Henzler (1987: 354356); Knobloch (1988:
424430); Graumann (1988); Musolff (1990: 2631); Vonk (1992: passim); Elffers
(2005; 2008: 2124).
17. For the contemporary context of the Saussure-reception in relation to Bhler see
Scheerer (1980: 3640), Koerner (1984), Musolff (1990: 4043) and Ehlich (1996).
18. Details of what Bhler had learnt from Wegener can be found in Bhler (1909b:
119123).
19. See Gardiner (1932: viii).
20. For detailed discussion of these and other pragmatic-functional influences cf. Graumann (1984), Herrmann (1984), Knobloch and Schallenberger (1993), Musolff (1993)
and Nerlich and Clarke (1996: 182183, 226236, 341343).
21. See Bhler (1933b: 90); in Strkers edition (1976: 116117).
22. See Laziczius (1939); Lohmann (19421943); Camhy (1980b); Ortner (1986).
23. For detailed discussion of Grnbaums influence on Bhler see Knobloch and Schallenberger (1993: 8590).
24. For Bhlers field-theoretical approach (as opposed to the contemporary notion of
semantic fields) cf. Wolf (1932); Heger (1984); Musolff (1990: 61120); Garvin
(1994).
25. For Bhlers influence on modern pragmatically oriented studies of deixis see Hrmann
(1978: 394424); Weissenborn (1988); Conte et al. (1989); Innis (1992: 556557);
Ehlich (1996: 959).
26. Garvin (1994: 6064) interprets the relationship as a superordination of the more general surrounding field over the two subcategories of symbolic and deictic field; he
248
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
bases this reading on an excellent comparison with the contemporary gestalt-psychological theories, which Bhler himself highlighted (1990: 175176, 1934: 154155) as
well as with early semiotics (Charles Morris). However, in the Sprachtheorie this classification is not explicated nor consistently followed it constitutes a further theoretical
(and highly plausible) development, the implications of which are still to be worked out
in detail.
For critical assessments see Klein (1984), Ortner (1983).
Living in Vienna at the time, Bhler of course chose an appropriate local variety (einen
schwarzen a black one).
But compare Nerlich and Clarkes (1996: 236237) critical comment that Bhler, despite being one of the most important figures among language researchers engaged in
pragmatics avant la lettre, never crossed the last hurdle on his way to a fully developed theory of linguistic pragmatics, especially on account of his [clinging] too much
to his three functions of language, overlooking to some extent the polyfunctionality of
speech acts, rediscovered by Wittgenstein.
Toman (1995: 7779). About Mathesiuss work, see further Trnka ([1946] 1966), Vachek (1982), Danes (1987, 2003), Fronek (1988), Graffi (1991: 179183, 249250),
Toman (1995: 7186, 97101 and passim), Leska (1995), Nekula (1999).
See Toman (1995: 8486, 97101), also Fronek (1988).
This notion features in the original title of the Czech article Mathesius ([1939] 1975b);
see the translators remark in Mathesius (1975b: 479). It came to be translated as information-bearing structure of the sentence, shortened to information structure and
alternating with functional sentence perspective (FSP).
The terminology differs; see Mathesius ([1929] 1983c: 126128; 1929: 202; 1930: 58).
The term rheme came into use only later, through the writings of the Brno scholar of
English Jan Firbas (19212000). See the explanation by Vachek in Mathesius (1975a:
185 n71), further Vachek (1982: 123124) and Danes (1987: 2326; 2003: 4041).
The Prague version of the information structure of the utterance has remained a distinctive one. See a short description in Hajicov (1994).
Jakobson referred to Bhlers three functions already in 1932 (Jakobson [1932] 1966:
2728). In its Theses of again some years earlier, the Prague Linguistic Circle had distinguished, under the heading On the functions of langugage, intellectual speech
from two types of emotional speech and had mentioned the different hierarchy of
functions for each utterance (Prague Linguistic Circle [1929] 1983: 8889). Jakobson
probably wrote (part of) this section (nr. 3a) of the Theses (cf. Toman 1995: 289 n13).
For the attribution to Jakobson and Mukarovsky, see again Toman (1995: 289 n13).
See Malinowski ([1923] 1946: 315316). Cf. Nekula and Ehlers (1996: 190, 192) on the
at first marginal place of the phatic function of language in Jakobsons model.
On Jakobsons model of language functions, see Holenstein (1976: 153164); Waugh
(1976: 2526); Waugh and Monville-Burston (1990: 1516); Nekula and Ehlers (1996);
Nerlich and Clarke (1996: 284286); Waugh and Rudy (1998: 22602261).
Conative: with the notion of striving to effect some act by the addressee; cf. Jakobson ([1932] 1966: 22, 27).
See Burks (1949). Among the signs having the property of indexicality, Burks views the
indexical symbol as the most fundamental type; a pure index cannot in effect point
to a well-defined object and is therefore a less fundamental type of sign (Burks 1949:
678688).
249
41. Jakobsons (1971: 132) statement to the effect that Bhler acknowledged the indexical
side of personal pronouns only seems not fully justified. It is when Bhler introduces the
indexical field (Zeigfeld) that he contrasts indexical and appellative words as to the
field in which, specifically, their meaning fulfilment (Bedeutungsprzision, Bedeutungserfllung, see above, 2.2) is located: in the indexical field and the symbol-field of
an utterance, respectively (Bhler 1990: 9194; 1934: 7881). This does not imply that
personal pronouns would not have a symbolic meaning aspect to begin with, such as
the sender of the word in question (Sender des aktuellen Wortes) (Bhler 1990: 91;
1934: 78); indeed, the notion of meaning fulfilment presupposes the opposite idea.
42. Considering the three other possibilities of combining the choices of Code/Message and
referring/referred-to item, Jakobson theorises further in the following way. When the
functions are turned round and it is rather the Message that refers and the Code that is referred to, we have to do with the elucidation of words and sentences (what would be
named the metalingual function of language a few years later); when a Message refers
to a Message, the result is the well-known phenomenon of reported speech; and lastly,
with Code referring to Code we could describe the general nature of the meaning of
proper names, Jakobson holds (Jakobson 1971: 130131).
43. On Jakobsons notion of shifters, see Holenstein (1976: 158, 162163); Waugh (1976:
2425, 32, 48, 52); Waugh and Monville-Burston (1990: 1719); Nerlich and Clarke
(1996: 282284).
References
Abse, D. Wilfred
1971
Speech and Reason: Language Disorder in Mental Disease. With a Translation
of The Life of Speech by Philipp Wegener. Charlottesville, VA: University
of Virginia Press.
Beck, Gtz
1980
Sprechakte und Sprachfunktionen: Untersuchungen zur Handlungsstruktur
der Sprache und ihren Grenzen. Tbingen: Niemeyer.
Benveniste, mile
1966a
La nature des pronoms. In: mile Benveniste, Problmes de linguistique gnrale, tome 1, 251257. Paris: Gallimard. First published The Hague: Mouton
[1956].
Benveniste, mile
1966b
De la subjectivit dans le langage. In: mile Benveniste, Problmes de linguistique gnrale, tome 1, 258266. Paris: Gallimard. First published Paris: PUF
[1958].
Brown, Gillian and George Yule
1983
Discourse Analysis. Cambridge: Cambridge University Press.
Brugmann, Karl
1904
Die Demonstrativpronomina der indogermanischen Sprachen: Eine bedeutungsgeschichtliche Untersuchung. In: Berichte ber die Verhandlungen der
Kniglichen schsischen Gesellschaft der Wissenschaften zu Leipzig, Philologisch-Historische Classe, No. 6: 3151.
250
Bhler, Charlotte
1984
Karl Bhler: Eine biographische Skizze. Transl. by Achim Eschbach. In:
Achim Eschbach (ed.), Bhler-Studien, Vol. 1, 2530. Frankfurt am Main:
Suhrkamp.
Bhler, Karl
1908
Tatsachen und Probleme zu einer Psychologie der Denkvorgnge: II) ber
Gedankenzusammenhnge. Archiv fr die gesamte Psychologie 12: 123; III)
ber Gedankenerinnerungen. Archiv fr die gesamte Psychologie 12: 2492.
Bhler, Karl
1909a
Review of Anton Marty, Untersuchungen zur Grundlegung der allgemeinen
Grammatik und Sprachphilosophie (1908). Gttingische gelehrte Anzeigen
171: 947979.
Bhler, Karl
1909b
ber das Sprachverstndnis vom Standpunkt der Normalpsychologie aus. In:
Bericht ber den 3. Kongre fr experimentelle Psychologie (Frankfurt 1908),
94130. Leipzig: Wilhelm Engelmann.
Bhler, Karl
1918
Kritische Musterung der neuern Theorien des Satzes. Indogermanisches Jahrbuch 6: 120.
Bhler, Karl
1927
Die Krise der Psychologie. Jena: J.A. Barth.
Bhler, Karl
1931
Phonetik und Phonologie. Travaux du Cercle Linguistique de Prague 4: 2253.
Bhler, Karl
1932
Das Ganze der Sprachtheorie, ihr Aufbau und ihre Teile. In: Gustav Kafka
(ed.), Bericht ber den 12. Kongre der Deutschen Gesellschaft fr Psychologie (Hamburg 1931), 95122. Jena: G. Fischer.
Bhler, Karl
1933a
Ausdruckstheorie: Das System an der Geschichte aufgezeigt. Jena: G. Fischer.
Bhler, Karl
1933b
Die Axiomatik der Sprachwissenschaften. Kant-Studien 38: 1990. (Reprinted
1969 and 1976 under the same title in a truncated version without the introductory chapter, edited by Elisabeth Strker, 25117). Frankfurt am Main: Klostermann.
Bhler, Karl
1934
Sprachtheorie: Die Darstellungsfunktion der Sprache. Jena: G. Fischer.
Bhler, Karl
1935
Forschungen zur Sprachtheorie: Einleitung. Archiv fr die gesamte Psychologie 94: 401412.
Bhler, Karl
1936a
Psychologie der Phoneme. In: Daniel Jones and Dennis B. Fry (eds.), Proceedings of the Second International Congress of Phonetic Sciences, 162169.
Bhler, Karl
1936b
Das Strukturmodell der Sprache. Travaux du Cercle Linguistique de Prague 6:
312.
251
Bhler, Karl
1938
Der dritte Hauptsatz der Sprachtheorie: Anschauung und Begriff im Sprechverkehr. In: Henri Piron and Emile Meyerson (eds.), 11e Congrs International de Psychologie (Paris 1937). Paris: Rapports et Comptes Rendus,
196203.
Bhler, Karl
1960
Von den Sinnfunktionen der Sprachgebilde. In: Richard Wisser (ed.), Sinn und
Sein: Ein philosophisches Symposion; 95112. Tbingen: Niemeyer:
Bhler, Karl
1990
Theory of Language: The Representational Function of Language. Translation
by Donald Fraser Goodwin. Amsterdam: Benjamins.
Burks, Arthur W.
1949
Icon, index, and symbol. Philosophy and Phenomenological Research 9:
673689.
Busse, Winfried
1975
Funktionen und Funktion der Sprache. In: Brigitte Schlieben-Lange (ed.),
Sprachtheorie, 207240. Hamburg: Hoffmann and Campe:
Camhy, Daniela G.
1980a
Karl Bhlers Sprachtheorie. PhD. Diss. Graz: University of Graz.
Camhy, Daniela G.
1980b
Sprechereignis als Handlung: Bhlers Beitrag zu handlungstheoretischen Anstzen in der Sprachphilosophie. In: Rudolf Haller and Wolfgang Grassl (eds.),
Sprache, Logik und Philosophie, 401405. Wien: Hlder-Pichler-Temsky.
Conte, Maria Elisabeth, Jnos Snder Petfi and Emel Szer (eds.)
1989
Text and Discourse Connectedness. Amsterdam: Benjamins.
Coupland, Nikolas
2007
Style: Language, Variation and Identity. Cambridge: Cambridge University
Press.
Danes, Frantisek
1987
On Prague School functionalism in linguistics. In: Ren Dirven and Vilm
Fried (eds.), Functionalism in Linguistics, 338. (Linguistic and Literary
Studies in Eastern Europe 20.). Amsterdam: Benjamins.
Danes, Frantisek
2003
Vilm Mathesius Konzeption der funktionalen Linguistik. In: Marek Nekula
(ed.), Prager Strukturalismus: Methodologische Grundlagen, 3147. (Slavica:
Monographien, Hand-, Lehr- und Wrterbcher 3.). Heidelberg: Universittsverlag Winter.
Duranti, Alessandro
1997
Linguistic Anthropology. (Cambridge Textbooks in Linguistics.) Cambridge:
Ehlich, Konrad
1996
Sprache als System versus Sprache als Handlung. In: Marcelo Dascal, Dietfried Gerhardus, Kuno Lorenz, Georg Meggle (eds.), Sprachphilosophie: Ein
internationales Handbuch zeitgenssischer Forschung, vol. 1, 952963.
(Handbcher zur Sprach- und Kommunikationswissenschaft 7.1). Berlin/New
York: de Gruyter.
252
Elffers, Els
1993
Philipp Wegener as proto-speech act theorist. Linguistics in the Netherlands
1993: 4959. Amsterdam: Benjamins.
Elffers, Els
2005
Reichling and de Groot: Two Dutch reactions to Bhlers organon-model. Historiographia Linguistica 32: 87116.
Elffers, Els
2008
Interjections and the language functions debate. The Henry Sweet Society
Bulletin 50: 1729.
Eschbach, Achim (ed.)
1984a
Bhler-Studien. 2 vols. Frankfurt am Main: Suhrkamp.
Eschbach, Achim
1984b
Karl Bhlers Axiomatik und das Axiomensystem der Zeichentheorie. In: Carl
F. Graumann and Theo Herrmann (eds.), Karl Bhlers Axiomatik: Fnfzig
Jahre Axiomatik der Sprachwissenschaften, 53104. Frankfurt am Main:
Klostermann.
Fronek, Josef
1988 T.G. Masaryk and Prague School linguistic structuralism. In: Josef Novk (ed.), On
Masaryk: Texts in English and German, 8196. (Studien zur sterreichischen
Philosophie, 13). Amsterdam: Rodopi.
Gabelentz, Georg von der
1875
Weiteres zur vergleichenden Syntax: Wort- und Satzstellung. Zeitschrift fr
Vlkerpsychologie und Sprachwissenschaft 8: 129165, 300338.
Gardiner, Sir Alan Henderson
1932
The Theory of Speech and Language. Oxford: Clarendon Press. (2nd ed. with
additions 1951.)
Garvin, Paul L.
1966
Karl Bhlers contributions to the theory of linguistics. The Journal of General
Psychology 75: 212215.
Garvin, Paul L.
1994
Karl Bhlers Field Theory in the Light of the Current Interest in Pragmatics.
In: Svetla Cmejrkov and Frantisek Stcha (eds.), The Syntax of Sentence and
Text: A Festschrift for Frantisek Danes, 5966. Amsterdam: Benjamins.
Graffi, Giorgio
1991
200 Years of Syntax: A Critical Survey. (Studies in the History of the Language
Sciences 98.) Amsterdam: Benjamins.
Graumann, Carl F.
1984
Wundt Mead Bhler: Zur Sozialitt und Sprachlichkeit menschlichen Handelns. In: Carl F. Graumann and Theo Herrmann (eds.), Karl Bhlers Axiomatik: Fnfzig Jahre Axiomatik der Sprachwissenschaften, 217247. Frankfurt
am Main: Klostermann.
Graumann, Carl F.
1988
Aspektmodell und Organonmodell: Die Problematik des Verhltnisses zwischen
Sprachwissenschaft und Psychologie bei Karl Bhler. In: Achim Eschbach
(ed.), Karl Bhlers Theory of Language, 107124. Amsterdam: Benjamins.
Grimm-Vogel, Irmingard Hildburg
1998
Philipp Wegener 18481916: Wesen Wirken Wege. Dissertation Rheinischen Friedrich-Wilhelms-Universitt zu Bonn.
253
Grnbaum, Abraham Anton

1932
Sprache als Handlung. In: Gustav Kafka (ed.), Bericht ber den 12. Kongre
der Deutschen Gesellschaft fr Psychologie (Hamburg 1931), 164176. Jena:
G. Fischer.
Habermas, Jrgen
1981
Theorie des kommunikativen Handelns. 2 vols. Frankfurt am Main: Suhrkamp.
Hajicov, Eva
1994
Topic/Focus and related research. In: Philip A. Luelsdorff (ed.), The Prague
School of Structural and Functional Linguistics: A Short Introduction,
245275. (Linguistic and Literary Studies in Eastern Europe 41.). Amsterdam:
Benjamins.
1978
Language as Social Semiotic: The Social Interpretation of Language and
Meaning. London: Edward Arnold.
Heger, Klaus
1984
Zeigfeld und Symbolfeld. In: Carl F. Graumann and Theo Herrmann (eds.),
Karl Bhlers Axiomatik: Fnfzig Jahre Axiomatik der Sprachwissenschaften, 105116. Frankfurt am Main: Klostermann.
Henzler, Stefan
1987
Die Funktionen der Sprache im Gebrauch: Zu Konsistenz und Funktionalismus von Karl Bhlers Sprachtheorie. In: Werner Abraham and Ritva rhammar (eds.), Linguistik in Deutschland, 353365. Tbingen: Niemeyer.
Herrmann, Theo
1984
Sprachstrukturen und ihre aktionale Realisation: Psychologische Erwgungen
zu Karl Bhlers Axiomatik der Sprachwissenschaften. In: Carl F. Graumann
and Theo Herrmann (eds.), Karl Bhlers Axiomatik: Fnfzig Jahre Axiomatik
der Sprachwissenschaften, 143172. Frankfurt am Main: Klostermann.
Holenstein, Elmar
1976
Roman Jakobsons Approach to Language: Phenomenological Structuralism.
Transl. by Catherine Schelbert and Tarcisius Schelbert. Bloomington: Indiana
University Press.
Husserl, Edmund
19001901 Logische Untersuchungen. Vol. I: Prologomena zur reinen Logik. Vol. II/1:
Untersuchungen zur Phnomenologie und Theorie der Erkenntnis. Vol. II/2:
Elemente einer phnomenologischen Aufklrung der Erkenntnis. Halle: Niemeyer.
Hrmann, Hans
1978
Meinen und Verstehen: Grundzge einer psychologischen Semantik. Frankfurt
am Main: Suhrkamp.
Hlzer, Heike
1987
Die Metapher: Kommunikationssemantische berlegungen zu einer rhetorischen Kategorie. Mnster: Nodus.
Hymes, Dell
1964
Introduction: Toward ethnographies of communication. In: John J. Gumperz
and Dell Hymes (eds.), The Ethnography of Communication, 134. (Special
issue of American Anthropologist 66, 6, part 2.)
254
Hymes, Dell H.
1968
The ethnography of speaking. In: Joshua A. Fishman (ed.), Readings in
the Sociology of Language, 99138. The Hague: Mouton. First published
Washington, D.C. [1962].
Hymes, Dell
1972
Models of the interaction of language and social life. In: John J. Gumperz
and Dell H. Hymes (eds.), Directions in Sociolinguistics: The Ethnography of
Communication, 3571. Oxford: Blackwell.
Hymes, Dell
1983
The pre-war Prague school and post-war American anthropological linguistics.
In: Dell Hymes, Essays in the history of linguistic anthropology, 331345.
Innis, Robert E.
1982
Karl Bhler: Semiotic Foundations of Language Theory. New York/London:
Plenum Press.
Innis, Robert E.
1992
Karl Bhler (18791963). In: Marcelo Dascal, Dietfried Gerhardus, Kuno Lorenz, Georg Meggle (eds.), Sprachphilosophie: Ein internationales Handbuch
zeitgenssischer Forschung. Vol. 1, 550562. (Handbcher zur Sprach- und
Kommunikationswissenschaft 7.1.) Berlin/New York: de Gruyter.
Jakobson, Roman
1950
Les catgories verbales. [Summary]. Cahiers Ferdinand de Saussure 9: 6.
Jakobson, Roman
1953
[From the point of view of linguistics.] In: Results of the Conference of Anthropologists and Linguists, 1121. (Indiana University Publications in Anthropology and Linguistics; International Journal of American Linguistics, Memoir 8.)
Baltimore: Waverly Press.
Jakobson, Roman
1960
Jakobson, Roman
1963
Efforts toward a means-ends model of language in inter-war continental linguistics. In: Christine Mohrmann, F. Norman and Alf Sommerfelt (eds), Trends
in modern linguistics, 104108. Utrecht/Antwerp: Spectrum.
Jakobson, Roman
1966
Zur Struktur des russischen Verbums. In: Eric P. Hamp, Fred W. Householder
and Robert Austerlitz (eds.), Readings in Linguistics II, 15. Chicago/London:
The University of Chicago Press. First published Prague: Prague Linguistic
Circle [1932].
Jakobson, Roman
1971
Shifters, verbal categories, and the Russian verb. In: Roman Jakobson, Selected Writings, vol. 2: Word and Language, 130147. The Hague/Paris: Mouton. First published Cambridge, MA: Harvard University [1957].
Jakobson, Roman
1985
Metalanguage as a linguistic problem. In: Roman Jakobson, Selected Writings,
vol. 7: Contributions to Comparative Mythology: Studies in Linguistics and
Philology, 19721982, ed. by Stephen Rudy, 113121. Berlin: Mouton de
Gruyter. First published [1976].
255
Jespersen, Otto
1922
Language: Its Nature, Development and Origin. London: George Allen and
Unwin; New York: Holt.
Juchem, Johann G.
1984
Die Konstruktion des Sprechens: Kommunikationssemantische Betrachtungen
zu Philipp Wegener. Zeitschrift fr Sprachwissenschaft 3: 318.
Juchem, Johann G.
1986
Wegener und Wundt. Kodikas/Code 9: 155166.
Juchem, Johann G.
1987
Einleitung. In: Gerold Ungeheuer, Kommunikationstheoretische Schriften,
vol. 1: Sprechen, mitteilen, verstehen. Ed. by Johannes G. Juchem, 117. (Aachener Studien zur Semiotik und Kommunikationsforschung 14.). Aachen:
Rader Verlag.
Kamp, Rudolf
1977
Axiomatische Sprachtheorie: Wissenschaftstheoretische Untersuchungen zum
Konstitutionsproblem der Einzelwissenschaften am Beispiel der Sprachwissenschaftstheorie Karl Bhlers. Berlin: Duncker and Humblot.
Klein, Wolfgang
1984
Bhler Ellipse. In: Carl F. Graumann and Theo Herrmann (eds.), Karl Bhlers
Axiomatik: Fnfzig Jahre Axiomatik der Sprachwissenschaften, 117141.
Frankfurt am Main: Klostermann.
Knobloch, Clemens
1988
Geschichte der psychologischen Sprachauffassung in Deutschland von 1850
bis 1920. (Reihe Germanistische Linguistik 86.) Tbingen: Niemeyer.
Knobloch, Clemens
1991
Introduction. In: Philipp Wegener, Untersuchungen ber die Grundfragen des
Sprachlebens. New edition by Konrad Koerner, with an introduction by Clemens Knobloch, xili. (Classics in Psycholinguistics 5.) Amsterdam: Benjamins.
Knobloch, Clemens and Stefan Schallenberger
1993
Sprechhandlung und Sprachbedeutung in der Sprachpsychologie um 1930.
Histoire pistmologie Langage 15/1: 81109.
Koerner, E. F. Konrad
1984
Karl Bhlers theory of language and Ferdinand de Saussures Cours. Lingua 62: 324.
Lazarus, Moritz
1884
Geist und Sprache: Eine psychologische Monographie. 3rd ed. Berlin:
Dmmler. First published Berlin [18561857].
Laziczius, Julius von
1939
Das sog. dritte Axiom der Sprachwissenschaft. Acta Linguistica 1: 162167.
Leska, Oldrich
1995
Vilm Mathesius (18821945). [In Czech.] Slovo a Slovesnost 56: 8197.
1983
Liberman, Anatoly
1987
Roman Jakobson and his contemporaries on change in language and literature
(the teleological criterion). In: Krystyna Pomorska, Elzbieta Chodakowska,
256
Hugh McLean and Brent Vine (eds.), Language, Poetry and Poetics: The
Generation of the 1890s: Jakobson, Trubetzkoy, Majakovskij, 143155. Berlin:
Mouton de Gruyter.
Lohmann, Johannes
19421943 Karl Bhlers Drittes Axiom. Acta Linguistica 3: 516.
Lyons, John
1977
Semantics. 2 vols. Cambridge: Cambridge University Press.
Malinowski, Bronislaw
1946
The problem of meaning in primitive languages. In: Charles K. Ogden and Ivor
A. Richards, The Meaning of Meaning: A Study of the Influence of Language
upon Thought and of the Science of Symbolism. 8th ed, 296336. London:
Kegan Paul. First published London: Kegan Paul [1923].
Marty, Anton
1908
Untersuchungen zur Grundlegung der allgemeinen Grammatik und Sprachphilosophie. Vol. 1. Halle: Niemeyer.
Masaryk, Toms Garrigue
1887
Versuch einer concreten Logik: Classification und Organisation der Wissenschaften. Wien: C. Konegen.
Mathesius, Vilm
1929
Zur Satzperspektive im modernen Englisch. Archiv fr das Studium der
neueren Sprachen und Literaturen 155: 202210.
Mathesius, Vilm
1930
On linguistic characterology with illustrations from Modern English. In: Actes
du premier Congrs International de Linguistes la Haye, du 1015 avril
1928, 5663. Leiden: Sijthoff.
Mathesius, Vilm
1966
Ten years of the Prague Linguistic Circle. Transl. by Josef Vachek. In: Josef
Vachek (ed.), The Linguistic School of Prague: An Introduction to its Theory
and Practice, 137151. (Indiana University Studies in the History and Theory
of Linguistics.) Bloomington/London: Indiana University Press. First published in Czech, Prague [1936].
Mathesius, Vilm
1975a
A Functional Analysis of Present Day English on a General Linguistic Basis.
Ed. by Josef Vachek, transl. by L. Duskov. (Janua Linguarum, Series Practica
208). The Hague/Paris: Mouton; Prague: Academia. First published in Czech,
Prague [1961].
Mathesius, Vilm
1975b
On information-bearing structure of the sentence. Transl. by Tsuneko Olga Yokoyama. Harvard Studies in Syntax and Semantics 1: 467480. First published
in Czech, Prague [1939].
Mathesius, Vilm
1983a
On the potentiality of the phenomena of language. Transl. by Josef Vachek. In:
Josef Vachek and Libuse Duskov (eds.), Praguiana: Some basic and less
known aspects of the Prague Linguistic School, 343. (Linguistic and Literary
Studies in Eastern Europe 12.). Amsterdam: Benjamins. First published in
Czech, Prague [1911].
257
Mathesius, Vilm
1983b
New currents and tendencies in linguistic research. In: Josef Vachek and
Libuse Duskov (eds.), Praguiana: Some basic and less known aspects of the
Prague Linguistic School, 4563. (Linguistic and Literary Studies in Eastern
Europe 12.). Amsterdam: Benjamins. First published Prague [1926].
Mathesius, Vilm
1983c
Functional linguistics. Transl. by Libuse Duskov. In: Josef Vachek and
Prague Linguistic School, 121142. (Linguistic and Literary Studies in Eastern Europe 12.). Amsterdam: Benjamins. First published in Czech, Prague
[1929].
Morpurgo Davies, Anna
1998
History of Linguistics. Vol. 4: Nineteenth-Century Linguistics. Ed. by Giulio
Lepschy. (Longman Linguistics Library.) London/New York: Longman. First
published in Italian, Bologna [1992].
Musolff, Andreas
1990
Kommunikative Kreativitt: Karl Bhlers Zweifelderlehre als Ansatz zu einer
Theorie innovativen Sprachgebrauchs. Aachen: Alano.
Musolff, Andreas
1993
Karl Bhlers and Alan Gardiners concepts of metaphor in the context of their
theories of speech and language. Beitrge zur Geschichte der Sprachwissenschaft 3: 255272.
Musolff, Andreas
1997
Karl Bhler. In: Jef Verschueren, Jan-Ola stman, Jan Blommaert, Chris Bulcaen (eds.), Handbook of Pragmatics 1997, 115. Amsterdam: Benjamins.
Nekula, Marek
1999
Vilm Mathesius. In: Jef Verschueren, Jan-Ola stman, Jan Blommaert, Chris
Bulcaen (eds.), Handbook of Pragmatics 1999, 114. Amsterdam: Benjamins.
Nekula, Marek, and Klaas-Hinrich Ehlers
1996
Jakobsons Bilanz des Strukturalismus: Ein Vortrag, der jahrzehntelang nicht
gedruckt werden konnte. Beitrge zur Geschichte der Sprachwissenschaft 6:
187199.
Nerlich, Brigitte
1990
Change in Language: Whitney, Bral, and Wegener. (Routledge History of
Linguistic Thought Series.) London/New York: Routledge.
Nerlich, Brigitte
1992
Semantic Theories in Europe 18301930: From Etymology to Contextuality.
(Studies in the History of the Language Sciences 59.) Amsterdam: Benjamins.
1996
Ortner, Hanspeter
1983
Karl Bhlers Stellung in der Ellipsenforschung. Zeitschrift fr Germanistische
Linguistik 11: 149165.
Ortner, Hanspeter
1986
Bhlers Vierfelderschema (Das dritte Axiom der Sprachtheorie) Grundgedanken und Rezeption. Kodikas/Code, Ars Semeiotica 9: 211226.
258
Prague Linguistic Circle

1983
Theses presented to the first Congress of Slavists held in Prague in 1929.
Transl. from Czech by Josef Vachek. In: Josef Vachek and Libuse Duskov
(eds.), Praguiana: Some basic and less known Aspects of the Prague Linguistic
School, 77120. (Linguistic and Literary Studies in Eastern Europe 12.). Amsterdam: Benjamins. First distributed as loose sheets in Czech, Prague [1929].
Rudy, Stephen
1978
Jakobsonian Poetics of the Moscow and Prague Periods. Doctoral dissertation
Yale University.
Saussure, Ferdinand de
1931
Grundfragen der allgemeinen Sprachwissenschaft. Berlin: W. de Gruyter.
Transl. by Hermann Lommel of: Cours de linguistique gnrale, ed. by
Charles Bally and Albert Sechehaye, 1916.
Scheerer, Thomas M.
1980
Ferdinand de Saussure: Rezeption und Kritik. (Ertrge der Forschung 133.)
Darmstadt: Wissenschaftliche Buchgesellschaft.
Schiffrin, Deborah
1994
Approaches to Discourse. Oxford: Blackwell.
Schlieben-Lange, Brigitte
1979
Linguistische Pragmatik. Stuttgart: W. Kohlhammer.
1949
Press.
Swiggers, Pierre
1981
Two dimensions in Karl Bhlers sign theory. Kodikas/Code, Ars Semeiotica 4:
5356.
Tenchini, Maria Paola
2008
Aspetti funzionali e pragmatici nel pensiero di Philipp Wegener. Con la traduzione antologica di Philipp Wegener, Untersuchungen ber die Grundfragen
des Sprachlebens. (Quaderni del Centro di Linguistica dellUniversit Cattolica 11.) Brescia: Editrice La Scuola.
Toman, Jindrich
1995
The Magic of a Common Language: Jakobson, Mathesius, Trubetzkoy, and the
Prague Linguistic Circle. Cambridge, MA: MIT Press.
Trnka, Bohumil
1966
Vilm Mathesius. Transl. by Vladimir Honsa. In: Thomas A. Sebeok (ed.),
Portraits of Linguists: A Biographical Source Book for the History of Western
Linguistics, 17461963, vol. 2, 474489. Bloomington: Indiana University
Press. First published in Czech, Prague [1946].
Tynjanov, Jurij, and Roman Jakobson
1972
Probleme der Literatur- und Sprachforschung. Translated by Rolf Fieguth und
Inge Paulmann. In: Texte der russischen Formalisten, vol. 2: 386391. First
published [1928].
Ungeheuer, Gerold
1987
Kommunikationstheoretische Schriften, vol. 1: Sprechen, mitteilen, verstehen.
Ed. by Johannes G. Juchem. Aachen: Rader Verlag.
259
Ungeheuer, Gerold
1990
Kommunikationstheoretische Schriften, vol. 2: Symbolische Erkenntnis und
Kommunikation. Ed. by H. Walter Schmitz. Aachen: Alano Verlag.
Vachek, Josef
1982
Vilm Mathesius living heritage to world linguistics. Philologica Pragensia
25:121127.
Vonk, Frank
1992
Gestaltprinzip und abstraktive Relevanz: Eine wissenschaftshistorische Untersuchung zur Sprachaxiomatik Karl Bhlers. Mnster: Nodus.
Waugh, Linda R.
1976
Roman Jakobsons Science of Language. Lisse: Peter de Ridder Press.
Waugh, Linda R., and Monique Monville-Burston
1990
Introduction: the life, work, and influence of Roman Jakobson. In: Roman
Jakobson, On Language. Ed. by Linda R. Waugh and Monique MonvilleBurston, 145. Cambridge, MA: Harvard University Press.
Waugh, Linda R., and Stephen Rudy
1998
Jakobson and structuralism. In: Roland Posner, Klaus Robering and Thomas
A. Sebeok (eds.), Semiotik: Ein Handbuch zu den zeichentheoretischen Grundlagen von Natur und Kultur, vol. 2, 22562271. (Handbcher zur Sprach- und
Kommunikationswissenschaft 13.2.). Berlin/New York: de Gruyter.
Wegener, Philipp
1902
Review of Berthold Delbrck, Grundfragen der Sprachforschung (Strassburg:
Trbner, 1901). Literarisches Centralblatt 12 (22. Mrz 1902): 401410.
Wegener, Philipp
1911
Exposition und Mitteilung: Ein Beitrag zu den Grundfragen des Sprachlebens.
In: Festschrift zur Feier des 350 jhrigen Bestehens des Gymnasiums zu
Greifswald, Bd. 2: Wissenschaftliche Aufstze, 121. Greifswald: Kunike.
Wegener, Philipp
1991
Untersuchungen ber die Grundfragen des Sprachlebens. New edition by
Konrad Koerner, with an introduction by Clemens Knobloch. (Classics in Psycholinguistics 5.) Amsterdam: Benjamins. First published Halle: Niemeyer,
[1885].
Weil, Henri
1887
The Order of Words in the Ancient Languages compared with that of the
Modern Languages. Transl. by Charles W. Super. Boston: Ginn. First published in French, Paris: Joubert [1844].
Weinreich, Uriel, William Labov and Marvin I. Herzog
1968
Empirical foundations for a theory of language change. In: Winfred P. Lehmann and Yakov Malkiel (eds.), Directions for Historical Linguistics: A Symposium, 95195. Austin/London: University of Texas Press.
Weissenborn, Jrgen
1988
Von der demonstratio ad oculos zur Deixis am Phantasma: Die Entwicklung
der lokalen Referenz bei Kindern. In: Achim Eschbach (ed.), Karl Bhlers
Theory of Language, 257276. Amsterdam: Benjamins.
Wettersten, John
1988
Klpe, Bhler, Popper. In: Achim Eschbach (ed.), Karl Bhlers Theory of
Language, 327347. Amsterdam: Benjamins.
260
Wolf, Kthe
1932
Darstellungsfelder in der Sprache. In: Gustav Kafka (ed.), Bericht ber den
12. Kongre der Deutschen Gesellschaft fr Psychologie (Hamburg 1931),
449453. Jena: G. Fischer.
Wunderlich, Dieter
1969
Karl Bhlers Grundprinzipien der Sprachtheorie. Muttersprache 79: 5262.
Wundt, Wilhem
1900
Die Sprache. 2 Parts (= Vol. 1 of Vlkerpsychologie). Leipzig: Engelmann.
Wundt, Wilhem
1908
Kritische Nachlese zur Ausfragemethode. Archiv fr die gesamte Psychologie
11: 445459.
9.
Foundations: ethnomethodology and

Erving Goffman
Christine Domke and Werner Holly
1.
Genesis and foundations of ethnomethodology
1.1.
Harold Garfinkels perspective on peoples everyday life activities
The most crucial assumption of ethnomethodology is that social reality exists only
through the everyday activities of a given societys members. Ethnomethodologists thus focus on the question of how social order arises in everyday life and is recognized as such. Answering this question entails identifying the actual practices
used by members of a society to create order. The goal of ethnomethodological
studies is to identify and reconstruct the methods used by individuals in everyday
life to produce meaningful and differentiable activities and to indicate how they interpret the activities of others.
Ethnomethodology was founded in the 1960s by Harold Garfinkel, who received his doctorate in 1952 for a dissertation entitled The perception of the other:
A study in social order under the direction of the sociologist Talcott Parsons (cf.
Heritage [1984] 1992; Schegloff 1992). While Garfinkel viewed Parsons studies
(alongside those of Alfred Schtz and Edmund Husserl) as crucial in the development of his own perspective on the world of everyday activities (Garfinkel
[1967] 1996: ix), he also insisted on distancing himself from his former mentors
theoretical assumptions about the crucial question of how social order arises. In
Parsons structural functionalism, social order is made possible by the orientation
of the individual to intersubjectively valid norms. According to Parsons, these cultural values and rules serve as a framework for the actions of the individual, who
perceives and interprets social reality on the basis of these shared norms. Garfinkel, however, rejected the notion that the individual is geared to given norms and
asked instead how he goes about interpreting situations. Garfinkel thus shifted the
focal point away from collectively held norms as constitutive of social order to
identify the continuous interpretation and attribution of meaning as central to the
genesis of social order.
Hence in his own analysis of social order, Garfinkel concentrated on the assessments made by ordinary people and the activities they produce which can be perceived by others (Heritage 1992: 736). Garfinkel understood ethnomethodology
in contrast to the then prevalent assumption of an objective reality. In the preface
to his book Studies in Ethnomethodology (1996), he remarks that the objective
reality of social facts should not be seen as given, but rather as the result of an ongoing accomplishment of the members of a society. Social order thus arises out of
262
the interactive activities of the participants. These activities also serve as accounts to explain how the activities of the participants should be understood.
On the basis on these reflections, everyday life became a legitimate subject of
sociological analyses. Moreover, the new focus directed attention at examining
how the order of everyday life, including the world of work, is organized. For as
Garfinkel (1996: 3575) pointed out in the late 1960s, although the familiar common sense world of everyday life occupied a central place as a topic of sociological inquiry, most scholarly literature included little data and few methods which
could be used to investigate the features of familiar scenes. His main concern
was thus to affirm and to (re)discover common sense activities in a societys everyday life as a relevant object of study.
1.2.
Early ethnomethodological investigations
One of the first ethnomethodological studies (see e.g. Turner 1974b; Sudnow
1972) examines the activities which jurors as non-professional adjudicators engage in so that others will recognize them as jurors and take their actions seriously
(Garfinkel 1974: 16). Using tape recordings and interviews with jurors as a basis,
this study focuses on becoming a juror (Garfinkel 1996: 110). This process includes, for example, using rules of the official line instead of the rules of daily
life (Garfinkel 1996: 110113). When describing their actions, the jurors present
their decisions in correspondence with the official rules as right, as based on law
and evidence, not personal preferences or interests. Moreover, they justify their
decisions by stressing the integrative features of the deliberations, thereby
neglecting the anomic or random ones (Garfinkel 1996: 113). These examples
demonstrate how a continuous perspective, that of the jurors, is successively generated.
Another early study about suicidal deaths examines the question of how the
involved parties, in this case coroners, arrive at the diagnosis suicide in cases of
sudden, unnatural death (Garfinkel 1996: 11). This work analyzes the process of
inquiry through which investigators establish what really happened as they provide an account of how a particular person died (Garfinkel 1996: 15). Ethnomethodological studies are thus concerned with how activities of everyday life are organized and how this organization is accomplished through peoples everyday life
activities.
The approach as well as the name ethnomethodology can be traced back to the
study on jurors (see above) which Garfinkel carried out together with Saul Mendlovitz in Chicago during the 1950s before going on to the University of California
at Los Angeles (UCLA) in 1954. When he wrote up the study about the tapes and
interviews, he was looking for a suitable name for the doing methodology as observed in the jurors (Garfinkel 1974: 16). The prefix ethno- is intended to refer to
a societys body of general knowledge that serves the individual as a point of orien-
263
tation. Hence the term ethnomethodology designates the procedures and practices,
that is, ethno-methods, which are valid for the members within a particular society
and which produce structure. While the approach of ethnomethodology is analogous to, for example, ethnobotany, it is not intended to stand for a specific knowledge domain such as, for example, ethno-physics (Garfinkel 1974: 1617). The
genesis of the term highlights one of the main thoughts behind this sociological research approach as laid out in Garfinkels 1967 volume Studies in Ethnomethodology: the search for the everyday methods through which a societys members
structure their activities and make them comprehensible and visible as such should
be oriented towards situations and conditions of everyday life (1996: vii). This direct orientation to the accomplishment of everyday activities was achieved with
the help of tape recordings and ethnographical observations from the field. Basing
their work on common sense activities and daily situations, Garfinkel and other
ethnomethodologists of his day vehemently denounced all forms of standardization of ethnomethodological research and stressed what they were not intending
to do: neither theories, nor advice, nor recipes, but rather the practices and features
of everyday life would be the focus of their investigations (Garfinkel 1996: viii; see
also Turner 1974a).
This avoidance of solely theoretical assumptions to focus on everyday activities represents one of Garfinkels central ideas, which can be linked to classical
American pragmatism: The relevance of real lived experiences, habitual practices,
and language with its indexically and socially constitutive characteristics had already been established in the work of Peirce, Mead, Dewey, and others. Seen from
this perspective, the questions arises as to whether certain elements of Garfinkel
and ethnomethodology can be seen as the fulfillment of the unfinished business
of pragmatism (Emirbayer and Maynard 2006).
In Studies in Ethnomethodology (1996) Garfinkel presents various studies
from the preceding twelve years, thereby demonstrating the possibilities for his
form of analysis of everyday activities. In addition to introductory articles and the
study of the everyday practices of jurors mentioned above, the volume also includes two of Garfinkels best-known areas of investigation: one about a psychiatric clinic and the methods used there to identify patients for treatment, and another known as Agnes. Agnes is a pseudonym for a young woman who was
born and raised as a male person (Garfinkel 1996: 120). At the age of 17, Agnes assumed a female identity by altering her appearance and dressing as a woman; three
years later she underwent a sex-change operation. Garfinkels study uses Agnes
story to illustrate that, contrary to commonly held notions about the fundamental
difference between the sexes, the perception of normal sexuality is actually the
result of a constant situational production which, in Agnes case, was constituted
by the daily management of herself as natural female (1996: 184).
264
1.3.
Basic assumptions of ethnomethodology: making activities accountable
While the establishment of a full-fledged research program was not a concern of

the first ethnomethodologists (Turner 1974a), several of their assumptions can be
seen as central.
As mentioned above, ethnomethodology assumes that the members of a society
provide accounts of their own actions. The main focus of ethnomethodological
studies is to show how people make their activities observable, reportable and visible so that social settings become accountable for their participants. Another basic
thought behind ethnomethology is that the actions themselves indicate how they
should be understood as they are being carried out. Because they constitute both a
part of the situation in progress and a means of interpreting it, these human actions
can be characterized by a basic reflexivity. Hence the context in which an activity
should be understood is locally produced in situ by the activities themselves. This
reflective character of the accounts makes it possible to assign sense and meaning
in spite of the indexicality of the expressions and the vagueness of language. Another central concept in the work of Garfinkel known as indexicality emphasizes
the contextuality of language. Because indexical expressions are so common and
in spite of the efforts of many scholars who are ironically mentioned by Garfinkel
irreplaceable, their properties provide important clues for understanding everyday
actions. Concerned only with objective expressions or context-free language, the
exact sciences describe indexical, that is, contextual expressions as simply
awkward (Garfinkel 1996: 5) in spite of their frequency of occurrence and their
utility; from this perspective, indexical expressions appear to be irremediable nuisances (Garfinkel 1996: 7). Garfinkel, however, is most interested in exploring the
rational properties (1996: 911) of indexical expressions and thus the following
central question: How is it possible to gain understanding in spite of the recognizable and irremediable contexuality of language and human action?
In order to illustrate this contextuality and reflexivity of language and human
activity, Garfinkel had his students conduct experiments in familiar scenes of
everyday affairs known as breaching experiments. He wanted his students to
examine ordinary conversations and episodes to test the degree to which everyday
events are characterized by seen but unnoticed features (Garfinkel 1996: 3649).
Garfinkel borrowed the term seen but unnoticed from the phenomenologist
Alfred Schtz, whose work about the world of everyday life influenced Garfinkel
profoundly (Heritage 1992: 3774). Schtz referred to the background expectancies which the members of a society use to understand and interpret human actions as the attitude of daily life. Garfinkels aim was thus to find out how commonplace scenes become familiar for the members of a society. The experiments
were designed to discover the mutual expectations and assumptions which people
share as a kind of basis. Garfinkels method was to examine what happens when
people depart from the usual everyday order by breaking social rules or norms, or
265
in his words: to start with familiar scenes and ask what can be done to make
trouble (Garfinkel 1996: 37). Garfinkels students were asked to engage someone
in an ordinary conversation and to then respond to indexical phrases which are
normally not further scrutinized by asking the person to clarify their remarks. For
example when asked How are you?, they would respond: How am I in regard to
what? My health, my finances, my school work []? (Garfinkel 1996: 44). Unaware of the experiment, most people reacted to this unexpected change in the normal course of conversation by showing confusion or annoyance or by changing the
subject. It also became clear that one explanation tended to result in another, indicating that indexicality cannot be suspended or healed. Reflexivity, which is, as
mentioned above, constitutive of all actions, normally prevents these kinds of deviations in everyday life and assures that the participants in a conversation alternately receive hints to help them deal with vagueness as a matter of course.
2.
Conversation Analysis
2.1.
Harvey Sacks Lectures on Conversation
These basic assumptions about the production of order through everyday activities
have also been fundamental for conversation analysis (CA), which grew out of ethnomethodology from the mid to late 1960s to became an independent area of research (see e.g. Sudnow 1972; Turner 1974b). Harvey Sacks, who, like Garfinkel,
was at UCLA from 1963, is known as the founder of CA. Through his work, he
paved the way for a growing interest in everyday conversations as an object of
study. Garfinkel and Sacks were engaged in close scientific exchange since they
first met in a seminar by Talcott Parsons at Harvard in 1954 (Schegloff 1992: xiii;
see also Bergmann 2000a). Known as the cornerstone of CA, Lectures on Conversation is composed of lectures which Sacks held between 19641972. Following his early death in 1975, Sacks former student Gail Jefferson edited the lectures
(Sacks 1989, 1992 Vol. I & II). Sacks used recorded conversations (including telephone recordings from an emergency psychiatric hospital) as a basis for intensely
analysing the structural characteristics of verbal interactions. In the course of his
work, he introduced and advanced the position that with respect to both verbal and
nonverbal activities, the generation of social order in everyday life is a legitimate
subject of sociology (Sacks 1972). Topics covered in his extensive analyses include lies to be expected in conversations (e.g. as an answer to the question How
are you?, Sacks 1992, Part V: 8, 9), the techniques used to choose a conversations
next speaker, and the procedures used by members to categorize someone or an
event. These techniques are used automatically by the participants in a conversation and assure, as a rule, two things: that conversations proceed not chaotically,
but according to a certain order, and that each party is able to assign meaning to the
utterances of the other. These order providing mechanisms are the focus of CA.
266
Sacks dealt extensively with questions about the everyday use of categories of
membership (Sacks 1992, inter alii Part I: 6, Part V: 11, 12). In his view, the procedures and devices which members use to categorize others constitute a central
machinery of social organization. Sacks developed an apparatus called MIR device (M for membership, I for inference-rich, R for representative) (1992:
41) which includes three categories of questions: questions about the chosen category sets (e.g. race, sex, age, occupation); questions about the conclusions which
are made possible by the categories used, e.g. baby and mommy in the wellknown example The baby cried. The mommy picked it up (see Sacks 1992,
Part III: 12; Sacks [1972] 1974); and questions about the activities ascribed to
particular members as category bound, e.g. crying to babies and picking up to
mothers.
2.2.
The sequential organization of verbal interaction
Sacks thus took what Garfinkel (and others) discussed about the relevance of
everyday activities for the production of social order and applied it to verbal interactions. CA empirically examines how order is generated in conversations. Proceeding from the basic principle of ongoing accomplishment (see above), the
question of the successive generation of various ordering structures comes into
focus. The processual creation of ordering structures during interactions is conceptualized as doing (see Sacks 1984 on doing being ordinary, Garfinkels analysis
of Agnes from 1996 on doing gender, Aya 2004 on doing interviewing).
A concentration on sequential organization constitutes a central and pivotal
point of analyses in the work of Sacks and within CA. The ever-present question
about the localisation of individual elements of conversation puts the focus on the
direct context of single utterances. By analysing the sequence of utterances generated locally within the situation, it is possible to determine what an utterance refers
to, what problem a particular turn solves, and which turn can be expected to follow.
Sequential organization is the core of the methodical approach of CA, which tries
to identify the processes which give order to conversation. In conjunction with
Emanuel Schegloff, a former fellow student who later became a colleague, and
Gail Jefferson, a former student, Sacks developed various research methods for
studying sequential organization. These methods constitute the centrepiece of CA
and prominently demonstrate how CA examines interactions within the context of
their own order, which is produced turn by turn. These assumptions about the selfreferential and reflexive processes involved in the production of a conversation
make it possible to describe the autonomous nature of interactions and to designate
them as systems (see below).
The first basic structure describes how people take turns in ordinary discourse
according to a turn-taking-machinery which functions automatically during conversation. Analysing this machinery indicates how the simplest systematic for
267
the organization of turn-taking in conversation functions and how the latent problem of choosing the next speaker at a transition relevance place can be solved,
e.g. through self-selection or other selection (Sacks, Schegloff, and Jefferson
[1974] 1978). The local, sequential organization of a conversation underscores
Garfinkels assumptions about the reflexivity of all actions in everyday life: understanding in verbal interaction is possible because the utterance of one speaker reflects and makes accountable his understanding of what was said before. One of the
central methods used by speakers to organize their conversation in the course of
conversation are so-called utterance pairs or adjacency pairs (Schegloff and
Sacks 1973; Schegloff 1972). Because some utterances provoke certain responses,
they are said to operate according to a property called conditional relevance
(inter alii Schegloff 1968), which compels the next speaker to respond in an expectable way in terms of both content and form (as a rule, questions are followed
by answers and greetings are followed by return greetings). Understanding is made
possible not only in the sequence of utterances and the structural order (question/
answer, accusation/defence), but also by virtue of the fact that utterances are retrospectively repaired. The sequential repair of utterances constitutes a special feature
of conversation: an utterance can be readily revised or repaired with respect to
formulations or details in a second turn. As shown in various studies (Schegloff,
Jefferson, and Sacks 1977; Schegloff 1979; Jefferson 1974), repairs can be differentiated according to who initiates them and who actually carries them out. The
analyses indicated that participants prefer self-initiated self-repair (as opposed to
other-repair), which means that the speaker of the utterance requiring repair makes
the correction without prompting from another. This idea has also been developed
in further studies of conversation under the name of a general preference organization and can be described with reference to common adjacency pairs such as
agreement/disagreement (inter alii Pomerantz 1984). In addition to these three central areas (turn-taking, adjacency pairs, repairs), a number of structural characteristics and specific tasks of verbal interaction (such as recipient design, laughter,
opening and ending of conversation, stories) as well as conversation types (everyday conversations, consultations, trouble talks, e.g. Jefferson 1984) have been analyzed since the beginnings of CA (see Schenkein 1978; Psathas 1979; Atkinson
and Heritage 1984; Hutchby and Wooffitt 1998). The initial work of Sacks, Sudnow, Schegloff, and Jefferson in Berkeley and at UCLA placed a sustained focus
on the regularities of everyday speech (also in working life) as well as on the
(ethno)-methods by means of which this verbal order is generated.
2.3.
The analysis of empirical data
Because talk-in-interaction (Schegloff 1987) had not been the subject of such
minute analysis up until this point, another achievement of the first CAs was to develop and promote appropriate methods for archiving and preserving conver-
268
sations (Bergmann 1981, 1985). Using recordings of natural data (initially telephone conversations) as a basis, transcripts were made which attempted to record
entire exchanges word for word. Hence word interruptions, reformulations and
ums were not edited out, but rather became a part of the transcript. Both the recordings and the transcripts provide the data used in the analysis of authentic conversations (Hutchby and Woofitt 1998; ten Have 2007). Both then and now, these
natural data stand in strong contrast to data which have been contrived or generated
in experimental settings (inter alii Heritage 2001). With the help of sound recordings and, more recently, video recordings (on this topic see Heath and Luff 2000;
Goodwin 2000), CA preserves interactions by recording them for future use. These
data transcripts then form the basis for exploring the central issue of ordering structures. Seen from this perspective, reconstructive methods (e.g. narrative interviews, memorized observations) commonly used prior to CA (and even today) do
not constitute data which have been recorded verbatim from an actual situation.
Already tainted by the perspective of the one reporting it, data collected in this way
cannot provide a basis for analyzing the genuine course of interactions (Bergmann
1985).
The strictly empirical approach of CA starts from the data themselves
(Schegloff and Sacks 1973: 291) in order to reconstruct the mechanisms und
methods through which the participants of verbal interaction generate turn-by-turn
order. This focus on the course of the interactions themselves disaffirms the practice of arbitrarily applying a methodological or normative apparatus which is always supposed to be universally valid. Just as ethnomethodologists reject theoretical claims and programmatic assertions (see Turner 1974b; Garfinkel 1996), there
is a consensus within CA about the inappropriateness of these kinds of standardizations. In the view of CA, verbal interactions should be studied using methods
that arise out of the analysis of the data: to use hypotheses or theoretical links as a
starting point would be to narrow ones perspective on the order of a given set of
interactions. As it seeks to identify the specific features of a particular interaction
using turn-by-turn analysis of the data, this position can also be understood as a
kind of analytic mentality (Schenkein 1978: 6).
This absolute concentration on the order of interaction as recorded by the transcript (which now includes nonverbal elements as well) emphasizes the special
empiric understanding of CA and its perspective on interactions. The fact that the
methods of the participants are concretely reconstructed from the data themselves not only makes it possible to study authentic processes by reconstructing
the creation of order and the sequential structure of talk turn by turn. This method
of analysis with its step by step procedure also highlights the autonomy of interaction. In other words, the particular empiric understanding points to certain assumptions about the object, the interaction itself. With each successive turn, the order of
interaction shows which options emerge for the participants, what the interaction
requires, and what interactively appears as relevant. From this perspective, self-
269
referential and reflexive processes of interaction move to the centre of CA. This
focus on the autonomy of interaction makes it possible to connect CA with more
recent system theory (Luhmann 1984, 1997), which presupposes the self-referential and autopoietic organization of a system (see Hausendorf 1992; Schneider
1997; Domke 2006). These interconnections correspond to CAs fundamental interest in interaction and open up empirical links to system theorys strictly theoretically based thoughts on the formation of systems.
One of the leading achievements and characteristics of CA is the way in which
it unlike mechanistic communication models points out the internal dynamics
of interactions. However, this focus on interaction as an autonomous system also
comprises a great challenge: the context of the interaction and the participants
must be taken into account appropriately and methodically examined. It is in these
points that CA differs from Erving Goffmans perspective on interactions.
3.
Erving Goffmans Interaction Order
Alongside Garfinkels ethnomethodology, Erving Goffmans interaction analysis

paved the way for numerous studies on the subject of everyday interactions and
analyses of how social order is generated by these interactions. Goffman taught at
the University of California at Berkeley from 19571968 before taking a position at
the University of Pennsylvania. His first book, The Presentation of Self in Everyday
Life was published in 1959 and is considered to be one of his best-known works.
Goffman had a great influence on Harvey Sacks, who studied with Goffman in Berkeley before going on to UCLA. In his introductions to Sacks lectures, Schegloff
emphasises that both sociologists influenced each other, an aspect which needs to
be explored more closely (Schegloff 1992: xxii). The special nature of their relationship is underlined by Goffmans reported answer to the question of whether
Sacks studied under him: What do you mean; I was his student! (Schegloff 1992:
xxiii). In spite of numerous ties and points of common ground (Drew and Wootton
1988; Widmer 1991) Goffmans work can neither be subsumed under the name of
ethnomethodology nor that of CA, but rather occupies an independent area of empirical social research with its own form of analysing everyday activities. However,
this does not mean that a Goffman School or methodology actually developed
out of his studies. It was primarily by pointing out the crucial role and function of
the interaction order as an institution in its own right (Goffman 1983b) that Goffman came to have such an important influence on conversation analysts like Sacks
and Schegloff (Heritage 2001; see also Bergmann 1991, 2000b).
According to Goffman, the features of traffic rules for pedestrians can be
studied in a crowded kitchen as well as anywhere else, and distinguishing between public and private settings is not crucial for understanding social interactions (Goffman 1983b). In his view, solely an understanding of face-to-face inter-
270
action as an independent domain with independent structures is necessary for the

analysis of everyday situations in the presence of others. However, this does not
mean in reverse that it can be effective to analyse the structure and rules of conversations out of context. On the contrary: in his studies on interaction order (initially
Interaction Ritual, 1967), Goffman focuses on the rituals and processes (such as
deference and demeanor, see also 1967) which participants adhere to in their
everyday presentations of self in order to create face and engage in moral enterprises (Drew and Wootton 1988: 67). From this perspective, the importance of
interactions for the formation of socially relevant structures (such as face, role) becomes clear. In order to differentiate social situations, Goffman uses, among
others, the term encounter to describe participants engaged in copresence-based,
mutually ratifying interaction, and the term gathering for unfocused situations
(such as waiting for a train) (Goffman 1964, 1983b).
Goffmans analyses work under the assumption that the individual self is created and shaped in every everyday face-to-face interactions. In his early essays
On Cooling the Mark Out (Goffman 1952) and On Face-work (Goffman
1955), Goffman introduces the concept of face-work as an essential process of
presenting the self and interacting with others. He developed the concept further in
later studies. For example in Stigma (1963b), he explores how persons who deviate
from normality (such as the handicapped or homosexuals) try to come to terms
with their threatened identity in everyday life. The exclusion of the individual in
public places, e.g. via the claiming of territories (such as seats or phone boxes) is
the focus of Behaviour in Public Places (Goffman 1963a). Goffman uses central
concepts from the theater to describe the production of self (Goffman 1959). Like
actors in a play, people constantly engage in impression management in order to
embody a certain role for their audience. This everyday dramaturgy includes
various methods and attributes such as the front stage, where the performers play
their roles in full view of the audience, and the backstage, a less accessible area
which is not available for public viewing.
The key points outlined above indicate a fundamental difference between Goffmans ideas and the core of CA. In CA, the focus is on the sequential structure of
interaction and the empirical analysis of interaction structures. This approach
brings the autonomy of interaction to the fore, which allows for an understanding
of interactions as systems as described in 2.3. Goffman, by contrast, is more interested in the individual and the effects of situations on his activities. The pertinent
concepts of face and ritual put the focus on the individual in relation to his surroundings. For Goffman, interaction within the structures of these surroundings is
less relevant than the individual participant in his emerging social identity. Hence,
by looking at the pertinence of interaction for both CA and Goffman, significant
differences can be identified. In light of the mutual influence between CA and
Goffman, these differences have prompted numerous attempts to explain and distinguish the two approaches (see Bergmann 1991).
271
Goffman addresses linguistic systems and dialogic structures more concretely

(see Bergmann 1991) in his later works. The essays collected in Forms of Talk
(Goffman 1981) identify the various possibilities for differentiating social situations and encounters and discuss the role of the participants. Of particular relevance are his thoughts on participation status and participation framework (Goffman 1981: 137), concepts which stand in contrast to the rigid ascriptions of a
speaker/hearer model (see Auer 1999: 155163; Bergmann 1991: 314317). In one
of the essays entitled Footing, which first appeared in 1979 in the magazine Semiotica, Goffman describes a range of different types of speakers and hearers
(Goffman 1981: 124157). The participant who is engaged in talking is at that moment merely the sounding box in use (Goffman 1981: 144); he or she functions
as an animator. The animator and the author of what is said are not necessarily
identical and must be differentiated. A third function is that of the principal, the
one who is responsible for what is said. Taken together, these ascriptions reveal
what Goffman refers to as the production format of an utterance (Goffman 1981:
145). Hearers can also be differentiated into different modes (Goffman 1981:
131137). There are two kinds of ratified hearers: those who are directly addressed
by the speaker and those who are unaddressed. Non-ratified participants may be
listening intentionally, which Goffman calls eavesdropping, or unintentionally,
which he refers to as overhearing. All of these elements constitute the footing of an
interaction. Displayed interactively, footing is characterized by constant change.
The footing of an interaction also acts as a kind of orientation aid similar to the
more general frame of events, a concept which Goffman developed in Frame
Analysis (Goffman 1974) with reference to the earlier work of Bateson. For Goffman, frames involves definitions of situations and cognitive structures which
guide participants and indicate what might be expected of them. Frame analysis
(e.g. Goffman 1974) includes Goffmans analyses of response cries, a term which
was also the title of an article (Goffman 1978) later published in Forms of Talk.
Speakers produce exclamatory utterances like Oops! or Brr! in everyday life
as self talk without directly addressing these interjections to a particular person.
Using his distinction between social situation and encounter (see above), Goffman
describes these response cries as creatures of social situation, not states of talk
(Goffman 1981: 121) and thus identifies them as conventionalised expressions and
not as primitive, unsocialised (Goffman 1981: 120. The open state of talk is yet
another of Goffmans important designations which is useful for describing certain
communicative situations. An open state designates a conversation in which the
participants have the right but not the obligation to talk, as is the case when individuals are involved in a joint physical task or a common meal (Goffman 1981:
134135; also 1983a).
272
4.
Recent developments
Taking the fundamental studies described here as a starting point, the analysis of
verbal interaction has developed further in many respects. In the following, a few of
the most prominent out of a large number of works will be mentioned, starting with
the studies of work which grew out of ethnomethodology. In the introduction to a
volume of studies (1986) which he edited, Garfinkel emphasises that the analysis of
work processes (e.g. within a jazz ensemble studied by Harvey Sacks) makes it
clear that there exists a locally produced order of works things (Garfinkel 1986:
vii) which has largely been ignored. The specific features of individual occupational activities have become the focus of studies on everyday practices in the working world including both verbal and nonverbal activities (see Bergmann 2000b).
So-called workplace studies, which concentrate primarily on the relationship
between society and technology, have also been influenced by studies of work
(Knoblauch and Heath 1999). This area of research focuses on computer supported
work environments and the consequences of the mechanisation of various processes (inter alii Heath and Luff 2000). The next step will be to extend the methods
used in these studies to carry out large-scale document and video analyses; this endeavour will provide adequate material for future works of conversation analysis.
Other important studies which were motivated by conversation analysis analyse the characteristics of work realms or specific forms of interactions by focusing
on the verbal component (see Drew and Heritage 1992 on Talk at work; Clayman
and Heritage 2002 on News Interview). In more recent studies, this concentration on speech has given way to a more holistic approach (Dausendschn-Gay
and Krafft 2002), which tries to incorporate facial expressions, gestures and the environment of interactions into each analysis. Conversation analytical work with
video recordings has led to a discussion of classical CA focus areas (such as turntaking, conversational openings) and elaborated the interplay and necessary coordination between speech, gesture and objects for various areas including medicine,
everyday conversations and the media (inter alii Goodwin 2000; Aya 2004; Kissmann 2009; Schmitt 2007; Streeck 2009).
5.
Influence of conversation analysis on linguistic pragmatics
Conversation analysis is one of the two cornerstones of pragmalinguistics. In retrospect, it appears to be the less prevalent of the two, but in the long run it may be the
more sustainable. In simplistic terms, one could say that linguistic pragmatics has
essentially followed two major lines in its reorientation. On the one hand, it has
taken its impulse from the philosophy of language, incorporating and integrating
not only ordinary language philosophy along with the speech act theory of Austin
and Searle, but also the concept of meaning and understanding as explicated by
273
their attributed intellectual father Ludwig Wittgenstein and further developed by

Grice. On the other hand, linguistic pragmatics has been influenced by sociological
impulses which were mediated chiefly through conversation analysis and its historical backdrop, including names ranging from Harold Garfinkel and Alfred
Schtz to Max Weber. In addition, the microsociological studies of Erving Goffman with their direct relevance to every kind of interaction have impacted the field
to a lesser degree.
No linguist working in the field of pragmatics has been able to avoid coming to
grips with speech act theory and its string of tradition because speech act is the
central concept that has defined not only the whole linguistic subdiscipline, but
also the refounding of the entire field of linguistics. The study of conversation
analysis (and especially Goffman), however, has been less obligatory for two reasons. First, because it applies only to verbal interaction, CA was thought to represent a limited subject area. Secondly, CA was in competition with other approaches which also attempted to analyze spoken interaction or which did not
make a rigorous distinction between (spoken) interactions and (written) texts. Another possible explanation for the resistance to CA may be the challenges it posed
to adherents of speech act theory from the philosophy of language who were accustomed to working with constructed and isolated example sentences. The consistent empirical approach to collecting data and diligently transcribing them
required a fundamentally different kind of methodology than was required by
grammar-oriented research. Finally, many were put off by what they saw as a kind
of unorthodox leftist movement, a reputation which CA earned with its wild
1960s California image, its antitheoretical mistrust of academic omniscience, and
its uniquely different perspective: the adherents of CA were searching for the
methods of the members of society themselves. Others, however, were attracted
by this unconventionality. For example, Bergmann (1981: 38) admires das Unsystematische und Unabgeschlossene, das Ungeschtzte und Riskant-Artistische, das
Fragmentarische, Labyrinthhafte und Antiakademische [the unsystematic and inconclusive, the unprotected and risky-artistic, the fragmentary, labyrinthian and
anti-academic]. In any case, it cannot be denied that the actual subject matter of CA
was paradoxically just what the more theoretical and less disputed branch of pragmatics actually had in mind without analyzing it empirically: ordinary language.
While CA researchers repeatedly denied having a primary or special interest in language and conversation (Schegloff and Sacks 1973, cited in Coulthard 1977: 52;
see also Sacks comment as quoted in Auer 1999: 137), they nonetheless were
compelled to and wanted to attract interest by designating themselves as conversation analysts, especially in light of the new pragmatic orientation of the field
of linguistics and the relevance and central importance of their subject matter, as
highlighted here by Levinson: It is not hard to see why one should look to conversation for insight into pragmatic phenomena, for conversation is clearly the
prototypical kind of language usage [] (1983: 284).
274
The gradual triumph of CA, particularly in the field of linguistics, was thus inevitable to a certain extent. The appearance of important studies by sociological researchers in linguistic journals (Sacks, Schegloff, and Jefferson 1978; Schegloff,
Jefferson, and Sacks 1977 in Language) or in a socioliguistic context (e.g. Sacks
1974, first in Gumperz and Hymes 1972) was no accident. The proximity and
kinship to newly developing sociolinguistic approaches (e.g. the ethnography of
speaking) created a situation in which the various communication-based approaches benefited each other in an area of overlap between sociology, anthropology, and linguistics; this interconnectedness also strengthened interest in works on
conversation analysis. Recognition of CA expanded beyond a small circle of interested scholars to the entire (Anglo-Saxon) field of linguistics after the publication
of the two chapters The ethnography of speaking and Conversation analysis
(following the chapter entitled Speech acts) in An Introduction to Discourse
Analysis by Malcolm Coulthard (1977). Coulthard founded the Birmingham
School together with John Sinclair on the basis of their study of classroom interaction (Sinclair and Coulthard 1975).
Within the field of German linguistics, two research structures in Bielefeld and
Freiburg were primarily responsible for the spread of studies in conversation
analysis. As early as 1973, researchers in Bielefeld published two volumes of texts
by American sociologists, including studies by Garfinkel, Sacks und Psathas on
ethnomethodology and by Hymes on the ethnography of speaking (Arbeitsgruppe
Bielefelder Soziologen 1973). The work of the group in Bielefeld also led to the
publication of a highly regarded article by the linguist Werner Kallmeyer and the
sociologist Fritz Schtze on Conversation Analysis (Kallmeyer and Schtze
1976). The Freiburg project Dialogstrukturen [dialogue structures] from 1974
under the direction of Hugo Steger investigated spoken language, thereby incorporating CA research in its investigation of spoken language (Berens et al. 1976; for
details see Schwitalla 2001a). Thanks to this early engagement in the field, the
1980 Annual Conference of the Institute for German Language on the topic of dialogue research included an overview article on ethnomethodological conversation
analysis (Bergmann 1981). This topic was also covered in later publications, including German guides to dialogue analysis and conversation linguistics (Bergmann 1994, 2001) as well as an introduction to linguistic interaction by Auer
(1999) on the basis of several classic authors in the field.
The research paradigm conversation analysis was effectively linguistically
canonized in the 1980s with the publication of the chapter on conversational structure in Pragmatics by Stephen Levinson (1983), the internationally best-known
book on the topic. Levinson deals extensively with the methods and results of CA
in this chapter, and similar introductions published later follow his example (e.g.
Green 1989; Mey 1993; Grundy 1995; Meibauer 1999). Levinsons work also
served as a model for other writers because he discusses CA in contrast to other approaches, which are combined under the label discourse analysis (DA).
6.
275
Conversation analysis vs. other approaches in the linguistics

of spoken interaction
While Levinson advocates CA, others have taken a position between reservation
and rejection. As a whole, the field of discourse analysis is quite diverse, especially
since it sometimes includes analyses of both written texts (not considered here) and
larger units such as discourse (in the sense of Foucault). This discussion, however,
will continue to focus on spoken language. Alongside the previously mentioned
approaches (Dell Hymes ethnography of speaking, Sinclair and Coulthards Birmingham School), John Gumperz developed his interactional sociolinguistics;
Levinson also cites Van Dijk (1972) and Labov and Fanshel (1977) as typical
representatives of discourse analysis; Bublitz (1991) and Ventola (2001) provide a
general overview of conversation and discourse analysis, and Hausendorf discusses the situation in German-speaking regions (2001). Some approaches are
based more strongly on speech acts: in Germany, for example dialogue analysis
and dialogue grammar (e.g. Hundsnurscher 1995, 2001; Hindelang 1994). The
volume Handbuch der Dialoganalyse [Handbook of Dialogue Analysis] (Fritz and
Hundsnurscher 1994) includes a discussion of the Geneva model of discourse
analysis (Moeschler 1994, 2001) and an action theoretically based concept of
linguistic communication analysis (Gloning 1994), which is associated with
Heringers practical semantics. Konrad Ehlich and Jochen Rehbein take their
own separate direction with their functional pragmatics in Diskursanalyse
(Rehbein 2001). Yet another example of DA is the critical discourse analysis
(Kritische Diskursanalyse) of the Wiener Schule under Ruth Wodak (e.g.
Wodak et al. 1990), an approach which is based on Faircloughs Critical Discourse Analysis (CDA) (e.g. Fairclough 1995).
Writing from the perspective of CA, Levinson (1983: 286294) targets text
grammatical or speech act theoretical approaches of DA with his criticisms of
methods and theoretical tools (Levinson 1983: 294). Claiming that DA sometimes falls back on constructed examples, he criticizes a lack of empirical focus as
well as a fundamentally intuitive approach which ultimately cannot tolerate the
possibility of falsifiability. He also takes issue with DAs acceptance of the familiar
theoretical principles of well-formedness (see also Kohrt 1986: 7475), a viewpoint which sees discourses as nothing more than strings of interlinked sentences.
In the eyes of DAs critics, this oversimplifying grasp (Kohrt 1986: 80) can
easily lead to the construction of rigid or even normative structures (Bublitz 1991:
271). According to Levinson, the practice of assigning individual utterances to one
particular speech act and then constructing structurally definable sequence mechanisms (as in sentences) is bound to fail because it cannot do justice to the complex
reality of the unpredictable dynamic interactions which are themselves embedded
in complex contexts (Levinson 1983: 288294). Proponents of CA defend their
method by pointing out that it is strictly empirical and inductive, with as little ap-
276
peal as possible to intuitive judgements (Levinson 1983: 287), and that it does not
immediately categorize individual cases, but rather attempts to compile functionally similar cases from the data in order to prove the validity of an analysis (Bergmann 2001: 923). The most effective argument for CA is definitely its indisputable
success in the recognition of the structures which organize speech.
However, CA has also had its critics. For example, Levinson (1983: 287)
makes reference to Labov and Fanshel (1977: 25) and Coulthard and Brazil (1979)
as he points out: DA theorists can accuse CA practitioners of being inexplicit, or
worse, plain muddled, about the theories and conceptual categories they are actually employing in analysis. It is indeed questionable whether all existing order
can be deduced solely from the object of investigation. In other words, it is doubtful that utterances are each reflexive and indexical enough to be understood without being categorized by the interpreter from the outside. In fact, it is commonly
recognized that by drawing on his or her analytical mentality, the ethnomethodologist is forced to draw repeatedly on the intuitive understanding which he or she
possesses as a competent member of a linguistic community (Bergmann 1981: 23).
At the same time, this demonstrative refusal to draw on additional outside knowledge in an attempt to immunize CA against possible allegations of interpretive arbitrariness has not only led several critics to label the field as anti-psychological
and individual-psychological (Redder 1990: 7), but it has also caused some to
cross the line to (unjustifiably) accuse CA of secret positivism (Flader and Trotha 1988). A further criticism cites an alleged failure to take social background into
account (e.g. Fairclough 1989: 12; see also Bublitz 1991: 271), but this assessment
is not particularly convincing in view of CAs sociological and sociocritical origins. Moreover, more recently, a growing number of studies in CA have focused on
institutional and work contexts as well as political communication.
On the whole, the field of linguistic pragmatics has not really seen an intense
dispute over theoretical and methodological issues surrounding CA; the predominant pattern has been rather a more or less reflected and more or less selective appropriation of methods and results. Evidence of these integrative tendencies can be
found in many textbooks and introductions, but they arise above all out of the experiences and needs of concrete research. Hence at the end of the 1990s, Hausendorf (2001: 977) came to the following conclusion with respect to the situation in
the German-speaking realm: To the extent to which the methodological praxis of
conversation analysis consolidated itself and the findings and conclusions of empirically sophisticated i.e. complex and long-term projects were presented in
the course of the 1990s, the methodological struggle between the various approaches has faded away.
7.
277
Influence of Erving Goffman on linguistic pragmatics
While CA quickly attained prominence within linguistic pragmatics on account of

its direct relevance for the study of spoken language, the linguistic importance of
Goffmans research initially appeared to be limited to marginal questions. However, his work has been widely recognized and discussed within many other disciplines because of its broad spectrum of topics (self-portrayal, asylums, interaction
rituals, role distance, gender advertising, frame analysis), so that gradually, his
studies have also entered the field of linguistics indirectly via other routes. At least
four strands of Goffmans work have been adopted by pragmatics in one form or
another over the course of time.
First, Goffmans concept of face was further developed by Brown and Levinson ([1978] 1987) into a politeness model with universal applicability. This model
was adopted in numerous studies on the topic of language and politeness, but it
was also vehemently criticized for its weak empirical basis of single utterances as
well as for the questionable way in which the model was universalized (for summary see Fraser 2001; Bublitz 2009: 270). However, the complexity of Goffmans
concept (see Tracy 1990) comprises at its core something that goes beyond mere
aspects of politeness to encompass the ritual negotiation of identities and relationships between interaction partners. This second Goffmanian strand was received
by conversation analysis quite early in a few instances (e.g. Holly 1979, 2001),
but it has taken more time to develop the concept fully for linguistics with respect
to routines (Coulmas 1981), rituals (e.g. Rauch 1992; Werlen 2001), conflict
settlement (e.g. Schank and Schwitalla 1987; Schwitalla 2001b) and further issues. The third important strand constitutes Goffmans frame analysis, which has
been applied more or less prominently in many pragmalinguistic studies (e.g.
Tannen 1984). Finally, the fourth strand involves Goffmans explicit investigations of conversations (Forms of Talk), in particular the deconstruction of
participant roles. This deconstructive aspect of Goffmans work was adopted by
Levinson (1988) and formed the basis of a new conceptualization of the idealized
and simplistic speaker-hearer-constellation (see Schwitalla 2001c), e.g. to analyse multiple addressing (Khn 1995) or testimonials in a television commercial
(Holly 2007). Moreover Goffmans analysis of different intensities of focus on
talk generated the useful concept of open state of talk, which is connected to
Bhlers observations on action-accompanying talk and is also essential for the
analysis of certain uses of speech (see e.g. on the reception of television texts
Holly, Pschel, and Bergmann 2001). The fact that four articles in the Handbook
of Linguistics of Text and Conversation focus primarily on Goffmanian concepts
(Werlen 2001; Schwitalla 2001b, 2001c; Holly 2001) is an index of how widely
Goffmans research has been received in the field of pragmalinguistics. Nonetheless, his importance for the analysis of spoken language continues to be underestimated in pragmalinguistics, probably because of his independence as a scholar,
278
which fostered a mentality against the formation of schools, and his conceptual
flexibility.
8.
The convergence of various approaches
Although ethnomethodological conversation analysis constitutes a markedly independent and distinctive school of thought in the study of spoken language, there is
a gradual tendency toward increasing convergence with other related approaches
which are anchored more firmly in the fields of anthropology, sociology or linguistics; Heritage (2001) combines the three under the label ethno-sciences.
These ethno-sciences share in common not only work with empirical data, but also
a series of core ideas which he characterizes in the following:
This research embodies the core notions that (i) communicative meaning is inherently
contextual in character, (ii) social context is unavoidably dynamic and is managed
through the participants actions, (iii) the specific contextual significance of actions is
structurally achieved by means of rules and practices of conduct which are systematically related and organized as systems, (iv) the contextual significance of action also involves a web of inferences which are inescapable, very often involve personal, moral
and social accountability, and thus connect interaction with culture, social structure and
personality, and (v) all this is managed through the integrated significance of talk, paralanguage and body movement. (Heritage 2001: 916)
With these kinds of common convictions as a basis, infighting, which also has seen
adherents of CA criticize a less than rigorous handling of methodological questions (Streeck 1983), could relax. At the same time, this convergence is being promoted in the field of linguistics by an increasing return to the original competencies, resulting in a common focus on linkages between grammatical and interactive
structures (Uhmann 1997; Ochs, Schegloff, and Thompson 1997), in particular
with reference to prosodic-intonational structures as foreseen in the concept of
interactional linguistics by Elisabeth Couper-Kuhlen und Margret Selting. This
approach sees itself as an interface between linguistics und interaction theory
(above all functional linguistics, conversation analysis, contextualization theory,
anthropological linguistics) (Selting and Couper-Kuhlen 2000: 76). Other efforts
concentrate more closely on the connection between grammar and interaction (e.g.
Deppermann, Fiehler, and Spranz-Fogasy 2006), so that the discipline of conversation linguistics reencounters many older orientations in the investigation of
spoken language (see Schwitalla 2001a), be it on the basis of Ehlich und Rehbeins
Functional Pragmatics which was originally intended to focus on grammar as
well, or on the basis of construction grammar (Gnthner and Imo 2006; Deppermann 2007).
279
References
Arbeitsgruppe Bielefelder Soziologen (ed.)
1973
Alltagswissen, Interaktion und gesellschaftliche Wirklichkeit. 2 vols. Reinbek:
Rowohlt.
Atkinson, John Maxwell and John Heritage (eds.)
1984
Structures of Social Action: Studies in Conversation Analysis. Cambridge:
Auer, Peter
1999
Sprachliche Interaktion. Eine Einfhrung anhand von 22 Klassikern. Tbingen: Niemeyer.
Aya, Ruth
2004
Konversationsanalytische Medienforschung. Medien und Kommunikationswissenschaft, 52 (1): 529.
Berens, Franz-Josef, Karl-Heinz Jger, Gerd Schank and Johannes Schwitalla
1976
Projekt Dialogstrukturen. Ein Arbeitsbericht. Mnchen: Hueber.
Bergmann, Jrg R.
1981
Ethnomethodologische Konversationsanalyse. In: Peter Schrder and Hugo
Steger (eds.), Dialogforschung. (Jahrbuch 1980 des Instituts fr deutsche
Sprache), 953. Dsseldorf: Schwann.
Bergmann, Jrg R.
1985
Flchtigkeit und methodische Fixierung sozialer Wirklichkeit. Aufzeichnungen als Daten der interpretativen Soziologie. In: Wolfgang Bon and Heinz
Hartmann (eds.), Entzauberte Wissenschaft. Zur Relativitt und Geltung soziologischer Forschung, 299321. Gttingen: Schwartz.
Bergmann, Jrg R.
1991
ber Erving Goffmans Soziologie des Gesprchs und seine ambivalente Beziehung zur Konversationsanalyse. In: Robert Hettlage and Karl Lenz (eds.),
Erving Goffman ein soziologischer Klassiker der zweiten Generation, 301
326. Bern/Stuttgart: Haupt.
Bergmann, Jrg R.
1994
Ethnomethodologische Konversationsanalyse. In: Gerd Fritz and Franz
Hundsnurscher (eds.), Handbuch der Dialoganalyse, 316. Tbingen: Niemeyer.
Bergmann, Jrg R.
2000a
Harold Garfinkel und Harvey Sacks. In: Uwe Flick, Ernst von Kardorff and
Ines Steinke (eds.), Qualitative Forschung. Ein Handbuch, 5162. Reinbek:
Rowohlt.
Bergmann, Jrg R.
2000b
Ethnomethodologie. In: Uwe Flick, Ernst von Kardoff and Ines Steinke (eds.),
Qualitativeforschung. Ein Handbuch, 118135. Reinbek: Rowohlt.
Bergmann, Jrg R.
2001
Das Konzept der Konversationsanalyse. In: Brinker et al. (eds.), 919927.
Brinker, Klaus, Gerd Antos, Wolfgang Heinemann and Sven Frederik Sager (eds.)
2001
Text- und Gesprchslinguistik. 2nd vol. (Handbcher zur Sprach- und Kommunikationswissenschaft 16.) Berlin/New York: de Gruyter.
280

1987
Politeness. Some Universals in Language Usage. Cambridge: Cambridge University Press. First published in: Esther N. Goody (ed.), Questions and Politeness, 56289. Cambridge: Cambridge University Press 1978.
Bublitz, Wolfram
1991
Discourse Analysis: The State of the Art. In: Claus Uhlig and Rdiger
Zimmermann (eds.), Anglistentag 1990 Marburg. Proceedings, 259284. Tbingen: Niemeyer.
Bublitz, Wolfram
2009
Englische Pragmatik. Eine Einfhrung. 2nd ed. Berlin: Erich Schmidt Verlag.
Clayman, Steve and John Heritage
2002
The News Interview. Journalists and Public Figures on the Air. Cambridge:
Coulmas, Florian
1981
Routine im Gesprch. Zur pragmatischen Fundierung der Idiomatik. Wiesbaden: Athenaion.
Coulthard, Malcolm
1977
An Introduction to Discourse Analysis. London: Longman.
Coulthard, Malcolm and David Brazil
1979
Exchange Structure. Birmingham: Birmingham University.
Dausendschn-Gay, Ulrich and Ulrich Krafft
2002
Text und Krpergesten: Beobachtung zur holistischen Organisation der Kommunikation. Psychologie und Sozialwissenschaft. Zeitschrift fr Qualitative
Forschung. Themenheft: Sprechen vom Krper Sprechen mit dem Krper 4:
3060.
Deppermann, Arnulf
2007
Grammatik und Semantik aus gesprchsanalytischer Sicht. Berlin/New York:
de Gruyter.
Deppermann, Arnulf, Reinhard Fiehler and Thomas Spranz-Fogasy (eds.)
2006
Grammatik und Interaktion. Untersuchungen zum Zusammenhang zwischen
grammatischen Strukturen und Interaktionsprozessen. Radolfzell: Verlag fr
Gesprchsforschung.
Domke, Christine
2006
Besprechungen als organisationale Entscheidungskommunikation. Berlin/
New York: de Gruyter.
Drew, Paul and John Heritage (eds.)
1992
Talk at work: interaction in institutional settings. Cambridge: Cambridge University Press.
Drew, Paul and Anthony Wootton (eds.)
1988
Erving Goffman. Exploring the interaction order. Cambridge: Polity Press. Reprinted Cambridge: Polity Press [1996].
Emirbayer, Mustafa and Douglas Maynard
2006
Pragmatism and Ethnomethodology. Paper presented at the annual meeting of
the American Sociological Association, Montreal Convention Center. Montreal, Canada: Online <PDF>. 2009-05-24 from http://www.allacademic.com/
meta/p95497_index.html.
281
Fairclough, Norman
1989
Language and Power. London: Addison Wesley.
Fairclough, Norman
1995
Critical Discourse Analysis: The critical study of language. London: Longman.
Flader, Dieter and Thilo von Trotha
1988
ber den geheimen Positivismus und andere Eigentmlichkeiten der ethnomethodologischen Konversationsanalyse. Zeitschrift fr Sprachwissenschaft 7:
92115.
Fraser, Bruce
2001
The form and function of politeness in conversation. In: Brinker et al. (eds.),
14061425.
Fritz, Gerd and Franz Hundsnurscher (eds.)
1994
Handbuch der Dialoganalyse. Tbingen: Niemeyer.
Garfinkel, Harold
1952
The Perception of the Other: A Study in Social Order. Ph.D. Dissertation. Boston: Harvard University.
Garfinkel, Harold
1974
The Origins of the Term Etnomethodology. In: Turner, Roy (ed.), Ethnomethodology. Selected Readings, 1519. Harmondsworth: Penguin Education.
Garfinkel, Harold (ed.)
1986
Ethnomethodological studies of work. London: Routledge and Kegan Paul.
Garfinkel, Harold
1996
Studies in ethnomethodology. Cambridge: Polity Press. First published Englewood Cliffs, NJ: Prentice Hall [1967].
Gloning, Thomas
1994
Praktische Semantik und Linguistische Kommunikationsanalyse. In: Fritz et
al. (eds.), 113129.
Goffman, Erving
1952
On Cooling the Mark Out: Some Aspects of Adaption to Failure. Psychiatry 15
(4), 451463. Reprinted in: Charles Lemert and Ann Branaman (eds.), The
Goffman Reader, 420. Oxford: Blackwell [1997].
Goffman, Erving
1955
On Face-work: An Analysis of Ritual Elements in Social Interaction. Psychiatry 18 (3), 213231.
Goffman, Erving
1959
The Presentation of Self in Everyday Life. Garden City, NY: Doubleday, Anchor Books.
Goffman, Erving
1963a
Behaviour in Public Places: Notes on the Social Organization of Gatherings.
New York: The Free Press.
Goffman, Erving
1963b
Stigma: Notes on the Management of Spoiled Identity. Engelwood Cliffs, NJ:
Prentice Hall.
Goffman, Erving
1964
The Neglected Situation. American Anthropologist 66: 133136. Reprinted in
Charles Lemert and Ann Branaman (eds.), The Goffman Reader, 229233. Oxford: Blackwell [1997].
282
Goffman, Erving
1967
Interactional Ritual: Essays on Face-to-Face Behaviour. Garden City, NY:
Doubleday, Anchor Books.
Goffman, Erving
1974
Frame Analyses. An Essay on the Organization of Experience. London: Harper
and Row.
Goffman, Erving
1978
Response Cries. Language 54 (4): 787815. Reprinted in Erving Goffman,
Forms of talk, 78122. Philadelphia: University of Pennsylvania [1981].
Goffman, Erving
1979
Footing. Semiotica 25 (1/2): 129. Reprinted in Erving Goffman, Forms of
talk, 124159. Philadelphia: University of Pennsylvania [1981].
Goffman, Erving
1981
Forms of talk. Philadelphia: University of Pennsylvania Press.
Goffman, Erving
1983a
Felicitys Condition. American Journal of Sociology 89 (1): 151. Reprinted in
Goffman, Erving
1983b
The Interaction Order. American Sociological Review 48: 117. Reprinted in
Goodwin, Charles
2000
Practices of seeing: visual analysis an ethnomethodological approach. In:
Theo van Leeuwen and Carey Jewitt (eds.), Handbook of visual analysis,
157182. London: Sage.
Green, Georgia M.
1989
Erlbaum.
Grundy, Peter
1995
Doing Pragmatics. London: Arnold.
Gumperz, John and Dell Hymes (eds.)
1972
Directions in Sociolinguistics: The Ethnography of Communication. New
York/London: Holt, Rinehart and Winston.
Gnthner, Susanne and Wolfgang Imo (eds.)
2006
Konstruktionen in der Interaktion. Berlin/New York: de Gruyter.
Hausendorf, Heiko
1992
Gesprch als System. Linguistische Aspekte einer Soziologie der Interaktion.
Hausendorf, Heiko
2001
Gesprchsanalyse im deutschsprachigen Raum. In: Brinker et al. (eds.), 971
979.
Have, Paul ten
2007
Doing Conversation Analysis. A Practical Guide. 2nd ed. Los Angeles: Sage.
Heath, Christian and Paul Luff
2000
Technology in action. Cambridge: Cambridge University Press.
283
Heritage, John
1992
Garfinkel and ethnomethodology. Cambridge: Polity Press. First published
Cambridge: Polity Press [1984].
Heritage, John
2001
Ethno-sciences and their significance for conversation linguistics. In: Brinker
et al. (eds), 908919.
Hindelang, Gtz
1994
Sprechakttheoretische Dialoganalyse. In: Fritz et al. (eds.), 95112.
Holly, Werner
1979
Imagearbeit in Gesprchen. Zur linguistischen Beschreibung des Beziehungsaspekts. Tbingen: Niemeyer
Holly, Werner
2001
Beziehungsmanagement und Imagearbeit. In: Brinker et al. (eds.), 13821393.
Holly, Werner
2007
Audiovisuelle Hermeneutik. Am Beispiel des TV-Spots der Kampagne Du
bist Deutschland. In: Fritz Hermanns and Werner Holly (eds.), Linguistische
Hermeneutik, 389428. Tbingen: Niemeyer.
Holly, Werner, Ulrich Pschel and Jrg Bergmann (eds.)
2001
Der sprechende Zuschauer. Wie wir uns Fernsehen kommunikativ aneignen.
Wiesbaden: Westdeutscher Verlag.
Hundsnurscher, Franz
1995
Some Remarks on the Development of Dialogue Analysis. In: Franz Hundsnurscher and Edda Weigand (eds.), Future Perspectives of Dialogue Analysis,
7993. Tbingen, Niemeyer.
2001
Das Konzept der Dialoggrammatik. In: Brinker et al. (eds.), 945952.
Hutchby, Ian and Robin Wooffitt
1998
Conversation Analysis: Principles, Practices and Applications. Cambridge:
Polity Press.
Jefferson, Gail
1974
Error corrections as an international resource. Language in Society 2:
181199.
Jefferson, Gail
1984
On the organization of laughter in talk about troubles. In: John Maxwell Atkinson and John Hertitage (eds.), Structures of social action: studies in conversation analysis, 346369. Cambridge: Cambridge University Press.
Kallmeyer, Werner and Fritz Schtze
1976
Konversationsanalyse. Studium Linguistik 1: 128.
Kissmann, Ulrike Tikvah (ed.)
2009
Video Interaction analysis. Methods and Methodology. Frankfurt am Main:
Lang
Knoblauch, Hubert and Christian Heath
1999
Technologie, Interaktion und Organisation: Die Workplace Studies. Schweizerische Zeitschrift fr Soziologie 25 (2): 163181.
Kohrt, Manfred
1986
Dialoggrammatik und/oder Konversationsanalyse. In: Franz Hundsnurscher
and Edda Weigand (eds.), Dialoganalyse. Referate der 1. Arbeitstagung
Mnster 1986, 6982. Tbingen: Niemeyer.
284
Khn, Peter
1995
Mehrfachadressierung. Untersuchungen zur adressatenspezifischen Polyvalenz sprachlichen Handelns. Tbingen: Niemeyer.
Labov, William and David Fanshel
1977
Therapeutic Discourse: Psychotherapy as Conversation. New York: Academic Press.
Lemert, Charles and Ann Branaman (eds.)
1997
The Goffman reader. Ed. with introductory essays by Charles Lemert and Ann
Branaman. Oxford: Blackwell.
1983
1988
Putting Linguistics on a Proper Footing. In: Paul Drew and Anthony Wootton
(eds.), Erving Goffman. Exploring the interaction order, 161227. Cambridge:
Polity Press.
Luhmann, Niklas
1984
Soziale Systeme. Grundriss einer allgemeinen Theorie. Frankfurt am Main:
Suhrkamp.
Luhmann, Niklas
1997
Die Gesellschaft der Gesellschaft, Vols 1 and 2. Frankfurt am Main: Suhrkamp.
Meibauer, Jrg
1999
Pragmatik. Eine Einfhrung. Tbingen: Stauffenburg.
Mey, Jacob L.
1993
Moeschler, Jacques
1994
Das Genfer Modell der Gesprchsanalyse. In: Fritz et al. (eds.), 6994.
Moeschler, Jacques
2001
The Geneva School. In: Brinker et al. (eds.), 952957.
Ochs, Elinor, Emanuel A. Schegloff and Sandra Annear Thompson (eds.)
1997
Interaction and grammar. Cambridge: Cambridge University Press.
Pomerantz, Anita
1984
Agreeing and disagreeing with assessments: some features of preferred/dispreffered turn shapes. In: John Maxwell Atkinson and John Heritage (eds.),
Structures of Social Action: Studies in Conversation Analysis, 57101. Cambridge: Cambridge University Press.
Psathas, George (ed.)
1979
Everyday language: studies in ethnomethodology. New York: Irvington.
Rauch, Elisabeth
1992
Sprachrituale in institutionellen und institutionalisierten Text- und Gesprchssorten. Frankfurt am Main: Lang.
Redder, Angelika
1990
Grammatiktheorie und sprachliches Handeln: denn und da. Tbingen:
Niemeyer.
Rehbein, Jochen
2001
Das Konzept der Diskursanalyse. In: Brinker et al. (eds.), 927945.
285
Sacks, Harvey
1972
An Initial Investigation of the Usability of Conversational Data for Doing
Sociology. In: David Sudnow (ed.), Studies in social interaction, 3174. New
York: The Free Press.
Sacks, Harvey
1974 [1972] On the Analysability of Stories by Children. In: Roy Turner (ed.), Ethnomethodology. Selected Readings, 216232. Harmondsworth: Penguin Education. First published in John Gumperz and Dell Hymes (eds.), Directions in
sociolinguistics: The ethnography of communication. New York: Holt, Rinehart, & Winston.
Sacks, Harvey
1984
On doing being ordinary. In: John Maxwell Atkinson and John Heritage
(eds.), Structures of Social Action: Studies in Conversation Analysis, 413429.
Sacks, Harvey
1989
Lectures 19641965. Ed. by Gail Jefferson. Dordrecht: Kluwer Academic Publishers.
Sacks, Harvey
1992
Lectures on Conversation. Vols I and II. Ed. by Gail Jefferson. With an introduction by Emanuel A. Schegloff. Oxford: Basil Blackwell.
Sacks, Harvey, Emanuel A. Schegloff and Gail Jefferson
1978
A simplest systematics for the conversation of turn taking for conversation. In:
Jim N. Schenkein (ed.), Studies in the organization of conversational interaction, 755. New York: Academic Press. First published Language 50: 696735
[1974].
Schank, Gerd and Johannes Schwitalla (eds.)
1987
Konflikte in Gesprchen. Tbingen: Narr.
Schegloff, Emanuel A.
1968 [1972] Sequencing in conversational openings. American Anthropologist 70:
10751095. Reprinted in John Gumperz and Dell Hymes (eds.), Directions
in sociolinguistics: the ethnography of communication, 346380. New York:
Rinehart & Winston.
1972
Notes on a Conversational Practice. Formulating Place. In: David Sudnow
(ed.), Studies in social interaction, 75119. New York: The Free Press.
1979
The Relevance of Repair to Syntax-for-conversation. In: Talmy Givn (ed.),
Syntax and Semantics 12: Discourse and Syntax, 261286. New York: Acadamic Press.
1987
Some sources of misunderstanding in talk-in-interaction. Linguistics 25:
201218.
1992
Introduction. In: Harvey Sacks: Lectures on Conversation. Vols I and II. Ed. by
Gail Jefferson, ixxii. Oxford: Basil Blackwell.
Schegloff, Emanuel A., Gail Jefferson and Harvey Sacks
1977
The Preference for Self-Correction in the Organization of Repair in Conversation. Language 53 (2): 361382.
286
Schegloff, Emanuel A. and Harvey Sacks

1973
Opening up closings. Semiotica 8: 289327.
Schenkein, Jim N. (ed.)
1978
Studies in the organization of conversational interaction. New York: Academic Press.
Schmitt, Reinhold (ed.)
2007
Koordination. Analysen zur multimodalen Interaktion. Tbingen: Narr.
Schneider, Wolfgang Ludwig
1997
Die Analyse von Struktursicherungsoperationen als Kooperationsfeld von
Konversationsanalyse, objektiver Hermeneutik und Systemtheorie. In: Tilmann Sutter (ed.), Beobachtung verstehen, Verstehen beobachten. Perspektiven einer konstruktivistischen Hermeneutik, 164227. Opladen: Westdeutscher Verlag,
Schwitalla, Johannes
2001a
Gesprochene-Sprache-Forschung und ihre Entwicklung zu einer Gesprchsanalyse. In: Brinker et al. (eds.), 896903.
2001b
Konflikte und Verfahren ihrer Bearbeitung. In: Brinker et al. (eds.), 1374
1382.
2001c
Beteiligungsrollen im Gesprch. In: Brinker et al. (eds.), 13551361.
Selting, Margret and Elizabeth Couper-Kuhlen
2000
Argumente fr die Entwicklung einer interaktionalen Linguistik. Gesprchsforschung Online-Zeitschrift zur verbalen Interaktion 1: 7695
Sinclair, John and Malcolm Coulthard
1975
Towards an analysis of discourse. London: Oxford University Press.
Streeck, Jrgen
1983
Konversationsanalyse. Ein Reparaturversuch. Zeitschrift fr Sprachwissenschaft 2: 72104.
Streeck, Jrgen
2009
Gesturecraft. The manufacture of meaning. Amsterdam: Benjamis.
Sudnow, David (ed.)
1972
Studies in social interaction. New York: The Free Press.
Tannen, Deborah
1984
Conversational Style: Analyzing Talk Among Friends. Norwood, NJ: Ablex.
Tracy, Karen
1990
The many faces of facework. In: Howard Giles and W. Peter Robinson (eds.),
Handbook of language and social psychology, 209226. Chichester: John
Wiley and Sons.
Turner, Roy
1974a
Introduction. In: Roy Turner (ed.), Ethnomethodology. Selected Readings,
712. Harmondsworth: Penguin Education.
Turner, Roy (ed.)
1974b
Ethnomethodology. Selected Readings. Harmondsworth: Penguin Education.
Uhmann, Susanne
1997
Grammatische Regeln und konversationelle Strategien. Fallstudien aus Syntax
und Phonologie. Tbingen: Niemeyer.
287
Van Dijk, Teun

1972
Some Aspects of Text Grammars. The Hague: Mouton.
Ventola, Eija
2001
Discourse studies in the English-speaking countries. In: Brinker et al. (eds.),
979994.
Werlen, Iwar
2001
Rituelle Muster in Gesprchen. In: Brinker et al. (eds.), 12631278.
Widmer, Jean
1991
Goffman und die Ethnomethodologie. In: Robert Hettlage and Karl Lenz
(eds.), Erving Goffman ein soziologischer Klassiker der zweiten Generation,
Wodak, Ruth, Johanna Pelikan, Peter Nowak, Helmut Gruber, Rudolf de Cillia and Richard
Mitten
1990
Wir sind alle unschuldige Tter!. Diskurshistorische Studien zum Nachkriegsantisemitismus. Frankfurt am Main: Suhrkamp.
10.

Maeve Cooke
1.
Introduction
Sociology, as the scientific or systematic study of society, emerged as an independent academic discipline only towards the end of the nineteenth century. From the
beginning it has been marked by a dispute as to the method of inquiry appropriate
to it. A central question is whether its methodology should model that of the natural sciences or that of humanistic and cultural studies. The first position is often attributed to Auguste Comte and nineteenth and twentieth century positivism; it
holds that social phenomena can be subjected to observable natural laws. The second position is associated with Wilhelm Dilthey and nineteenth and twentieth century historicism and hermeneutics; it asserts an essential difference between the
methodology of the natural sciences and humanistic and cultural studies, emphasizing the interpretative component of social scientific inquiry.
The first position adopts an external perspective on the production of social
order, viewing it as a quasi-natural process that can be observed in its empirical
regularities and explained by way of nomological hypotheses. The second position
adopts an internal perspective on the production of social order, viewing it as a
meaningfully structured reality for social agents that should be interpreted in terms
of intentional actions.
To be sure, the lines of division are by no means clear cut. Max Weber, for
example, defines sociology as a science that attempts the interpretative understanding of social action in order thereby to arrive at a causal explanation of its course
and effects (Weber 1969). In line with this, he holds that rational interpretation, in
the sense of objective or impartial appraisal, is possible, when the agent and her interpreter equally accept the standards of judgement as valid (Habermas 1988).
Moreover, Weber is ambivalent on the question of whether the methods of the
social sciences should model those of the natural sciences. In this regard we find
striking differences in attitude and emphasis in his writings (McCarthy 1978:
141145).
We cannot explore this debate further here. For our present purposes, it is sufficient to note that, today, there is widespread, if not universal, agreement that the
methods of the natural sciences cannot simply be mapped onto sociology. Furthermore, among those who accept this, a significant number agree that the difference is due in large part to the intentional component of human action.1 By this is
meant, first, that the concept of society is internally linked with the concept of
human action and, second, that in the case of such action it is always in principle
290
Maeve Cooke
possible to ask what the agent(s) intended meant when carrying it out. The implication here is that a description of a particular human action that failed to consider the intentions of the agent would lack an essential ingredient.2 This implies in
turn that intention, understood as subjective meaning, is a sociological primitive:
an irreducible element of sociological analysis. It should also be noted that even if
we accept the reference to subjective meaning as constituting an essential methodological difference between the natural sciences and sociology, a number of important questions remain open, for example, the questions of whether causal explanations of social phenomena are possible and the degree to which objectivity is
achievable in sociology. Thus, accepting the irreducibility of subjective meaning
does not commit the sociologist to a purely internal perspective on social phenomena and require her to abstain from empirical-analytic analyses and causal explanations. However, it presents her with the challenge of developing a theoretical
framework, and corresponding methodology, that negotiates the tensions between
the external and internal perspectives.
In the early 1970s, as part of an endeavour to meet this challenge, Jrgen Habermas, a philosopher and sociologist in the Frankfurt School tradition of critical
social theory, turned to pragmatics.3 The importance of pragmatics in contemporary
sociology is largely due to Habermas contribution to the study of society. This contribution is so significant that no sociologist today can ignore his social theory and,
by extension, the programme of pragmatics underpinning it. While Habermas owes
a significant debt to his colleague Karl-Otto Apel, who introduced him to pragmatics in the late 1960s, Apels work has been more influential in philosophy than
social theory (Apel 1980). Furthermore, unlike Habermas, Apel did not develop his
reflections on pragmatics into a systematic programme. For these reasons, we will
take Habermas as our reference point in the remainder of our discussion.
In turning to pragmatics, Habermas concerns were not only methodological.
He sought, in addition, to provide an answer to central questions of sociology, such
as how we should conceptualize social action, how social order is possible and how
social order is reproduced. A further, equally important, aim was to provide a basis
for the rational critique of society without appealing to metaphysical notions of
truth, justice or the good. The Theory of Communicative Action, first published in
two volumes in 1981, attempts to meet these aims. In the following, we will examine the role of pragmatics, first, from the point of view of method, where we will
consider the conceptual strategy he deems most appropriate for sociological inquiry; then, from the point of view of the interconnected questions of social action,
social order and social reproduction; finally, from the point of view of social
critique, where we will look at the role pragmatics plays in critique of social pathologies and in developing a postmetaphysical conception of reason, which is the
normative reference point for this critique. In the concluding section we will look
briefly at some challenges to Habermas appeal to pragmatics.
Pragmatics in Habermas' critical social theory
2.
291
Method
In the course of the 1970s Habermas develops a theory of pragmatics that, as indicated, is supposed to help him answer central sociological questions relating to social action, social order and social reproduction as well as provide a normative
basis for his critical theory of society. In a preliminary essay based on lectures delivered at Princeton University in 1971, he focuses specifically on methodology,
identifying a number of metatheoretical issues as centrally important in the endeavour to develop a conceptual strategy appropriate for sociology.4 In each case,
in line with his concern to avoid a simple decision for either explanation or interpretation, Habermas adopts a position that seeks to retain the most valuable elements in the apparently conflicting approaches. For example, in his reflections on
the debate between methodological individualism and holism, he does not straightforwardly endorse either of these approaches but attributes to them complementary
strengths and weaknesses that make each useful for different purposes and in different contexts (Habermas 1971: 1314). Most relevant in the present instance is
the methodological decision he makes in favour of subjective meaning as a sociological primitive (Habermas 1971: 310). This decision leads him to view society
as a continuous process of generating a meaningfully structured reality, whereby
meanings are produced according to an underlying system of abstract rules and
have a sociologically relevant, intentional component. Having generally endorsed
theories that take subjective meaning as an irreducible element of sociological
analysis, Habermas now makes a distinction within this set between constitutive
and communicative theories.5 It is at this point that pragmatics enters the scene.
In his account, constitutive theories base the process of the generation of meaning on a transcendental subject. Some, taking their lead from Husserls account of
the constitution of the everyday world of lived experience (the lifeworld), conceive of this transcendental subject as an intelligible ego modelled on the empirical
individual subject. Others, taking their lead from Hegel and Marx, conceive of it as
a species-subject that constitutes itself in history. In each case, they encounter the
problem of how to move from the level of experience to the level of society. More
precisely, they run up against the problem of how adequately to conceptualize intersubjective meanings within a subject-object model of cognition and action. In
models of this kind, knowledge and action are conceived monologically, as the result of a relationship between a subject and an object that is distinct from it. However, if meaning is conceived of essentially as the product of a relationship
between a subject and some distinct object, commonly shared, intersubjective
meanings appear to have a secondary status. From a social-theoretical point of
view this gives rise to considerable difficulties. One difficulty is that it makes it
hard to explain how society, as a process of generating intersubjective meanings,
can be granted the primordial status it has in phenomenological accounts. This
weakness of constitutive theories, together with other, related ones, prompts Ha-
292
Maeve Cooke
bermas to look to communicative theories, which construe meaning as intersubjectively generated from the beginning.
In this preliminary essay, Habermas finds himself in the unwelcome situation
of being unable to refer to an existing body of work that would support the kind of
communicative theory he is seeking; he admits that the theoretical approach that he
deems methodologically most appropriate for social scientific inquiry has not yet
been adequately developed. However, he is clear as to the direction it is necessary
to take. What is required in his view is a sociological theory that would start from
the premise that everyday communication, as a process of generating shared meanings, is the basis for social action (and socialization); the use of ordinary language
would be seen as constituting the intersubjective relations within which action
takes place (as well as the personality structures of the subjects of action). His
name for this project is universal pragmatics.
Habermas programmatic essay What is Universal Pragmatics? first appeared in German in 1976. The adjective universal indicates the difference between his reconstructive approach to the study of linguistic behaviour and empirically oriented analyses. In his account, empirical-pragmatic approaches set
themselves the task of describing the speech acts typical of a certain milieu, which
may then be analyzed from sociological, ethnological and psychological points of
view. Habermas favours a broader perspective. He aligns himself with general
pragmatic theory, which is concerned not with specific instances of language use,
but rather with formally reconstructing the rule system that underlies the ability of
a subject to utter sentences in any relevant situation whatsoever (Habermas 1998:
54). Accordingly, he characterizes the task of universal pragmatics as being to
identify and reconstruct universal conditions of possible Verstndigung, which we
can translate here as mutual understanding6 (Habermas 1998: 21). Soon after the
publication of this essay, however, he distanced himself from the name universal
pragmatics. In a footnote to the English translation, published three years later in
1979, he writes that he is no longer happy with his original terminology, suggesting
that his project instead be named formal pragmatics. In opting for the word formal he wishes to emphasize the continuity between his approach and formal semantics, notwithstanding his focus on the use of language in speech acts or utterances as opposed to the semanticists concern with the properties of isolated
sentences.7 His choice of the word formal also makes sense in the context of his
adoption of the procedure of rational reconstruction which, as indicated, is the
method he attributes to his universally oriented, pragmatic approach. By rational
reconstruction he means the systematic reconstruction of the intuitive knowledge
of competent subjects (Habermas 1998: 2941). Reconstructions relate to pretheoretical knowledge of a general sort, to universal capabilities and not merely to the
competences of particular groups or individuals. Thus, what begins as an explication of meaning ultimately aims at the reconstruction of species competences.
Habermas sees it as the great merit of Noam Chomsky to have developed this idea
293
in the case of grammatical theory (Chomsky 1969). He also commends Jean Piaget
and Lawrence Kohlberg for, respectively, their theories of cognitive and moral development, which, like Chomskys, rationally reconstruct the pretheoretical and
implicit knowledge and competences of speaking and acting subjects in general.
In distinguishing his formal-pragmatic approach from an empirical-pragmatic
one, Habermas does not want to deny the importance of empirical validation. As
indicated, reconstructive theories belong to the category of general theory in the
sense that they seek to reconstruct the rule system that underlies the ability of a
subject to generate well-formed sentences in any language whatsoever. At the
same time, they belong to the category of empirical theories. Their explications of
general structures and universal conditions have a merely hypothetical status and
must be subjected to the usual methods of testing. Thus, Chomskys and Habermas reconstructions of linguistic competence are open to checking against speakers intuitions, scattered across as broad a sociological spectrum as possible (Habermas 1984:138). As is appropriate for scientific investigations of this sort, the
checking procedure is circular: it moves in a circle between theory formation and a
more precise rendering of the object domain (Habermas 1998: 40).
From this it is clear that Habermas places reconstruction, as an explicative exercise, on the side of interpretation rather than explanation, thereby connecting it
with an internal rather than external perspective on social phenomena. The point of
reconstructions is neither to describe behaviour nor to systematize their findings in
law-like explanations, but rather to bring to consciousness implicit knowledge
through a maieutic method of interrogation, which involves the choice of suitable
examples and counterexamples, the use of contrast and similarity relations, translation, paraphrase and so on (Habermas 1998: 40). This is not to deny that the procedures employed in constructing and testing hypotheses, in appraising competing
reconstructive proposals and in gathering and selecting data, are not in many ways
like the procedures used in the nomological sciences (Habermas 1998: 46). Like
such sciences, reconstructive theories usually operate at a high level of abstraction,
seeking to explicate basic conceptual frameworks and to formalize assumptions
built into everyday behaviour. The key difference, to recall our earlier discussion,
is that reconstructive social theories take subjective meaning as an irreducible element of social scientific analysis whereas theories that adopt a purely external
perspective do not.
We have seen that Habermas turns to pragmatics in order to find a method of
sociological analysis that takes subjective meaning as a sociological primitive,
while avoiding the weaknesses of constitutive models which attribute the process
of generating a meaningfully structured reality to some sort of transcendental subject. His aim is to develop a communicative model that replaces the subject-object
paradigm underlying constitutive theories with an intersubjective paradigm in
which the generation of shared meanings is seen as fundamental in the constitution
of social reality. He sees the use theory of meaning developed by Ludwig Wittgen-
294
Maeve Cooke
stein in his later work as making an important contribution towards developing

such a model (Wittgenstein 2001). But for reasons that will become clearer in the
next section, he finds speech act theory as developed by Austin and by Searle (in
his early work) more fruitful for his purposes.
Speech act theory shares with reconstructive language analysis a concern to explicate the rules that a competent speaker must master in order to form grammatical sentences and to utter them in an acceptable way. However, by contrast with
Chomskian linguistics, instead of starting from the postulate of an implicit ability
to produce sentences, speech act theory starts from the assumption of an implicit
ability to employ sentences in speech acts (utterances). Thus, it is concerned with a
competent speakers ability to communicate as opposed to produce a grammatical
sentence. Accordingly, a general theory of speech acts explicates the system of
rules that competent speakers master when they fulfil the conditions for the successful employment of sentences in utterances, no matter to which particular languages the sentences may belong and in which random contexts the sentences may
be uttered (Habermas 1998: 47). The importance of speech act theory in the development of Habermas conceptual strategy is twofold. First, by virtue of its focus on
utterances as communicative actions, it enables him to show how the meaning of
speech acts is produced by way of an action involving speaker and hearer(s); this
provides support for his shift in paradigm from a subject-object based, monological model of meaning production to an intersubjective one. Second, it enables him
to show how the production of meaning is internally connected with a number of
different kinds of claims to validity. The connection between meaning and validity
in a number of dimensions emerges from speech act theorys analysis of the successful employment of sentences in utterances. Habermas contends that for an utterance or speech act to be successful it must count as true for the speaker and hearers, in the sense of representing something in the world, must count as normatively
right, in the sense of conforming to socially recognized expectations and must
count as truthful, in the sense of expressing something intended by the speaker. As
we shall now see, the connection between meaning and validity claims is of crucial
importance in Habermas accounts of social action, social order and social reproduction and in his proposal for a postmetaphysical conception of reason that permits social critique.
3.
Social action, social order and social reproduction
Formal pragmatics forms the basis for the kind of communicative theory that Habermas thinks necessary in order to avoid the weakness of constitutive social theories and those that share their subjectivist orientation. In his view, such theories
not only have difficulties adequately conceptualizing meanings shared in common;
they are also not well suited to the task of adequately conceptualizing social action,
295
social order and social reproduction. This is because they tend to privilege instrumental action, understood as action undertaken as a means to achieving an end that
is specified in advance of the undertaking. Accounts of social order and reproduction that take instrumental action as the fundamental mode of social action run up
against the problem of how to explain long-term social cohesion, when the preservation and reproduction of the social order in question does not evidently equally
enable everyone to obtain their particular ends (cf. Habermas 1998: 234). A central
aim of Habermas theory of communicative action is to show that instrumental action is not the primary mechanism in social cohesion and social reproduction. He
draws on formal-pragmatic analyses of everyday language in order to achieve this
aim.
As the background to our discussion in this section and the next, it is important
to note that in The Theory of Communicative Action Habermas offers an account of
the developmental path of modern societies in terms of two categorially distinct
modes of societal integration. In what he calls the lifeworld, co-ordination of action takes place primarily by way of communicative action, which relies on intentionally acting human subjects who establish validity-based relationships with one
another. In the system, by contrast, co-ordination of action takes place primarily
by way of the functional interconnection of action consequences, bypassing the intentions of the agents concerned. For Habermas, rationalization processes, by
which he means the development of the internal logic of a particular mode of societal integration, are possible and desirable in each case. The rationalization of the
system refers to increasing complexity and increasing capacity to take on steering
functions such as material reproduction and administration. By contrast, the
rationalization of the lifeworld is a matter of communicative rationalization: the
development of the internal logic of communicative action. It takes place in the domains of cultural tradition, social integration and socialization and entails the increasing independence of processes of justification from traditional normative contexts of validity, together with increasing reliance on action oriented to achieving
understanding (Verstndigung).
Communicative action is a form of social interaction in which the plans of various agents are co-ordinated through an exchange of speech acts. Drawing on the
speech act theory of Austin and the early Searle, Habermas maintains that the success of a speech act is not just a matter of comprehending the sentence embedded in
the utterance, but of responding in the appropriate manner to what the speaker does
in performing the action. In addition to commending these speech act theorists for
emphasizing the action component of linguistic understanding, he praises them for
directing attention to modes of language use other than the assertoric and descriptive modes. In this respect, too, they take their lead from the later Wittgenstein.
However, Habermas finds speech act theory more promising than Wittgensteins
use theory of language, due to its assertion of a connection between meaning and
validity in a context-transcending sense. For Habermas, truth and truth-analogous
296
Maeve Cooke
concepts such as moral rightness have a context-transcending aspect in the sense

that they assert a validity that transcends particular places and times. This is one
reason why he refers to them as universal validity claims.8 The connection between
meaning and universal validity disappears from view in Wittgensteins theory, since
validity is held to be relative to particular, historically specific, language games. To
be sure, Habermas finds Austins and Searles approaches to universal validity
claims unsatisfactory. He criticizes Austin for blurring the distinction between clear
cut universal validity claims such as truth and normative rightness and a host of
other evaluative criteria (Habermas 1998: 290291). He admonishes Searle for acknowledging only the representational function, together with the claim to truth
corresponding to it, as the only clear cut universal validity claim (Habermas 1998:
291292). Against Austin and Searle Habermas seeks to demonstrate that everyday
language use is connected with three clear cut universal validity claims.
This is one of the two main respects in which Habermas moves beyond speech
act theory. The other is his emphasis on the relationship of obligation into which
the speaker and hearer(s) enter with the performance of a speech act. Taking his initial lead from Austin, he finds it instructive to consider what a speaker does in saying something. He begins by discussing Austins distinction between locutions and
illocutions (Austin 1962). Initially Austin reserved the term locution for the
propositional content of what is said and the term illocution for the performative
act of uttering sentences with a propositional content. He later reformulated the
distinction as one between constatives and performatives, defining constatives as
speech acts in which we concentrate on the truth or falsity of what is said and performatives as speech acts in which we concentrate on the illocutionary force of the
utterance (Habermas 1998: 6681). Habermas finds this way of distinguishing between constative and performative speech acts unhelpful, proposing instead that
we distinguish between utterances on the basis of the kinds of validity claims they
raise. In his reformulated version of Austins classificatory schema, constatives involve the cognitive use of language and raise claims to truth (examples are assertions and descriptions); here we thematize the content of the utterance as a statement about what is or could be the case. Regulatives involve the interactive use of
language and raise claims to normative rightness (examples are promises, requests
and warnings); here we thematize the relationship of obligation into which speaker
and hearer enter with the speech act. Habermas adds a third category, expressives,
to these revisions of Austins schema; as the name indicates, these involve the expressive use of language and raise claims to truthfulness (examples are confessions
and disclosures); here we thematize the truthfulness or sincerity of the speakers
intentions. He proposes, further, that we understand the illocutionary force of the
utterance what the speaker does in performing any speech act as a rationally
binding force (bindende Kraft), in the twin sense of binding and compelling (Habermas 1998: 8188). This rationally binding force is held to be operative in all
three modes of language use: the constative, regulative and the expressive.
297
This, then, is Habermas answer to the question of what we do in performing a

speech act: we enter into a relationship of rationally-based obligation with the
hearers. The obligation is rationally-based since it arises from the speakers undertaking to support what she says with reasons. He conceptualizes this in terms of a
validity claim raised by the speaker. The speaker claims that what she says is true
or that she is normatively right to say it or that she is sincere in saying it, thus
raising a claim to validity that calls for a yes or no response from the hearer.
The success of the interaction depends on the speakers ability to supply reasons to
support the validity of the claim, if challenged, and on the hearers ability to provide reasons in support of her yes or no. For Habermas, therefore, the illocutionary force of an utterance is not simply an aspect of its meaning conveyed by a
performative verb; it is a co-ordinating power that is based on reciprocal obligations to provide reasons. This is what he means by the rationally binding force of
linguistic interaction.
Habermas holds, furthermore, that every speech act raises all three claims simultaneously. More precisely, he holds that with every speech act, the speaker raises
one of these three kinds of validity claims directly, while the other two are raised
indirectly. Thus, for example, if a speaker states she has no money to pay the rent,
she is raising a claim to truth that calls for a yes or no response from the hearer(s) or for an abstention on grounds of insufficient knowledge of the facts. At the
same time she raises a claim that her utterance is appropriate in the context in question (normatively right) and that she is sincere in uttering it. In this case, a claim
to truth is raised directly and claims to normative rightness and sincerity are raised
indirectly. If a speaker requests a hearer to lend her some money, she raises a claim
to normative rightness directly and claims to truth and truthfulness indirectly (Habermas has clarified that indirect claims to truth should be understood as claims
that certain existential presuppositions are met, 1998: 146). A speaker who confesses that she has spent all her money raises a claim to truthfulness directly and
claims to normative rightness and truth indirectly.
The thesis that every speech act raises three claims to validity simultaneously is
important for Habermas attempt to show that an orientation to a plural multi-dimensional conception of validity is built into everyday language use. As we shall
see, this plural conception of validity is the basis for his non-logocentric conception of communicative rationality, which does not privilege propositional truth but
conceives of reason as an interplay of truth, moral rightness and truthfulness/authenticity. The thesis of three simultaneous validity claims is also important from the
point of view of social integration and reproduction. Its purpose is to show that the
three validity claims raised with every speech act amount to three structural components of speech acts that correspond to the three functions of the communicative
use of language: (a) representing states of affairs (or something in the world that
confronts the speaker), (b) entering into a relationship with a hearer and (c) giving
expression to the speakers intentions or experiences. Referring to the communi-
298
Maeve Cooke
cation-theoretic approach expounded by the German psychologist Karl Bhler

(Bhler 1934), Habermas posits these as three mutually irreducible but internally
connected linguistic functions.9 His thesis is that these linguistic functions, carried
out with every speech act, serve to maintain and reproduce the lifeworld in its three
domains of cultural tradition, social integration and socialization.
For this, however, the thesis of three simultaneous validity claims is not sufficient. Even if successful (see Cooke 1994: 9091), it would show merely that the
communicative use of language fulfils functions of representation, of establishing
a relationship with the hearer and of expression; it would not show that the communicative use of language is of fundamental importance in maintaining and reproducing the lifeworld. Showing the latter is central to Habermas critical project.
We will recall his claim that the colonization of the communicatively regulated domains of the lifeworld by the functional rationality of the system is a deeply troubling pathology of modernity. If he were unable to show that communicative action
is indispensable for the purposes of establishing, maintaining and reproducing
social order, he would lack grounds for referring to the shrinking of the communicatively regulated domains of the lifeworld as a pathology and for adopting the
metaphor of colonization to describe the expansion of functional rationality into
the communicatively regulated domains. In order to meet this challenge, Habermas
seeks to demonstrate the primacy of communicative action over instrumental (and,
in particular, strategic) action. Once again, he employs a formal-pragmatic strategy. This entails showing the primacy of the communicative use of language over
other modes, specifically the strategic one.
Just as the communicative use of language amounts to communicative action,
the strategic use of language amounts to strategic action. Strategic actions are instrumental insofar as one agent uses another as a means of achieving his respective
goals; he treats other persons as though they were objects or entities in the physical
world. In contrast to communicative action, in which the success of the interaction
is not at the disposal of the individual agent but depends on all parties involved,
strategically acting agents are not dependent on the co-operation of others (Habermas 1998: 220227). While Habermas acknowledges that strategic action, like
communicative action, is a mode of action within the lifeworld, he insists that it is a
merely secondary mode of social integration. His position, in other words, is that
communicatively regulated interactions must predominate in the lifeworld if this is
to maintain and reproduce itself. He seeks to show the priority of communicative
action by demonstrating that the strategic use of language is parasitic on the communicative one.10
In The Theory of Communicative Action Habermas attempts to show that the
use of language oriented to reached understanding (Verstndigung) is the primary
or original mode of language use by drawing on Austins distinction between locutions and perlocutions (Habermas 1998: 122129) In Austins account, the locutionary act is the act of saying something, the illocutionary act is the action per-
299
formed by saying something and the perlocutionary act is the effect the speaker
produces upon the hearer. Initially, Habermas took this to mean that perlocutionary
acts are an indication of the integration of speech acts into contexts of strategic action (Habermas 1998: 126). This interpretation neglects the importance of distinguishing between different types of perlocutions. It is important, above all, to distinguish between the contingent, often unintended effects on the hearer(s) of
particular speech acts and the intended, but concealed effects of particular speech
acts. An example of the former is the pleasure a speaker gives to his hearers husband by lending him money for the rent (for the act might have given him displeasure); an example of the latter is the speakers intention to use his loan of
money to blackmail the hearer. Habermas has subsequently modified his position,
clarifying that only perlocutionary effects of the latter kind are relevant for his argument (Habermas 1998: 329333). He sees them as examples of latently strategic
action. As illustrated by the case where the speaker intends to blackmail the hearer
by lending her money, the success of latently strategic action depends on the speakers success in hiding his intentions from the hearer; he must successfully pretend
that he is using language with an orientation towards mutual understanding (Verstndigung), as opposed to strategically, using the interaction with the hearer as a
means to furthering his own particular ends.
The latently strategic mode of language use is by definition parasitic on the
communicative use (Cooke 1994: 24).11 Thus, demonstrating its secondary character will not be sufficient for Habermas purposes. The more interesting and difficult case is the manifestly strategic use of language. In this mode of language use,
the presupposition of an orientation towards reaching understanding is overtly suspended (Habermas 1998: 225226). A speaker who backs up his demand that a
hearer lend him some money with the threat of sanction (for example, by threatening to harm her daughter) illustrates such language use. In order to make good his
claim that the communicative use of language is the primary mode of language use,
Habermas has to show that this manifestly strategic use of language, too, is parasitic on the communicative one. This part of Habermas argument is underdeveloped (Cooke 1994: 2326).
In this section we have considered the importance of formal pragmatics in Habermas accounts of social action, social order and social reproduction. We have
seen that its role is threefold: to demonstrate that everyday linguistic communication has an in-built rationally binding force, to show that this binding force is operative in three dimensions corresponding to the three linguistic functions of representation, establishing a relationship with the hearer(s), and expression, and to
demonstrate that the communicative use of language is the primary mechanism in
establishing, maintaining and reproducing the social order. In the next section, we
will look at the importance of formal pragmatics for Habermas project of social
critique.
300
4.
Maeve Cooke
Social critique
Habermas is first and foremost a critical social theorist. His sociological concerns
are intimately bound up with his concern to diagnose and criticize the pathologies
of modernity. We have already mentioned the colonization of the lifeworld as one
of these pathologies. Habermas, drawing on Weber, sometimes refers to this as the
modern loss of freedom (Habermas 1987: 301302). The second principal pathology he diagnoses is a loss of meaning (here, too, he draws on Webers analysis of
modernity). By this he means the cultural impoverishment of the lifeworld that results from the differentiation of the cultural value spheres of science, law/morality,
and art under conditions of modernity; these value spheres become fields for the
specialized treatment of questions of truth, normative rightness and beauty/authenticity by experts. While these specialized discourses promote a growth in cognitive-instrumental, moral-practical and aesthetic-expressive knowledge, thus giving rise to significant learning processes, they are also a potential source of
pathology. This is because with progressive modernization, the gap between the
expert cultures and the general public becomes ever wider. Learning processes
within the expert cultures do not automatically flow back into the communicative
practices of the lifeworld, contributing to their renewal and regeneration, but remain cut off from them. Lacking revitalizing impulses from the specialist cultures
(or, indeed, from other sources12), everyday communicative practices threaten to
become impoverished.
His critique of these pathologies is guided by a corresponding normative vision. At the core of this vision is a conceptual triad that has defined his work from
his earliest major publication The Structural Transformation of the Public Sphere
(Habermas 1989): the concepts of the public sphere, discourse and reason. The interplay of these concepts yields a utopian vision of a communicatively rationalized
lifeworld (Habermas 1991: 6469). In the projected society, cultural traditions
would be reproduced through processes of intersubjective evaluation of validity
claims, legitimate orders would be dependent on critical and open argumentative
practices for justifying laws/norms and for making political decisions, and individual identities would be self-regulated through processes of critical reflection. We
must assume, furthermore, that in such a communicatively rationalized lifeworld
there would be ongoing processes of semantic renewal in the form of continuous
flows of knowledge from the specialized, expert cultures into the communicative
practices of everyday life.13 This utopian projection of a rationalized lifeworld is in
turn part of a vision of society in which the development of systems rationality and
communicative rationality would be evenly balanced: systems rationality would
no longer encroach into the communicatively regulated domains of the lifeworld.
Formal pragmatics contributes to Habermas critical project in two main respects. His formal-pragmatic investigations are intended, first, to provide a normative reference point for the critique of the two principal pathologies of modern
301
societies: the colonization of the communicatively regulated domains of the lifeworld by systems rationality and its cultural impoverishment due to the emergence
of specialized discourses among experts in the areas of science, law/morality and
art. They are intended, second, to provide a normative reference point for criticism
of the conduct and outcomes of actual practices of deliberation in everyday life. In
each case, the normative reference point is given the name communicative
rationality.
As we have seen, Habermas thesis of the colonization of the lifeworld makes
the claim that the uneven developmental path of modern societies is pathological.
What is required, instead, is an evenly balanced pattern of societal rationalization.
Habermas expresses the idea of a non-pathological, evenly balanced process of societal rationalization in terms of the harmonious, free interplay of the three cultural
value spheres of science, morality/law and art, in which no one sphere would predominate at the expense of the other (Habermas 1984: 240). His formal-pragmatic
investigations are used as a normative underpinning for this idea.
The relevant formal-pragmatic finding is that everyday linguistic communication fulfils three mutually irreducible functions: representation, establishing a relationship between speaker and hearer(s), and expression; as we have seen, this is
the conclusion Habermas draws from his thesis that the speaker raises three validity
claims simultaneously with every speech act. He asserts, in addition, that these
three linguistic functions can be correlated with the kind of knowledge defining
each of the three cultural value spheres: scientific knowledge, moral-practical
knowledge and aesthetic-expressive knowledge. He then makes a further correlation between the three cultural value spheres and the three moments of a once substantively unified concept of reason, which have become separated from one another under conditions of modernity;14 the image of harmonious, free interplay is
extended to these three moments of reason, evoking an idea of reason whose unity is
defined purely formally as opposed to substantively, and that is non-logocentric, expanding the traditional philosophical focus on theoretical reasoning and truth to include other dimensions of validity (such as justice and beauty) on an equal footing.
If the colonization thesis suggests as its opposite evenly balanced processes of
functional rationalization and communicative rationalization, the thesis of cultural
impoverishment suggests as its opposite the continuous feedback back into everyday communicative practices of the knowledge produced by experts in their various fields. However, whereas Habermas evidently wishes to use his formal-pragmatic analyses to provide empirically based, normative support for the idea of
evenly balanced processes of societal rationalization, he gives no indication of any
formal-pragmatic finding that would lend support to the idea of the free flow of
knowledge between expert cultures and the lifeworld. Nor is it easy to see which
part of his analyses could help him in this regard.
In addition to these two principal pathologies he diagnoses and attributes to
deficiencies in modern processes of societal rationalization, Habermas identifies a
302
Maeve Cooke
number of further potential pathologies (Habermas 1987: 142144); by contrast

with the pathology of colonization, which threatens it from the outside, these potential disturbances arise within the lifeworld (in this they are similar to the pathology of cultural impoverishment). In addition to the loss of meaning that results
from the encapsulation of specialist cultures, he mentions the anomie that results
from disturbances in the process of social integration and the mental illnesses (psychopathologies) that result from disturbances in the process of individual socialization. At least initially, Habermas held that formal pragmatics could play a role in
the critique of these disturbances. In The Theory of Communicative Action and in
the preliminary studies that led up to it, he suggested that formal-pragmatic analyses could help to establish a normative basis for criticizing the phenomena in
question as pathologies. The specific strategy he envisaged was a formal-pragmatic
account of systematically distorted communication (Habermas 1974). Despite occasional remarks in his writings up to the middle of the 1980s that indicate that he
had not yet abandoned this strategy, he has not since then made any serious attempt
to develop such an account. His key intuition seemed to be that systematically distorted communication violates what formal pragmatics shows to be the internal organization of speech, by undermining the three linguistic functions of representation, establishing a relationship between speaker and hearer(s) and expression.
The violation, which takes the form of disconnecting meaning from validity, speaking from acting and/or meaning from intention, is due to the overwhelming pressure exerted on the internal organization of speech by its external organization; by
the external organization of speech he means the regulation of the normative context in which the communicative interaction takes place: for example, the determination of who is allowed to take part in which discussion, who can initiate topics,
who can bring the discussion to a close, who can contribute and in which order, how
the topics are ordered, and so on. To be sure, in order to develop a convincing account of systematically distorted communication, Habermas would have had to
embed it within his critical account of societal rationalization as colonization (at
the time of publication of his 1974 essay, he had not yet provided such an account);
specifically, he would have had to show how the infiltration of functional rationality into the communicatively regulated domains of social life affects the external
organization of speech, contributing to a violation of its internal organization. Formal-pragmatic arguments would not be helpful here. Nor would they be able to
ground the normative force of the thesis implicit in Habermas account of systematically distorted communication: the thesis that the regulation of the context in
which speech takes place should conform to demanding normative standards of inclusivity, fairness and openness. In short, even a well-developed formal-pragmatic
account of systematically distorted communication would show only what the internal organization of speech and its disruption would look like; on its own it would
not be sufficient to provide the sought-for normative basis for critique of disturbances such as anomie and psychopathologies (Cooke 1994: 148150).
303
As indicated, the normative reference point for Habermas diagnosis and

critique of social pathologies is the concept of communicative rationality. Habermas sets himself the task of elaborating a concept of reason that responds to two
(closely connected) challenges facing contemporary critical social theorists:
maintaining a dialectics of immanence and transcendence and maintaining concepts of validity that transcend times and spaces while acknowledging the historical situatedness of all actual conceptions of context-transcending validity. The
first is specific to theorists in the Frankfurt School tradition who take seriously the
Hegelian thesis that reason is embedded within history itself; the second arises for
any social theorists today who seek to uphold an idea of reason while taking on
board the post-Enlightenment suspicion of transcendental ideas such as truth, justice and the good. Formal pragmatics plays a role in meeting each of these challenges.
The term Left-Hegelian is sometimes used to characterize modes of critical
social theory that, taking their lead from Marx and Lukcs, are concerned to re-articulate Hegelian speculative thinking in a materialist way. Most critical social theorists in the Frankfurt School tradition fall into this category. The aspect of Hegels
thinking most relevant here is his emphasis on reasons simultaneous immanence
and transcendence. Hegel endeavours to show a) that reason is immanent to human
history in the sense that the latter is a process in which reason actualizes itself by
way of the progressive overcoming of contradictions and b) that reason always
transcends its particular historical self-actualizations.15 The attempt to maintain a
dialectical relationship between immanence and transcendence is taken up explicitly by Theodor W. Adorno, Max Horkheimer, Herbert Marcuse and other first
generation critical social theorists (see, for example, Adorno 1981) and continues
to play a central role today in the writings of Axel Honneth, the most prominent
member of the third generation of critical social theorists (Honneth 1994, 2003).
Habermas, the most prominent member of the second generation is less inclined to
describe his project in explicitly Left-Hegelian terms; nonetheless, a concern to
sustain a dialectics of immanence and transcendence is certainly also one of its
most important elements (see, for example, Habermas 1996:125). In Habermas
work, this amounts to the elaboration of a conception of reason that is at once context-transcending, in the sense of transcending times and places, thus extending to
human beings in general, and immanent, in the sense of being anchored in practices
within everyday historical practices, specifically in everyday practices of using
language communicatively.
Habermas theory of communicative action is of crucial importance to his attempt to elaborate a conception of reason that is at once immanent and transcendent. We have seen that he uses a formal-pragmatic analysis of everyday language
use as the basis for an account of social action, order and reproduction that has a
normative as well as an empirical content. It is empirical in the sense that it uses
empirical evidence to reconstruct the relevant structural features of everyday lan-
304
Maeve Cooke
guage use; it is normative in the sense that the resulting account of language is used
as the basis for a normative idea of a communicatively rationalized lifeworld that
co-exists harmoniously with a well-functioning system. An analogous interplay of
the empirical and normative characterizes the concept of communicative rationality. This concept expresses the rational potential implicit in immanent to everyday communicative action. As a practice of raising and responding to validity
claims, communicative action is conceptually tied to more or less rudimentary
practices of argumentation. Habermas maintains that the validity claims raised in
communicative action have a context-transcending, critical force that derives from
idealizations built into the concept of argumentation (see Cooke 2006: 4753).
One set of idealizations refers to the procedure of argumentation: to the normative
promise contained in the general understanding of how argumentation should be
conducted. Examples here are the idealizing presuppositions that all participants
are motivated by a concern to find the single right answer, that no force is exerted
except that of the better argument, that all relevant arguments are considered, that
no one should be excluded from the deliberation on grounds such as race, class and
gender and that everyones voice should be given an equal hearing. Another set
refers to the outcome of argumentation: to the normative promise contained in the
general understanding of the aim of argumentation. An example here is the idealizing presupposition that a discursively reached agreement warrants the truth of
propositions and constitutes the justice of norms. In both cases, the tension between the normative promise contained in these idealizations and what actually
happens in everyday communicative practices provides a basis for criticism: in the
one case, they permit criticism of the ways in which the outcomes of argumentation are reached; in the other case, they permit criticism of the outcomes from the
point of view of an idea of context-transcending validity (e.g. truth or justice). The
critical power of communicative action resides primarily in the tension between an
idealized notion of argumentation and what happens in actual empirical practices.
It should be noted that this critical power is not restricted to a particular socio-cultural context. Since it is grounded in universal features of language use, it expresses a critical perspective with context-transcending force, in the sense that its
validity would have to be accepted by everyone, everywhere, irrespective of sociocultural context. It should be noted, too, that the critical power of the concept of
communicative rationality refers to an idealized practice of communication that
has the status of a methodological fiction. Habermas warns against essentialist
misunderstandings of the ideal speech situation, emphasising that it is merely an
idealizing projection and best understood as enabling a fruitful thought experiment
(Habermas 1996: 322323). In this way, his formal-pragmatic analyses of language enable Habermas to assert an idea of reason that is at once immanent to history, in the sense that it is built into everyday practices of linguistic communication, and transcendent, in the sense that it provides a reference point for critique
of all actually existing communicative practices.
305
As part of his engagement with the post-Enlightenment critique of reason, from

the Romantics, through Nietzsche and historicism, up to contemporary post-structuralism, Habermas emphasizes the postmetaphysical character of the concept of
communicative rationality. The role played by formal pragmatics here is very similar to its contribution to sustaining a dialectics of immanence and transcendence.
By postmetaphysical he means, principally, that the concept of communicative
rationality breaks with traditional views of the substantive unity of reason; refrains
from specifying the content of truth, justice and the good life; moves from a subject-object model of cognition to an intersubjective one; gives up the traditional
fixation on truth and theoretical reason; and renounces all dependency on otherworldly projections such as God or the Good. Our discussion in the foregoing has
cast light on how his formal-pragmatic investigations help him to achieve these
aims. The reference to reason built into everyday language use is a formal as opposed to substantive conception of reason, for it amounts only to a procedure for
determining what is true or right and leaves open the question of the respective
content of these concepts; it is non-subjectivist in the sense that it starts from interaction in communicative practices rather than an individual subjects relation to
an object independent of it; it is non-logocentric in the sense that it does not assert
the primacy of theoretical or propositional truth but comprises a plurality of dimensions of validity that interact with each other and relate to each other on an
equal footing; it is historically rooted in the sense that it is contained within actual
practices of using language; and it is innerworldly in the sense that the ideal reference point to which it appeals is not some otherworldly idea of God or the
Good but an idealized projection of a speech situation in which the idealizing presuppositions relating to the conduct of intersubjective deliberation would be realized in practice and in which the outcomes of such deliberation would contribute
constructively to validity in a context-transcending sense.
In this section we have considered the ways in which formal pragmatics enables Habermas to develop a concept of communicative rationality that is the indispensable normative reference point for his critical social theory, both with regard to the critique of social pathologies and with regard to criticism of the
shortcomings and distortions of actual practices of intersubjective deliberation. In
my final remarks I will outline the principal lines of objection that have been raised
against Habermas use of formal pragmatics as the normative foundation for his
critical social theory.
5.
Objections
There are three principal lines of objection to Habermas use of formal pragmatics
as the underlining for his critical project. The first casts doubt on the ability of formal pragmatics to provide a foundation for social critique that is sufficiently robust
306
Maeve Cooke
for his purposes. The second criticizes the specific shape of the conception of reason that emerges from his formal-pragmatic analyses. The third claims that his appeal to formal pragmatics, and the vision of a communicatively rationalized lifeworld that he derives from these analyses, amounts to a foreshortened idea of
human flourishing.
1) The first line of objection challenges the ability of formal pragmatics to provide an empirically based, normative foundation for social critique; this objection
can be further divided into three sub-objections: a) an objection to the universality
of the idealizations built into everyday communicative practices; b) an objection to
the idea of the harmonious, free interplay of the three cultural value spheres/moments of reason as a standard for criticizing the colonization of the lifeworld; c) an
objection that Habermas does not show the primacy of the communicative use of
language over the strategic one.
a) The criticism here is that the concept of communicative action that Habermas
extracts from his formal pragmatic analyses is too weak for the critical purposes he attributes to it (Benhabib 1992; Cooke 1994; Cooke 2006). Some of
the idealizations he identifies as necessary presuppositions of the communicative use of language are not a feature of language use in general; they orient
communicative practices only in certain socio-cultural contexts as a result of
specific historical developments; in fact, they seem to be features of language
use only under conditions of modernity, in societies in which there has been a
desacralization of knowledge, a secularization of authority and an internalization of the principle of universal moral respect. Since the critical force of the
concept of communicative rationality depends to a considerable degree on
idealizations relating to social inclusiveness and equality, it is particularly
worrying that they appear to be features only of certain kinds of modern society. This presents Habermas with the following dilemma: on the one hand, if
he wants to maintain a formal-pragmatic basis for his critical perspective, he
will have to acknowledge the socio-cultural specificity of the resulting concept
of communicative rationality, significantly restricting the scope of its critical
power; on the other hand, if he wants to offer a critical perspective that is not
restricted to particular socio-cultural contexts but rather universal in scope, he
will have to supplement his formal-pragmatic analyses with arguments of other
kinds (Cooke 2006: 5761).
b) This objection queries the usefulness of the metaphor of the harmonious, free
interplay of the three moments of reason for the purposes of critique of the colonization of the lifeworld by functionalist rationality. It points out that Habermas critique of colonization appeals to a normative idea of evenly balanced,
non-selective processes of societal rationalization that calls for a balanced relationship between communicative rationality and functional rationality. The
difficulty is that the balance evoked by the metaphor of three harmoniously interrelating moments of reason is not a balance between communicative
307
rationality and functional rationality, but rather a balance between cognitiveinstrumental, moral-practical and aesthetic-expressive rationality. For the
metaphor to work, we would have to be able to equate functional rationality
with cognitive-instrumental rationality. Habermas, however, insists that they
are not equivalent: whereas instrumental rationality is the reasoning employed
by social agents to achieve particular, intended ends, functional rationality bypasses the intentions of the agent (see Cooke 1994: 144146).
c) The nub of this objection is that in order to assert the primacy of the communicative use of language Habermas needs to show that manifestly strategic action, too, is parasitic on the communicative use. The objection is not just that
Habermas shows only the parasitic nature of latently strategic action; it is the
stronger point that it is impossible in principle to show which use of language is
most fundamental without appealing to controversial philosophical presuppositions concerning the rational use of language; indeed it is impossible to decide whether there is a fundamental use of language at all (Apel 1998). In other
words, formal-pragmatic arguments will never suffice to establish the primacy
of any mode of language use. A related objection is that even successful demonstration of the primacy of the communicative use of language would leave
open a number of important questions; these include questions such as how
much communicative action is necessary for the maintenance of social order
and social reproduction, the extent to which the erosion of communicative action can count as colonization and the degree of reflexivity, inclusiveness and
openness communicative action has to display in order for social reproduction
to count as successful (Cooke 1994: 2428).
2) The second line of objection relates to the specific shape of the concept of communicative rationality that emerges from Habermas formal-pragmatic analyses.
Critics have drawn attention, in particular, to his sharp separation between the use
of language oriented to mutual understanding and the use of language to disclose
new perspectives (Cooke 1994: 7484). This produces an impoverished account of
communicative reason that excludes modes of language use that bring us to see the
world in a new way from the category of validity-oriented and, hence, rational communication; ultimately, it leads to an inability to explain the power of critical social
thinking to engage the imagination as well as the reason of its addressees, thereby
motivating them to work towards social change for the better (Cooke 2006).
3) The final line of objection is directed at the very project of formal pragmatics. One of the strongest criticisms of Habermas in this regard comes from
within the Frankfurt School tradition of critical social theory. Honneth, described
above as a third generation Frankfurt School critical theorist, commends Habermas
for the paradigm shift he introduces into this tradition of thinking through replacing a subject-object model of cognition and action by an intersubjective one. However, he discerns certain biases in Habermas critical social theory, which he traces
308
Maeve Cooke
back to its reliance on formal pragmatics. The key point is that by reducing practical reason to a conception of rationality that is extracted from an analysis of
everyday linguistic practices, Habermas adopts a critical perspective on society
that is blind to forms of human suffering, and corresponding forms of human flourishing, that cannot be conceptualized as failures of linguistic communication. His
critical perspective has no room, for example, for critique of violations of bodily
integrity, and the corresponding idea of human flourishing as involving bodily selfconfidence, or for critique of the disaffection and alienation that result from lack of
recognition of the value of ones work, and corresponding idea of human flourishing as involving self-esteem based on ones achievements. For Honneth, this leads
to a regrettable narrowing of the scope of social criticism, permitting only normative deficits of certain kinds to come into view. The objection, in sum, is that
Habermas theory fails to exploit the full potentials of an intersubjective framework, leading to a form of critical social theory that is too thin and too weak to provide a basis for critique of the principal pathologies of contemporary capitalism.
His alternative proposal is for an intersubjective model that does not prioritize the
linguistic aspects of human flourishing but takes account of its bodily-affective and
creatively-productive ones as well.
Habermas has responded to some but not all of these points of criticism. Moreover, in those places where he appears to have acknowledged the force of the criticisms, his responses have often been indirect. They manifest themselves in a gradual
moving away from the strong reliance on formal pragmatics that informed his theory
of communicative action towards more critical-hermeneutic and historicist lines of
argument (Cooke 2006: 71). To be sure, Habermas continues to afrm the importance of formal-pragmatic arguments in his critical project (Habermas 1996: 109). We
may take this as evidence of his unwavering commitment to a conception of reason
that, though historically situated, transcends times and places. Historicist and hermeneutic approaches, although productive in many ways, are unable to account for context-transcending validity. Were Habermas to dispense with formal pragmatics, or
even signicantly to reduce his projects reliance on it, he would have to nd some
other way of accounting for the context-transcending power of rational critique.
Notes
1. Important exceptions include sociologists who embrace systems theory as inspired by
Niklas Luhmann and structuralism as inspired by Ferdinand de Saussure. Both systems
theory and structuralism conceive of the deep structures that generate society as subjectless systems of rules. This also holds for poststructuralists such as Michel Foucault.
2. As we shall see, this position by no means implies the need for intentionalist accounts of
linguistic behaviour, which privilege the speakers intentions and neglect both the importance of what is said and the intersubjective relationship established in every com-
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
309
municative use of language. Nor does it imply that the speakers account of her own intention is incontrovertible.
Habermas views the tension as potentially positive insofar as it serves as a reminder that
both perspectives have their strengths and weaknesses. Although, as we shall see, he
makes a methodological decision in favour of the internal perspective, his theory of
communicative action acknowledges the importance of the external perspective, most
obviously when it offers a causal explanation of the pathologies of modern societies in
terms of the colonization of the communicatively regulated domains of social life by the
functional rationality of the economic and administrative systems.
These questions are further expounded in his longer study, The Logic of the Social
Sciences (Habermas 1988), an earlier version of which appeared as an article in the Philosophische Rundschau in 1967.
I have summarized Habermas discussion here. In the essay in question, Habermas distinguishes between three kinds of generative theory of society because he includes theories that do not take subjective meaning as a sociological primitive. Thus, in addition to
the two I mention, he identifies system theories that lack a subject (Habermas 1971:
18). See note 4 above.
Verstndigung (n.): reaching understanding, mutual understanding or communication. The corresponding verb is sich verstndigen. The term is ambiguous. It embraces linguistic comprehension (Verstehen) but goes beyond this to refer to the process
of reaching understanding, in the sense of reaching an agreement with another person or
persons.
When Habermas refers to formal semantics he is thinking of the philosophical theories of meaning that take their inspiration from Gottlob Frege and Betrand Russell and
in the second half of the twentieth century were associated with thinkers such as Donald
Davidson and Michael Dummett; Habermas mentions Peter Strawson, but his place in
this tradition is not straightforwardly evident (Habermas 1998: 99, note 62).
Habermas sometimes describes such claims as universal in the sense of transcending
times and spaces (e.g. Habermas 1994: 42). However, when he refers to universal validity claims on other occasions, he seems to mean universal in the sense of present in
every speech act (e.g. Habermas 1998: 89).
On Bhler (and Jakobson) cf. Daalder and Musolff this volume.
Habermas also asserts the primacy of the communicative use of language over what he
calls the indirect use of language such as the figurative and the symbolic (Habermas
1998: 166). Since his argument to this effect is cursory, I will not discuss it in the following. His neglect of the figurative and the symbolic uses of language, and demotion of
them to a subordinate status, is symptomatic of his overly sharp distinction between reason and imagination (see Cooke 1994: 7484 and Cooke 2009).
This is true irrespective of whether the intention to deceive or manipulate is conscious
or unconscious. Habermas distinguishes between conscious and unconscious latently
strategic action. Unconscious latently strategic action is correlated with what he calls
systematically distorted communication (Habermas 1998: 169). I cannot adequately
discuss the latter topic in the present context; I return to it briefly in the next section.
See note 13 below.
For twenty years subsequent to the publication of The Theory of Communicative Action,
Habermas had little to say about the problem of cultural impoverishment; in his most recent writings, however, he directly addresses the related question of semantic renewal,
310
Maeve Cooke
drawing attention to the potential semantic productivity of the beliefs and traditions of
the major world religions (Habermas 2008).
14. Habermas sees Immanuel Kant as the first philosopher to take seriously the differentiation of reason into its moments (Habermas 1990: 120, esp. 23).
15. I leave open the question of whether Hegel thought this process of self-actualization
would ever be concluded.
References
Adorno, Theodor W.
1981
Cultural Criticism and Society. In: Theodor W. Adorno Prisms. Trans. by
Samuel and Shierry Weber. Cambridge, MA: MIT Press.
Apel, Karl-Otto
1998
Openly Strategic Uses of Language: A Transcendental-Pragmatic Perspective.
In: Peter Dews (ed.), Habermas: A Critical Reader, 272290. Oxford: Blackwell.
Apel, Karl-Otto
1980
Towards a Transformation of Philosophy. Trans. Glin Adey and David Frisby.
London: Routledge and Kegan Paul.
Austin, John L.
1962
How to Do Things with Words. Oxford: Oxford University Press.
Benhabib, Seyla
1992
Situating the Self. New York: Routledge.
Bhler, Karl
1934
Sprachtheorie. Jena: Fischer.
Chomsky, Noam
1969
Aspects of the Theory of Syntax. Cambridge, MA: MIT Press.
Comte, August
1893
The Positive Philosophy of Auguste Comte. Trans. H. Martineau, 3rd edition.
London: Kegan Paul, Trench, Trbner and co.
Benhabib, Seyla
1992
Cooke, Maeve
1994
Language and Reason. A Study of Habermass Pragmatics. Cambridge, MA:
MIT Press.
Cooke, Maeve
2006
Re-Presenting the Good Society. Cambridge, MA: MIT Press.
Cooke, Maeve
2009
Zur Rationalitt der Gesellschaftskritik. In: Rahel Jaeggi and Tilo Wesche (eds.),
Was ist Kritik?, 117133. Frankfurt/Main: Suhrkamp.
Dilthey, Wilhelm
1996
Hermeneutics and the Study of History. Edited by Rudolf Makkreel and Frithjof Rodi. Princeton, NJ: Princeton University Press.
Fraser, Nancy and Axel Honneth
2003
Redistribution or Recognition? A Political-Philosophical Exchange. London:
Verso.
311
Habermas, Jrgen
1971
Reflections on the Linguistic Foundation of Sociology: The Christian Gauss
Lectures (Princeton University, February-March 1971). In: Habermas 2001,
1104.
Habermas, Jrgen
1974
Reflections on Communicative Pathology. In: Habermas 2001, 129170.
Habermas, Jrgen
1984
The Theory of Communicative Action, vol. 1. Trans. Thomas McCarthy. Boston: Beacon Press.
Habermas, Jrgen
1986
A Reply. In: Axel Honneth and Hans Joas (eds.), 214264.
Habermas, Jrgen
1987
Habermas, Jrgen
1988
On the Logic of the Social Sciences. Trans. Shierry Weber Nicholsen and Jerry
A. Stark. Cambridge, MA: MIT Press.
Habermas, Jrgen
1989
The Structural Transformation of the Public Sphere. Trans. Thomas Burger
with the assistance of Frederick Lawrence. Cambridge, MA: MIT Press.
Habermas, Jrgen
1990
Moral Consciousness and Communicative Action. Trans. Christian Lenhardt
and Shierry Weber Nicholsen. Cambridge, MA: MIT Press.
Habermas, Jrgen
1991
The New Conservatism: Cultural Criticism and the Historians Debate. Trans.
Shierry Weber Nicholsen, Cambridge, MA: MIT Press.
Habermas, Jrgen
1992
Postmetaphysical Thinking. Philosophical Essays. Trans. William Mark Hohengarten, Cambridge, MA: MIT Press.
Habermas, Jrgen
1996
Between Facts and Norms. Trans. William Rehg, MIT Press, Cambridge, MA:
MIT Press.
Habermas, Jrgen
1998
On the Pragmatics of Communication. Trans. and edited by Maeve Cooke.
Habermas, Jrgen
2001
On the Pragmatics of Social Interaction. Preliminary Studies in the Theory of
Communicative Action. Edited by Barbara Fultner. Cambridge, MA: MIT
Press.
Habermas, Jrgen
2008
Between Naturalism and Religion. Trans. Ciaran Cronin. Cambridge, MA:
MIT Press.
Honneth, Axel
1994
The Social Dynamics of Disrespect: On the Location of Critical Theory Today.
Constellations 1(2): 255269.
Honneth, Axel
2003
The Point of Recognition. A Rejoinder to the Rejoinder. In: Nancy Fraser and
Axel Honneth, 237247.
312
Maeve Cooke
Honneth, Axel and Joas, Hans (eds.)

1986
Communicative Action. Cambridge: Polity.
McCarthy, Thomas
1978
The Critical Theory of Jrgen Habermas. Cambridge: MA: MIT Press.
Weber, Max
1969
Basic Concepts in Sociology. Trans. H. P. Secher. Westport, Conn: Greenwood
Press.
2001
Philosophical Investigations. Trans. Gertrude E. M. Anscombe. Oxford:
Blackwell.
Part III
Key topics in pragmatic description
11.

William F. Hanks
1.
Introduction
The study of deixis is central to pragmatics, because deictic systems define points
of intersection between linguistic structure and the social settings in which speech
takes place. It is standard in contemporary English language sources to distinguish
Spatial (here, there), Temporal (now, then, tense), Person (pronominals), Discourse
(coreference or reference to prior talk) and Social Deixis (honorification and any
indicators of social identity or status relations among participants and contexts).1
Any one of these functional foci may be formally marked in utterances morphologically, lexically, or by construction type. Deictic phenomena inevitably interact
with gesture (conventionalized or not) and with many other parts of the grammar,
including evidentiality, status, modality, aspect, noun classification, possession,
spatial and temporal description. This chapter focuses on those forms whose primary function is to individuate objects of reference (including events, material
things, talk itself or any individuated concept). Referential deixis is found in all
human languages and includes at least demonstratives, person markers, locative,
directional and temporal markers, but excludes much social deixis such as honorification (where social status is indexed but usually not singled out for comment)
and standard sociolinguistic markers (where social factors are indexed but not
singled out for comment). By starting from the distinctive structure of referential
deixis, one can better characterize its relation to grammar and to other varieties of
indexicality. The main grammatical question is how deictic functions are encoded
in languages. The challenge for pragmatics is to determine the constant capacity of
deictic types to contribute to token acts of situated referring, and the relation between such referring and other aspects of utterance context.
Although they appear in various grammatical categories, deictics are all part of
a single field, which I will call the deictic field (Bhler 1934). The deictic field is
a psycho-social construct composed of: (1) the roles of communicative agents
relative to the participant frameworks they occupy, that is, the relations between
Speaker (Spr), Addressee (Adr), and Others, as defined by the language and communicative practices of its speakers; (2) the positions occupied by objects of reference, that is, the kind of object and its role in the activity in which the deictic
utterance is produced; and (3) the multiple dimensions whereby participants have
cognitive access to objects, such as perceptibility, relative proximity, memory, anticipation and prior talk. When a Spr performs an act of deictic reference, as in
thats John right over there (punctual point), we came from over there (vague
316
William F. Hanks
point), or heres your coffee (extending in hand), she produces a social relation
between herself, her interlocutor and the object. The deictic field is a template
for the elementary social relations on which conversation rests. The remainder of
the paper is organized into two parts, theoretical approaches and typological approaches.
2.
Theoretical approaches to deixis
Deictics are what Sacks (1992) called indicator terms and what linguists and
philosophers have called variously indexicals,2 shifters3 or, as in the present
paper, deictics (Fillmore 1997, Hanks 1983, 1990). In his theory of semiotics (cf.
Nth this volume), CS Peirce (1955) distinguished indexes from icons and symbols, according to the relation between the perceptible sign vehicle (or representamen) and its object (that for which it stands). Icons relate to their objects by way
of similarity, symbols by way of sheer convention, and indexicals by way of dynamical contiguity. As Peirce (1955:107) put it, the index stands for dynamical
(including spatial) connection [of the deictic form] both with the individual object,
on the one hand, and with the senses or memory of the person for whom it serves as
sign, on the other. Certain indexes are in a causal relation to their object (smoke
indexes fire, a weather vane indexes wind direction) and others are conventionalized (this indexes objects close at hand, there places at a remove from the location of utterance, road signs index road conditions). In the prototypical cases, the
index does more than stand for its object, it directs an addressees attention to it,
and this directive force is especially salient in certain types of deixis (e.g. presentatives, such as Here it is (take it!), There he goes! (look!)). Such expressions do
more than individuate an object; they direct the addressee to it, and the proper uptake is usually a gesture or a shift in attention focus.
Morris (1946) simplified the Peircean framework, distinguishing characterizing signs (icons and symbols, both of which characterize their object) from indices
(which are devoid of descriptive information about their object) and universal
signs (anyone, anywhere, etc., which are applicable to any item in a class). From
this perspective, what is noteworthy about deictics is that they contribute to individuated referring without in any way describing their objects. As Sacks (1992:
520) put it, such terms, which he called indicators, are altogether abstract.
[] capable of invoking the sheer fact of the setting, without in any way formulating it.
The definition of deixis in terms of individuation without meaning is common
in philosophical literature (Searle 1969:80, Nunberg 1993, Bar Hillel 1954), but
has less prominence in the linguistic literature, where cross-language comparison
and detailed description reveals many conventionalized features in deictic systems
(see below). While Morriss framework played a pivotal role in the development of
317
linguistics (cf. Nth this volume), his definition of indexicality is too oversimplified to describe natural language deixis. By contrast, Peirce insisted that his three
sign types designate relations which typically combine in actual signs. The conventionality of any language guarantees that the pairing of form with meaning in
the language is conventional, no less so for deictics than for other kinds of expression. Moreover, this conventionality encompasses deictic features such as
punctate vs. regional space, centripetal vs. centrifugal movement relative to utterance context, kinds and degrees of perceptual access to the objects denoted by demonstratives, the presence or absence of gender, animacy and noun class in demonstratives, and any special presentative forms in a language (see French voici,
voil, Yucatec Maya hebe there it is (listen!).The result is that actual deictic
forms and arrays always combine sheer indexical functions with other kinds of
information (indexical, iconic or symbolic). If one maintains the more austere
philosophical definition of deixis as pure indexicality, we are led to conclude that
natural language systems are always impure blends between indexical and non-indexical functions. A more useful and realistic alternative is to conclude that deixis
in natural language requires a more robust Peircean definition of the phenomenon,
and challenges researchers to explain why certain bundles of functions cluster
around indexicality in language after language. This point is all the more clear
when we consider that deictics are typically organized in closed class paradigms,
where obviously indexical functions are bundled with a wide array of other information.
Natural language deictics are what Bloomfield (1933) called category representatives, that is, they bear features for morphosyntactic category and function as
syntactic heads. Some but not all deictics may be used either exophorically, for reference to extra-linguistic context, or endophorically, for anaphoric co-reference
with prior talk (A: I prefer this one, B: Oh, I like that too) or cataphoric co-reference with subsequent talk (the two statements have this in common, they are
terse) (see Halliday and Hasan 1976). While there are borderline cases where this
distinction is difficult to draw, there are also regular patterns in which deictic types
are specialized for one or the other. Anaphora, for instance, is often associated with
forms which when used exophorically indicate remoteness of the object, while spatial adverbial paradigms often include one or more forms that can be used only exophorically. They also differ as to whether they may be lexically expanded, as in
here vs. here in my room, this vs. this blue pencil, and so forth. It is common
to describe the simple forms as pro-locative, pronominal, and so forth, and the expanded forms as adlocative, adnominal and so forth. Moreover, deictics in many
languages may be combined in distinctive n-tuples of coreferential or mutually reinforcing forms (this here book, Yucatec leti e hela < the one the here.it.is =
this here one, te waya < right.there here = right here [point within a region]).
It is also common to find distinctions in spatial deixis according to the regional
scope of the place referred to, as in Yucatec tela right here (punctual) vs. waye
318
William F. Hanks
here (regional). By contrast, English here may denote a point on the speakers
nose, the room, house, neighborhood or country in which it is uttered, depending on
context, and similar extensions apply to there, now, then, and others. In short,
when one examines the pragmatics of actual deictic systems, many features arise
that are elided in the austere definition of indexicality as semantic emptiness.
In his classic work Sprachtheorie, Bhler (1934) initiated what would become
another line of thought concerning deixis in language. Bhler distinguished three
dimensions in the elementary speech event, the Speaker, the Addressee and the Object talked about. Intersecting all three poles is the division between what he called
the Zeigfeld, the demonstrative field in which objects can be shown, and the Symbolfeld, the discourse field to which description of the object, and other discourse
belongs. This relatively simple schema already generates six values (three in each
field) and provides the elementary framework for a functional analysis of deixis:
deixis and demonstration belong to the field of exophoric reference to objects
available in the immediate perceptual field. Anaphora, reference to prior talk and
to remembered objects are treated as blends of the Zeigfeld and Symbolfeld.
Jakobson (1960, 1971) expanded and refashioned Bhlers distinctions into the
division between the speech event (Es) in which talk occurs and the narrated event
(En) talked about. Jakobson eschews the term deixis, but treats all referential indexicals under the rubric of shifters (a term coined by Jespersen 1965: 290).
Shifters are linguistic categories whose distinctive feature is that they individuate a
referent (in the En) in relation to the situation of speech (Es). Thus if Ben tells Katherine I saw him right here yesterday, every item but the verb see is a shifter
because each one establishes a relation between the utterance and the event narrated: I equates the current speaker with the seer, the past tense of the verb places
the En in a time prior to the moment of utterance (reinforced by yesterday, the
day prior to the day of utterance) and him indicates that the person seen is neither
the current Spr nor Adr. Jespersen adopted the label shifter in light of the distinctive behavior of these grammatical categories in reported speech. If Katherine subsequently reports Bens utterance to Madeleine, she says Ben told me he saw him
there yesterday. In reported speech the shifters shift to accommodate the new
speech event frame of the quoted speech, while maintaining sameness of reference
to the event of seeing and its participants. Inversely, in direct quotation, the linguistic forms remain constant, but the reference is dislocated. If Katherine directly
quotes Ben, she says Ben said I saw him here yesterday. Notice that despite the
I, here in this utterance, Katherine does not therein refer to herself or to the location in which she reports the speech. In effect, the introductory phrase Ben said
projects a prior Es in which the utterance was produced, and it is that prior event
that serves as the ground of reference for the deictics. Hence there is no shift in the
forms, but the speech event relative to which they are interpeted has been transposed. Silverstein (1976) follows Jakobsons shifters model and expands it to a
wide array of indexical effects, only some of which contribute to reference. Essen-
319
tially the same set of distinctions underlies the still more elaborate analysis of the
speech event in the ethnography of speaking.4
Phenomenologists were centrally concerned with indexicality (Husserl 1978,
Schutz 1970, Merleau-Ponty 1967) as evidence of the unavoidable situatedness
of speech in the immediate experiential settings of its production. While phenomenology has little impact in English language linguistics, it is fundamental to ethnomethodology (Garfinkel 1967, 1972) in its claim that all meaning is (indexically) produced in the ad hoc processes of interaction, and similarly to the
approaches to interaction pioneered by Cicourel (1964, 1974, 1985, 1992), Schegloff (1972, 1982), Gumperz (1982) and Goodwin (1981, 2003). From a psycholinguistic perspective, Rommetveit (1974) combines phenomenology with linguistics
to argue that linguistic production relies not on abstract semantics, but on the
experiential worlds of speakers. Conversation analysis as developed by Sacks,
Schegloff and others carries forth the basic premise that speech is pervasively indexical and grounded in the interaction between co-participants. Developing the
structural linguistics of Meillet (1926), Martinet (1960) and the Prague School,
Benveniste (1965, 1974) incorporated phenomenology in his landmark papers on
subjectivity in language.
This range of approaches to indexicality raises a number of issues. There is
widespread agreement that indexicality designates the context dependency between utterances and speech contexts, and that deictic types are specialized in individuating single objects or groups of objects in relation to the participants in the
event of speaking. We can schematize this as in Figure 1, where the Indexical
ground is the zero-point, or pivot relative to which the deictic points to the object
denoted. Thus for example, in an utterance like its sunny here, the locative here
projects a relation of relative immediacy to or inclusion of the Speakers location at
the moment of utterance.
Relation
Indexical ground
Object
Figure 1. Relational structure of referential deixis
This three-way distinction raises questions on which there is considerable debate

in the literature. First, what kinds of relational features are encoded in deictic
types? The spatialist view holds that relative contiguity (this = proximal, that =
non-proximal and so forth) is fundamental, and other pragmatic effects are incidental to deixis or derived by metaphorical extension (Anderson and Keenan1985)
or contextual inference (Levinson 1983). For careful assessment of these alternatives, argued with data from Lao, see Enfield (2003, 2009). By combining a spatialist definition with Gricean implicature (Grice 1967, 1978) and conversational
organization, the enriched values of deictic tokens can be calculated as implica-
320
William F. Hanks
tures. By contrast, recent non-spatialist views reject the a priori primacy of space,
and substitute for it something closer to accessibility, that is, how the participants
have or gain access to the object.5 Spatial contiguity sometimes provides a support
for accessibility, but is only one possible parameter alongside others, including
perception (via any of the senses), prior talk and memory, salience (cognitive or
perceptual), evaluative stance, relative familiarity with the object, and in some
cases ownership of the object by a participant. Deictics may be relatively lean in
terms of descriptive information, but the pragmatics of deixis nonetheless responds
systematically to the features of actual objects. Where the spatialist approach presumes that space is basic across all contexts, the accessibility approach looks to
situational relevance, the interactive move made by the deictic utterance, the social
definition of the object and other features of context (Hanks 2005a).
The second major question bears on how one defines the indexical ground.
Traditional treatments assume that the zero-point of deixis is the speaker, as in
here means a place close to me. One of the earliest statements of what can be
called the egocentric approach is Russells description of egocentric particulars
(1940), which cover the range of deictics (compare Evans 1982: Chapter 6, Gale
1968). Most traditional grammars and typologies (eg. Anderson and Keenan 1985)
assume egocentricity, and Benvenistes (1966, 1974) anchoring of indexical reference in subjectivity makes the same move (cf. Lyons 1982: 121). With the exception of Schutz, whose emphasis was intersubjectivity, most phenomenology also
proceeds from the ego-subject (albeit with different inflections). Egocentricity is
consistent with the commonsense idea that the solitary speaker is the one who produces her utterance, and the minimal requirement for utterance production is a
speaker. The challenge for this position is to explain the many cases in which a
deictic responds not to speaker accessibility, but to the addressees access to the
object, or to the relative symmetry of access between the co-participants. Moreover, while the idealization of the solitary speaker has been productive for certain
purposes, Conversation Analysis played a leading role in showing it to be patently
inadequate as a basis for pragmatics.6
There are various alternatives to egocentricity. Reichenbach (1947) described
indexicals under the heading of token-reflexivity, by-passing the problem of the
Speaker in favor of the utterance itself as zero-point. The utterance here denotes
a place immediate to the utterance token here, just as I denotes the individual
who produces the token I. This removes the problem of the zero-point from the
realm of interaction, stating in effect that these forms denote in relation to instances of themselves. Close description of actual deictic usage indicates that elements such as common ground,7 collaboration between participants (Goodwin
1981, 2003), and the relative symmetry of participant access (whether speaker and
addressee have roughly the same or significantly different access to the object in
situ; Hanks 1990, 1992) all impact on the use and understanding of deictic tokens.
Any of these factors may be conventionalized in the semantics of deictic types.
321
This suggests that the zero-point of indexicality is best conceived as the participation framework in which talk is produced. Just as space may be a key kind of access for some uses, the speaking ego may be the ground for some uses. But neither
space nor egocentricity is a priori privileged.
These different approaches to deixis imply different background pictures of utterance context. The egocentric spatial picture holds that deictic acts occur when
the Spr, the Adr and the Object are physically copresent and mutually perceptible.
The zero-point of the deictic utterance is the Sprs body and the utterance orients
the Adrs attention to a specific sector of the broader sphere of spatial proximity,
not unlike a pointing gesture (which may indeed be part of the utterance). To
understand a deictic is not to interpret it but simply to recognize by observation
what it singles out in the physical situation of utterance. The space of dialogue
may be interperspectival, but it is the Spr who produces the utterance and his or
her body that serves as the indexical anchor-point: a self-contained individual,
oriented in physical space and endowed with a sphere of contiguity and sensory access. This picture is dominant in much of the English language literature on deixis.
It is found in most modern linguistic descriptions, especially typological and psycholinguistic work, for which it seems to offer a controlled basis of cross-linguistic
comparison.8 Despite its appeal, however, this picture omits critical features of
deictic practice, including the mutual orientation of interactants, all non-perceptual
modes of access such as background knowledge, memory or anticipation, and all
extra-physical aspects of social settings. Non-spatial aspects of deictic speech are
either ignored, treated as metaphorical extensions from space, or derived from
other, non-deictic principles.
There is a second, widely attested background picture for deixis, according to
which the core of context is social interaction.9 In this view, deictic utterances involve at least the following factors.
(i) The reciprocal perceptual fields of the parties, especially reciprocal vision in the
face-to-face (Schutz 1970, Goodwin 1981, 1994, 2000, Kendon 1992).
(ii) The broader reciprocity of perspectives whereby each party assumes that the
other has a perspective and that if (s)he adopted that perspective, the world would look
to her as it does to the other (Schutz 1970). This is precisely reflected in the alternation
of I, you, here, there among Spr and Adr: each party to talk is an I, but only to himself, and each one is a you, but only to the other, my here is your there and so forth.
(iii) Participants bodies are expressive through gestures, as well as receptive through
the senses, and gesture is a key aspect of deixis.10
(iv) Any interaction has sequential organization, perhaps the best studied feature of
conversation. Sequential organization is embodied in turn taking, adjacency, contiguity
and proximateness.11
(v) The interaction has a motivational structure: the Spr speaks in response to something and in order to achieve something else; the Adr responds in the light of the aim of
the first utterance, in order to achieve his or her own aim, which then serves as the because motive of the next turn, and so on back and forth (Schutz 1970).
322
William F. Hanks
(vi) Any utterance has a relevancy structure according to which certain things matter
and others dont.12
(vii) Utterance meaning must be negotiated or worked out by the co-engaged parties.
It is not given in advance, nor is it fixed by the intentions of the Spr (Clark, Schreuder
and Buttrick 1983). For deixis, this means that, for instance, the reading of a deictic
token as spatial proximity, perceptibility, cognitive salience, egocentric or altercentric
can vary by context and must be worked out by the participants.
(viii) Participants display to one another their sense of the current situation, the relevancy structure they are assuming, and their current relation to their own speech. Deixis is a primary resource for such display precisely because it points into the situation
and thereby positions the Spr.
In the Interactionist picture, there is an emergent space of interaction, but it is not

the kind of egocentric, physical space presumed by the Spatialist view (Kendon
1992), and to study it, we must attend to sequential organization of talk, to situational variation and to the micro-ethnography of everyday usage. In the course of
deictic practice, interactants must jointly establish and display the relevance of
spatial perceptual, discursive or other contextual frames. Moreover, in much ordinary referring, sequential placement of a deictic utterance in talk combines with
spatial location of the object in the course of identifying the denotatum. Ordinary
practice therefore blends endophoric with exophoric functions.
For the purpose of deictic analysis, situational relevance has two main sources:
(1) local relevance turns on the immediate circumstances of utterance, including
sequential placement of the utterance, the location of the object and the positions
of the interactants in the interactive field, and (2) global relevance turns on the embedding of the deictic field in a larger social field. The broader social setting can
define the boundaries of here for the purposes of talk, and factors such as social
identity or hierarchy and activity spaces may license extensions, transpositions, or
they may constrain usage. For instance, if a foreman on a work site says I am
installing the electric outlet here, the utterance can be fully normative even if
the foreman does none of the actual work. The hierarchical role of the foreman
licenses first person singular usage for tasks of which he is the leader but not literally the one who does the work. Similarly, spatial references in the extended
households of Yucatec speakers respond systematically to the social organization
of space, so that here may refer to the entire household or exclusively to the portion occupied by the Spr. Shamans and doctors who use imaging technologies
(divining crystals, x-ray images and the like) routinely make deferred ostensions
in which pointing to the image is pragmatically equivalent to pointing to its bodily
counterpart. In cases like these, the activity space in which the deictic practice
takes place provides specific affordances and relevancies that help motivate practice.
The Spatialist and Interactionist picures of utterance context can be combined
so long as we treat them as complementary: spatialism is a claim about the re-
323
lational features of deixis, and interactionism is a claim about the indexical ground.
We can grant interaction as the ordinary context of utterances, while still claiming
that the semantics of deixis is egocentric and spatial. In this case, something akin to
Gricean implicature is invoked to derive contextual effects from what is literally
spatial reference. At the same time, conventionalized non-spatial features, such
as perceptual access, are taken to be non-deictic. Recent work by a variety of
scholars has combined grammar and interaction, thereby opening the possibility of
a blended account of deixis.13
While some combination of linguistic and interactional analysis is undoubtedly
necessary to describe the pragmatics of deixis, it is not without cost. If the interactionist grants that verbal deixis is fundamentally spatial in its semantics, then interaction becomes an epiphenomenon of grammar; it is merely how people ordinarily use linguistic forms whose meanings are fixed apart from context. This is
surely contradictory to the work of ethnomethodologists such as Garfinkel (1967),
who took the strong position that all meaning is produced in the course of ordinary
talk, and it also contradicts the more recent position that talk is inherently situated.
Similarly, the spatialist position is reduced to an idealization if one grants that
neither egocentricity nor spatial relations are necessarily salient in ordinary interaction. Without recourse to the counter-evidence of actual usage, and protected by
the vagueness of Gricean implicature, it is hard to see what evidence could be adduced either for or against spatialism.
What the two background pictures have in common is that they take the conditioning circumstances of pragmatics to be local to the utterance: in the spatial view,
it is the field of proximity to the speaker, and for the interactionist, it is the immediate sequential context, augmented by those features of context to which participants are oriented (Schegloff 1987). Linguistic anthropologists have questioned
the localism of both approaches, in an effort to understand the how broader social
values and definition of the occasions of talk systematically affect deictic practice
(Hanks 1990, 2005a, Enfield 2003, Goodwin 2006, Silverstein 1976). Cross-cultural descriptions of talk suggest that there is no value-neutral way to define the
dynamical contiguity that is at the core of deixis. Starting from the linguistic
structure of deictic systems, to the local circumstances of deictic acts, this third approach proceeds to the social fields in which the deictic field is embedded (Hanks
2005a, 2006a). This includes the identities of participants, the genres of practice of
which deixis is a part, the social definition of place and time and the values attaching to objects of reference. The resulting framework is less elegant than formal accounts, but more empirically precise to the varieties of deictic practice that we find
in actual talk.
324
3.
William F. Hanks
Typological approaches
Deictic systems can be compared along a number of axes according to their grammatical and functional dimensions. The first point of comparison is an inventory of
the grammatical categories for which the forms are marked: Noun (pronouns, articles, demonstratives, e.g. I, the, this, that), Adverbs of space (here, there), time
(now, then), manner (like this, like that), Presentative forms (French voici, voil,
Maya hela here it is (giving), helo there it is (pointing), hebe there it is
(sniffing, cocking ear), Verbs (especially motion verbs like come, go, see Fillmore 1997), and various elements that attach to major categories (directionals, as
in Mayan verb stems; clitics and utterance or phrase initial or final particles).
Prototypical deictics are category representatives that can substitute for, or be expanded by lexical descriptions (this vs. this red pen). In the former case, the
deictics is pro-nominal, and in the latter it is ad-nominal. In some languages the
pronominal forms and the adnominal qualifying usages correspond to different
series, as in Mulau, a Daic language, and Japanese (Diessel 1999: 59). Diessel
(1999: 58) reports that in twenty-four of the eighty-five languages he sampled, pronominal and adnominal usages are marked by different categories, and in languages such as Turkish and Lezgian, in which the same deictic roots are used in
both functions, the pronominal forms inflect for case, whereas the adnominal ones
do not.14
The grammatical classification of deixis raises various questions, of which two
deserve mention here. The first is whether all or part of certain category-specific
subsystems are derived from forms in other categories. For instance, adverbial
deictics in some languages are derived from nominal ones, as in Quechua (kay
this > kay-pi here, chay that [medial] > chay-pi there, haqay that [dist] >
haqay-pi there [dist] (Calvo Perez 1999, Lasater 2009). Temporal deictics are
often derived from spatial ones, as in Yucatec tol-akhak-o back then [remote
past] < tol-o out there [remote], or from non-deictic lexemes (cf. today, yesterday, aujourdhui, Maya behelae today (lit. road.here.it.is).
The second question is the degree to which the different subsystems in any language are similar or different in structure. Do we find the same series of oppositions in different subsystems, as in French ceci, cela (nominal), ici, l (spatial)
voici, voil (presentative) or English this, that, here, there, now, then? The apparent proportionality of deictics across catgories has encouraged many scholars to
propose paraphrase relations among deictics, such as this is the one here, that
is the one there, here is the place occupied by I, and so forth.15 Both the derivational relations between subsystems and the relative proportionality of oppositions go to the larger question of the degree to which deixis in any language is a
single system unified by shared forms, regular formations or recurrent arrays of
distinctions. Comparative research has demonstrated that languages differ on this
axis.
325
A closely related axis is the degree of compositionality of deictic forms. Languages like English and Hausa have monomorphemic deictics.
nn
nan
cn
can
Prox to Spr (immediate field)

Distal to SPr/Prox to Adr
Prox to Spr
(non-immediate field)
Distal to Spr (Hill 1982, Jaggar and Buba 1994)
By contrast languages such as Yucatec show bimorphemic stems in all categories,

with a number of more complex stems among the temporal adverbs (Hanks 1990):
lel-a
tel-a hel-a
bey-a
lel-o
tel-o hel-o
bey-o
wal-ak-hak-a at this time of day in the past
tol-ak-hak-o at that time in the remote past
In Quechua, nominal demonstratives mark a three way contrast between Prox,

Medial, Dist with monomorphemic forms, and the spatial adverbs are derived from
the demonstratives by suffixation of [-pi], as in:
kay
chay
haqay
this
that
that
kay-pi here
chay-pi there
haqay-pi
over there
Similarly, Mandarin demonstratives are monomorphemic, whereas spatial adverbs

are derived by suffixation (Tao 1999, Yang 2006).
zh
z
this, hi accessibility
that, low accessibility
zhr, zhli
nr, nli
here
there
At the high end of compositionality are languages like Nungubuyu, Santali and
Malagasy, where deictic words may be composed of five or more morphemes, each
of which contributes to the meaning of the whole. For Nungubuyu, Heath (1980)
reports that four basic roots distinguish Proximal, Accessible (to Adr), Distal and
Anaphoric in both demonstrative and adverbial categories, with the adverbs
derived by suffixation. These are cross-cut by + Concrete (precise vs. vague),
+ Kinetic (trajectory vs. location), + Centric (where +Centric is subdivided
into Centripetal Centrifugal, and Centric is used for transverse relative to the
Sprs field of vision). The nominal forms are further subdivided by + Absolute
(which Heath glosses definiteness), gender (masc, fem), number (sg, dual, pl) for
human referents, and an additional five noun classes for non-humans. Each of these
distinctions is marked by morphological exponents, resulting in relatively compositional forms.
In Santali all deictics are composed of minimally three morphemes, but potentially more (Zide 1972). There are three bases (proximal, medial, distal) that combine with three regions (immediate, medial, remote) to yield nine different spatial
values. The resulting forms may be further marked as Lateral (for objects at the
edge of the perceptual field, like Nungubuyu non-Centric), Intensive (for referen-
326
William F. Hanks
tial precision, like Nungubuyu Concrete), Visible Invisible (two distinct morphemes), Animate Inanimate, and a three way number distinction. Denny (1982:
371372) reports that in Inuktitut, there are twelve deictic roots, distinguishing
Out of field In-field and within In-field, Away (non-prox) At (prox), and
within Away, Bounded (Exterior Interior) Vertical (InferiorSuperior) Horizontal. The In-field categories are further divided by a binary division between
Restricted (equidimensional, precise) Extended (elongate, mobile, regional).
Further modifications (by inflection or derivation) include grammatical category
(nominal, adverbial, predicative), transposition of the indexical ground from Spr to
Adr or Other, case, number, existence in a place and motion yield over six
hundred and eighty-five deictic words.
In general, the greater the degree of compositionality, the greater the functional
transparency of the composed forms and the more likely it is that some deictic morphemes (root or affix) will recur across subsystems, resulting in a greater degree
of proportionality (although non-compositional languages, like English, can be
highly proportional). Grammatical categories, derivational relations, proportionality of oppositions and compositionality jointly make up the paradigmatic skeleton of deixis in any language and provide a basis for more detailed comparison
of functional structures in different subsystems and across languages. Are there recurrent gaps in the paradigms, markedness asymmetries, or domains in which
functions are hyperdifferentiated? Individual subsystems in any language and
across languages differ widely in the relative delicacy of the distinctions they
mark.
Many linguists have sought to organize this variability in typologies based on
the number of opposed terms given deictic systems display. Frei (1944) first formulated this approach, noting that the most commonly attested systems are two- or
three-term systems, often augmented by an additional term that is neutral as to distance. Anderson and Keenan (1985) build on Freis schema, as does Fillmore
(1997) and more recent typological work done by scholars at the Max-Planck Institute for Psycholinguistics. Among languages with four or more terms, Anderson
and Keenan (1985) cite Malagasy, Quileute and Sre, to which we can add Yucatec
Maya (Hanks 1990), Tlingit (Boas 1917, Swanton 1911), Crow (Graczyk 1986)
and Santali (Zide 1972).
There are several challenges facing this kind of typology. First, how to distinguish between multi-term systems in which each term indicates a degree of spatial
remove vs. ones in which an erstwhile medial is actually neutral or marks a shift in
the indexical ground. The former are treated as distance-centric, the latter as person-centric (Anderson and Keenan 1985). A second challenge is to decide on
which forms belong in the inventory. In highly compositional systems or ones in
which multiple deictics are combined in relatively fixed phrases, the line between
basic deictics, on which the typology is focused, and non-basic ones can be difficult to draw (a difficulty pointed out by Frei 1944: 114, 120). A language that ap-
327
pears two-term from a restrictive definition of the paradigm may in fact systematically mark three or more distinctions at the level of conventionalized phrases.
Inversely, a system with three forms may resolve to two binary distinctions. Hence
the number of formal distinctions need not correspond to the number of functional
distinctions. Finally, the number of terms in one subsystem of a language need not
be identical to the number in another. In Maya for instance, the Manner deictics
show a binary split, plus a neutral term, whereas the Spatial deictics make a fiveway distinction and the nominal deictics a four-way split (Hanks 2005a).16
Diessel (1999: 51) presents a useful summary of what he treats as semantic,
syntactic and pragmatic features of deixis. Under deictic semantics he lists Distance (prox, medial, dist), Visibility (vs. Invisibility), Elevation (up, down relative
to origo), Geography (uphill, downhill, upriver, downriver) and Movement (toward, away from, across the visual field of the Spr). Under quality semantics he
cites Ontology (location, object/person), Animacy, Humanness, Sex, Number and
Boundedness (bound, unbound). The difference is that the deictic features apply to
the relation between the ground or origo and the denotatum, whereas qualitative
features characterize the denotatum itself. Grammatical categories, case marking
and agreement are grouped under Syntax, while Pragmatic features consist of Use
(exophoric, anaphoric, discourse deictic, recognitional) and Reference (emphasis,
contrast, precision). While this way of arraying the features is heuristically useful,
it leaves out features of the indexical ground (origo as speaker, addressee, common
ground or other) and draws an artifical distinction between semantics and pragmatics (since conventionalized indexicality is both semantic, because encoded,
and thoroughly pragmatic, because linked to usage). Furthermore, by splitting
deictic from use features, the typology obscures the fact that spatial location,
vision, endophoric and recognitional features denote alternative modes of access
(perceptual and cognitive) to the referent. For our purposes, it is more useful to
classify features in terms of the three-part relation between the indexical ground,
the relational features and the object denoted (see Figure 1).
4.
The relational values
The relational values of deictics are widely discussed in the literature. They include Relative Immediacy (in space or time), Interiority (inside, outside, lateral),
Location vs. Trajectory, Perception (visual or other), and several varieties of Cognitive Access (anaphora, reference to discourse, relative salience, perspective
[prospective, retrospective] and recognitional). Temporal deixis poses numerous
problems beyond the scope of this essay, including tense, aspect, mode and the
status of quasi-deictic shifters like today, tomorrow, yesterday, ago. Like pronoun
systems, temporal shifters are functionally, and sometimes formally related to
other kinds of deixis, but sufficiently different to merit separate treatment.
328
4.1.
William F. Hanks
Spatial immediacy
Frei (1944) noted that most deictic systems distinguish two or three degrees
of relative proximity between the ground and the object. Examples are English
(here there), Vietnamese (Close vs. Far; Diessel 1999: 36), West Greenlandic
(Prox, Medial, Dist; Fortescue 1984 cited by Diessel 1999: 46) and Turkish (Prox,
Med, Dist; see Diessel 1999: 60). Among three-term systems, Anderson and Keenan (1985: 282285) and Diessel (1999: 50) distinguish two kinds: (1) distanceoriented systems, in which the medial term marks mid-range from Spr and the Distal term marks remote from both Spr and Adr, and (2) person-oriented systems in
which the medial marks Prox or Identifiable to Adr. The relation between deictic
distinctions and participants was addressed by Brugmann (1904), Boas (1911),
Bhler (1934), Frei (1944), Kurylowicz (1972) and Fillmore (1982: 4748). In the
current framework, the so-called person oriented systems illustrate not distinct relational features, but distinct indexical grounds, a point to which we return below.
According to Anderson and Keenan (1985: 286289), numerous languages (including Malagasy, Santali, Sre, Quileute, Tlingit) distinguish four or more degrees
of distance from ground. Both Fillmore (1982: 4849) and Diessel (1999: 4042)
argue that no language distinguishes more than three degrees of distance, and such
expanded systems always involve some other dimensions (such as perception, Inside Outside or a shift in the origo). Debates over the status of the medial term
and hyper-differentiated distance values both illustrate the difficulty of isolating
purely spatial distinctions in the deictic field.
In addition to degrees of remove from the indexical origo, some languages distinguish values on a dimension that can be called interiority, that is, whether the
referent is inside, outside or lateral to the indexical origo. West Greenlandic for instance makes a three-way spatial distinction in the simple demonstratives, but also
has a special deictic root for reference to objects outside the utterance space, as
well as four separate roots that combine distance with verticality: medial.up, medial.down, distal.up, distal.down relative to origo (Fortescue 1984: 259262). Denny
(1982: 361) reports a five-way distinction for the distal locative there in Inuktitut:
up.there, down.there, in.there, out.there, over.there (horizontal). In the spatial
deictics in Yucatec Maya, there is no verticality, but regional forms (as opposed to
punctual ones) distinguish Inside (inclusive of origo) from Outside (exclusive of
origo)17.
In a number of languages, including Nungubuyu, Inuktitut, Santali and Kiowa,
there are specialized deictics for reference to objects in motion, for which the origo
serves as source space for centrifugual motion, the goal space for centripedal, or
the ground for lateral motion (transverse or along the edge). Heath (1980: 152) describes the Nungubuyu forms as Kinetic, and shows the three way distinction between [-ala-] approaching, [-ali-] going away and [-waj-] across. In Inuktitut,
deictics may combine with a motion infix [-qa-] plus case marking to distinguish
329
source, goal and path readings (Denny 1982: 373). For Crow, Graczyk (1986) cites
the postposition [-ss(ee)] which marks centripedal trajectory. Zides (1972:
267268) description of Santali deixis includes the infix [-Vh-], which marks that
the referent is off to the side (lateral) of the origo space. Diessel (1999: 46) cites a
similar distinction in Kiowa between centrifugal and centripedal deixis. In other
languages, such as English and many Mayan languages, directionality is encoded
in the verb complex, either lexically or by way of affixes (see Fillmore 1975 on
English come, go). Dixon (1972: 48) reports an uphill, downhill distinction in
Dyirbal deictics, and Matisoff (1973: 110111) sites a similar one in Lahu.
4.2.
Perceptual access to the object
In theory, perceptual access to the referent may play a role in any exophoric deixis,
and vision is a key source in spatial deixis. While noting the importance of visual
access in some deixis, Fillmore (1982: 51) adopted a restrictive definition in which
such perceptual conditioning is not deictic per se. By contrast, Bhler (1934) starts
from the premise that sound is critical, because to identify the referent of here one
need only follow the speaking voice to its source. In some languages, such features
are conventionalized: certain deictics or bound morphemes indicate the presence
or absence of perceptual access to the object. Despite the variation in perceptual
coding, it appears that in all languages, if perceptual access is encoded, it is access
for the Spr (and not the Adr or some Other) that is in play (Diessel 1999: 4142). In
other words, no language specifically marks that the Object is perceptually available to the Adr, even though there are many cases in which spatial contiguity is
computed relative to the Adr (see next section). Diessel cites seven languages in
his sample in which Visibility is overtly marked, and there are many more cases.
In some languages, both Visible and Invisible are coded in separate roots or
stems. For example, Quileute shows a root distinction between Visible and Invisible, where Visible is further subdivided into Prox (Spr), Prox (Adr), Medial (Spr +
Adr) and Distant (Andrade 1933). Chinook makes a basic distinction between
Visible (also used for Present) and Invisible (also used for Past) (Boas 1911:
617623). Kwakwala shows a similar distinction (Anderson and Keenan 1985:
290). Crow (Graczyk 1986) shows a slight variant in the four-way distinction between Immediate, Medial, Distal (+ Vis), Distal (-Vis). These oppositions between
presence and absence of visual access are all equipollent in that both terms are
overtly marked.
By contrast, many languages show privative oppositions, in which one term is
marked and the other unmarked. Santali appears to show a three-way split between
stems unmarked for perception, ones marked for lateral visual and ones explicitly marked for unseen objects (Garvey et.al. 1964: 115, Zide 1972: 267). West
Greenlandic has a special root [-im] for Invisible objects (Fortescue 1984: 259
262, Diessel 1999: 46), as do the Malagasy locative deictics (Anderson and Kee-
330
William F. Hanks
nan 1985: 293). Diessel (1999: 42) cites Ute to illustrate what he considers the
most common alignment: Prox Distal Invisible. It appears that in any language
with a privative opposition on the parameter of vision, it is the Invisible form that
gets special marking. No language has a special form marking Visible unless it also
has one for Invisible. The implication is that visual access is the normal state of affairs, and the use of simple proximal or distal forms in a language with an overt Invisible, pragmatically implies that the object is within sight. This is consistent with
the fact that vision plays a central role in locating objects in space.
The close association between visual and spatial identification is reflected in
the widespread tendency among linguists to overlook the role of other senses in
identifying objects of deictic reference. With the exception of Bhler (1934) and
Hanks (1983, 1990, 1993a, 2005a) none of the works cited in this chapter describes
auditory, olfactory, tactile or gustatory access of deictic referents. While such distinctions are evidently rarely encoded in deictics, they frequently part of the phenomenal conditions in which deictic acts occur. But in Yucaec Maya presentative
deictics, there is a three way division between Tactual (for objects within touch of
the Spr), Visual (objects visible to both participants) vs. Peripheral sensory (objects perceived by hearing, smell or taste) (Hanks 1993a, 2007b).18
4.3.
Cognitive access
Deictics in any language may individuate objects that are cognitively accessible,
but neither perceivable nor locatable in the current utterance space. This includes
for instance, cognitive salience, as in the Hausa distinction between High and Low
focus deictics (Anderson and Keenan 1985: 289, cf. Jaggar and Buba 1994). It also
includes oppositions between Prospective (here we go, Im going this way) and
Retrospective (so thats my story [resumptive conclusion of narrative], I came
that way). Prospect and retrospect are closely related to future and past respectively, and when deployed in Discourse, they merge with cataphora and anaphora.
In many languages, such as English, non-proximal forms are used for anaphora (recall) and proximal ones for cataphora (anticipation), but in some, there are special
anaphoric deictics.
Heath (1980) cites a special Anaphoric base in Nungubuyu and Hanks (1990)
shows that Yucatec spatial deictics have a form unmarked for distance or interiority, but used for reference to places known (either anaphoric or recognitional). The so called recognitional usage, identified by Schegloff (1972) and cited
by Diessel (1999), denotes usages of names, descriptions and deictics in which the
Spr assumes that the Adr will be able to identify the referent without further specification. Essentially, this is a species of cognitive access based on common
ground. Hanks (1990) and Diessel (1999: 103, 113) distinguish anaphora proper
(coreference with a prior expression in the discourse) from reference to prior discourse (as in English Thats an outrageous claim, said in response position). In
331
any spatialist account, these uses are treated either as secondary, non-deictic uses
(as in Diessel 1999), or as metaphorical extensions from space (as in Anderson and
Keenan 1985). In an interaction-based approach in which the core function of deixis is to signal participant access and direct the Adrs attention to referents, they
demonstrate that cognitive access functions alongside spatial and perceptual access. Terminological differences aside, any language in which indexical reference
were limited to the exophoric field of utterance would be too constrained to meet
the demands of ordinary talk. This is obvious in any language with dedicated forms
for cognitive access, but is a factor of the pragmatics of deixis in any language.
4.4.
Features of the indexical origo
Among the highly significant contributions made by Bhler (1934) was his discussion of what we have called deictic transpositions. These are uses in which the
origo of deictic reference is shifted so that forms normally understood in relation to
the Spr are grounded on the Adr, some other object, or a narrative space. An easy
example is provided by talk to young children, in which a caregiver speaks for the
child e.g. a mother saying to a grimacing infant Oh Mommy, I dont like that. As
in quoted speech, the transposed deictics are grounded in the origo of another, in
this case the addressee. A narrator can project herself into a narrative space, as
when one retells an event from the past, saying So here I am walking down the
street and this guy comes up to me and asks me my name. The use of the proximal
deictic and the verb come are grounded in the narrative, not in the current situation of speech. A related transposition takes place whenever a distant place or object is imaginatively projected into the current indexical ground of utterance by
using unexpected proximal forms: So this is Paris, this is Berlin and this is Strasbourg (all on a table-top). Similarly, even very small utterance spaces can be
treated as if they were large, thus licensing use of normally remote forms for objects actually within reach: youve left your queen way over there (in reference to
a chess piece on a board between the interlocutors). What these examples have in
common is that deictic usage in effect manipulates the indexical ground of utterance. In the examples cited so far, these are purely pragmatic effects with little or
no grammatical conventionalization.
By contrast, many languages have special deictic forms whose function is to
signal shifts in the indexical ground. Anaphora and recognitional uses can be seen
in this light if we consider that they function only when both Spr and Adr share
cognitive access to the object. Thus anaphoric forms are symmetric in that they
presuppose that both participants can recover the antecedent. From a pragmatic
perspective, symmetry of access is the prototypical condition for deixis. In the
face-to-face, what is close to one party is also close to the other, and their perceptual fields are convergent (overlapping or reciprocal). Key aspects of cognitive access either rest on common ground or are co-constructed. By contrast, actual ego-
332
William F. Hanks
centric or altercentric uses are asymmetric because the origo attaches selectively to
one party but not the other. Whenever participants are spatially, perceptually or
cognitively separated, the ground is de facto asymmetric. Languages have various
ways of encoding different types of origo, as is clear in the so-called person-centered systems.
According to Frei (1944), Bisaya marks a six-way distinction between Distal
and Proximal relative to three distinct origos: Spr, Adr and Spr+Adr. Anderson and
Keenan (1985:287) cite demonstratives in Bemba, which mark a five-way distinction: Immediate (Spr), more Immediate to Spr than to Adr, Immediate (Spr + Adr),
Immediate (Adr) and Distal (Spr + Adr). These two languages appear to be maximal in that they distinguish both Proximal and Distal relative to all three origos.
Bemba is evidently unique in conventionally distinguishing degrees of relative
symmetry of access to the object. Chinook (Boas 1911) and Kwakwala (Anderson
and Keenan 1985: 290) are similar in that they lack a distal category, but mark
Proximal relative to three distinct grounds: Spr, Adr and Other. Japanese (Diessel
1999: 59), Ewondo (Diessel 1999: 40), Spe (Anderson and Keenan 1985: 287) and
Quileute (Andrade 1933) are nearly identical in distinguishing Prox (Spr) vs. Prox
(Adr) vs Medial or Distal (Spr+Adr), and Diessel cites several other languages
with a similar paradigm. Inuktitut (Denny 1982: 362) illustrates yet another solution, with a special affix marking transpositions from the Speaker-ground to
Other (including Adr or third).
There are several reasons that it is challenging to actually demonstrate these different values for the indexical ground. Detailed observation of ordinary usage is
requisite, and most of the published literature simply fails to provide the necessary
evidence. In any language lacking dedicated symmetric forms (for which the common ground is the origo), one must infer or guess at how the attested forms are used
when access to the object is roughly the same for both participants. In languages
with a three-way split among SprAdrOther, either the Other category covers
common access, or the use of either egocentric or altercentric forms pragmatically
implies common ground. Similarly, in languages like Japanese, the question is how
one denotes an object proximal to both parties, since shared access is apparently
signalled in the distal form only. Heath (1980) does not mention the ground for
either Proximal or Distal deictics in Nungubuyu, whereas separate forms mark Prox
(Adr) and Anaphoric (which implies shared cognitive access). Presumably the Prox
and Distal categories may be either egocentric (Spr) or symmetric (Spr+Adr).
If we take the published descriptions at face value, the following generalizations emerge: (1) No language has a distinctive category for the Spr-origo, unless
it also has a form for Adr-origo. In the absence of an Adr-origo form, the attested
forms are simply unmarked and likely to be used for egocentric, altercentric and
shared ground. Adr-centered deixis is more pragmatically marked than Spr-centeric deixis, and more likely to be overtly signalled. (2) Medial and Distal categories tend strongly to relate to symmetric origo (Spr+Adr). Even if Speaker-cen-
333
tric and Addressee-centric forms are distinguished in the Proximal deictics, Distal
forms are likely to be symmetric (Spr+Adr). (3) In languages lacking a dedicated
Anaphoric form, the non-Proximal forms are most likely to subserve this function
(because they share the indexical symmetry of Distal categories). (4) Immediate
forms mark higher salience than their non-Immediate counterparts, and are more
likely to be used for new information, high-focus and prospective cognitive access.
There are many more subdivisions among deictics in the published literature,
and long-term fieldwork in any language demonstrates the complexity of decitic
practices, in which all of the foregoing factors may be simultaneously in play. In
numerous languages for instance, there are special forms to distinguish precise or
punctate from vague or regional reference. Compare Punctual in Yucatec Maya
(Hanks 1990), Concrete in Nungubuyu (Heath 1980), Precise in Tsimshian (Boas
1911, Dunn 1979), Intensive in Santali (Zide 1972) and Restriced in Inuktitut
(Denny 1982: 372). Insofar as these categories serve to mark the relative precision
with which the object is individuated, as in this very one vs. this one, they nicely
display one of the functional requisites of deictic reference, that is, the resolution
of reference. Such resolution relies on the co-operation of spatial, perceptual and
cognitive access, along with the range of indexical origos to which they are anchored.
Language key to items in bibliography

N ATIVE N ORTH A MERICA
Tsimshian
Tlingit
Chinook
Wasco-Wishram
Bella Coola
Crow
Mohawk
Oneida
Inuktitut
Aleut
Quileute
Greenlandic
M ESOAMERICA
Tlapanec
Yucatec
Toba
Quechua
AND
A RCTIC
Boas 1911b, Dunn 1979
Boas 1917, Swanton 1911
Boas 1911a
Silverstein 1978
Davis and Saunders 1975, 1976
Graczyk 1986
Bonvillain 1981
Abbott 1981
Denny 1978, 1982
Bergsland 1951
Andrade (1933)
Fortescue 1984
AND
L ATIN A MERICA
Wichmann 1993
Hanks 19832009 (various)
Manelis Klein 1981
Calvo Perez 1999
334
William F. Hanks
I NDO -E UROPEAN
French
Germanic
Russian
A FRICA
Hausa
Swahili
Malagasy
Morel and Danon-Boileau, eds. 1992

Brugmann 1904, Bhler 1934
Friedrich 1979
Hill 1982, Jaggar and Buba 1994

Leonard 1982, 1985
Cousins 1894, Garvey 1964, Dez 1980, Anderson
and Keenan 1985, Hanks 1986
A SIA (S OUTH , S OUTHEAST ), P APUA N EW G UINEA

Mandarin
Tao 1999
Japanese
Kuno 1973
Santali (Munda)
Zide 1972
Belhare (Sino-Tibet) Bickel 2000
Lao
Enfield 2003, n.d.
Lahu
Matisoff (1973)
Jahai
Burenhult 2003, n.d.
Yl Dnye
Levinson n.d. (to appear)
A USTRALIA
Nunggubuyu
Ngandi
Mara
Guugu Yimithirr
Dyirbal
Wardaman
Kayardild
Heath 1980
Heath 1978
Heath 1981
Haviland 1993
Dixon 1972
Merlan 1994
Evans 1995
Notes
1. Cf. Fillmore 1971, Silverstein 1976, Hanks 1984, 1990, 2005a, Diessel 1999, Enfield
2003.
2. Cf. Morris 1946, Peirce 1955, Eco et al. 1988, Husserl 1978, Benveniste 1974, cf. Searle
1969.
3. Cf. Jespersen 1965 [1924], Jakobson 1971 [1957], Silverstein 1976.
4. Cf. Hymes 1974, Gumperz and Hymes 1964, Sherzer 1983, 1991.
5. Cf. Hanks 1990, 2005a, Himmelmann 1996, Janssen 2000, 2002, Leonard 1985, Mondana 2005.
6. See Garfinkel 1967, Schegloff 1972, Enfield 2009, Sidnell 2010.
7. See Clark and Marshall 1981, Clark, Schreuder and Buttrick 1983, Enfield 2006.
335
8. Cf. Frei 1944, Anderson and Keenan 1985, Diessel 1999: 158, Lyons 1977, Senft and
Smits 2000: 6580, Levinson 2003.
9. See Sacks 1992, Goodwin 1981, Schegloff 1984, Heritage 1984, Sidnell 1997, 2005.
10. Cf. Haviland 1993, 2001, 2003, Goodwin 1994, 2000, Kita 2003, Enfield 2009.
11. Cf. Sacks, Schegloff and Jefferson 1974, Schegloff 1987, 1992, Sidnell 2005, 2010,
Stivers 2007.
12. Cf. Schutz 1970, Goodwin 1994, Goodwin and Heritage 1990, Sacks 1992, Schegloff
1972, 1992.
13. See Haviland 1996, 2003, Goodwin 1981, 2003, Sidnell 1997, 2005 and Ochs, Schegloff and Thompson 1996.
14. See Diessel 2005 for similar generalizations over a larger sample of languages.
15. Compare Russell 1940, Reichenbach 1947, Frei 1944, Kurylowicz 1972, Lyons 1982,
Fillmore 1997.
16. For recent discussions of deictic typology, see Himmelmann 1996, Fillmore 1997 and
Diessel 1999.
17. Cf. Hanks 1990, where I note that Maya and Inuktitut illustrate interiority as well as
does Greenlandic; Santali indicates lateral to origo, which I suggest is related to interiority. There is something like a family of effects around interiority, inclusiveness,
edge-ness (at boundary of origo). My use of interiority, perhaps too vague, is meant to
cover this range.
18. From the viewpoint of ordinary usage and native metalinguistic commentaries, it is
clear that Yucatec speakers distinguish a three-way split between Tactual, Visual and
Peripheral Sensory access, although the comparative evidence suggests that the Peripheral Sensory category may be better analyzed as [+ Perceived, +Invisible], similar to the
systems in other languages. In any event it is clear that vision is the most commonly
marked modality and it is the absence of visual access that usually gets special marking.
References
Abbott, Clifford
1981
Here and there in Oneida. International Journal of American Linguistics 47(1):
5057.
Anderson, Stephen and Edward Keenan
1985
Deixis. In: Timothy Shopen (ed.), Language Typology and Syntactic Description, vol. 3: Grammatical Categories and the Lexicon, 259308. Cambridge:
Andrade, Manuel J.
1933
Quileute. Handbook of American Indian Languages 3, 151292. New York:
Columbia University Press.
Bar Hillel, Yehoshua
1954
Indexical expressions. Mind LXIII: 359379.
Benveniste, Emile
1965
Language and Human Experience. Diogenes 51: 112.
Benveniste, Emile
1966
De la subjectivit dans le langage. In: Problmes de Linguistique Gnrale,
vol. 1, 258266. Paris: Gallimard.
336
William F. Hanks
Benveniste, Emile
1974
Le langage et lexprience humaine. In: Problmes de Linguistique Gnrale,
Benveniste, Emile
1974
Problmes de Linguistique Gnrale, vol. 2. Paris: Gallimard.
Bergsland, Knut
1951
Kleinschmidt Centennial IV: Aleut demonstratives and the Aleut-Eskimo relationship. International Journal of American Linguistics 17(3): 167179.
Bickel, Balthazar
2000
Deictic transposition and referential practice in Belhare. Journal of Linguistic
Anthropology 10(2): 224247.
Bloomfield, Leonard
1933
Language. New York: Holt.
Boas, Franz
1911a
Chinook. In: Handbook of American Indian Languages. Smithsonian Institution, Bureau of American Ethnology, Bulletin 40, ed. by Franz Boas. 2 vols, I:
559678. Washington: Government Printing Office.
Boas, Franz
1911b
Tsimshian. In: Handbook of American Indian Languages. Smithsonian Institution, Bureau of American Ethnology, Bulletin 40, ed. by Franz Boas. 2 vols. I:
Boas, Franz
1917
Grammatical Notes on the Language of the Tlingit Indians. Vol. VIII, No. 1.
Philadelphia: Anthropological Publications of the University Museum.
Bonvillain, Nancy
1981
Locative semantics in Mohawk: Time and space. International Journal of
American Linguistics 47(1): 5865.
Brugmann, Karl
1904
Demonstrativpronomina in den indogermanischen Sprachen. Leipzig: Teubner.
Bhler, Karl
1934
Sprachtheorie. Jena: Springer.
Burenhult, Niclas
2003
Attention, accessibility and the addressee: The case of the Jahai demonstrative
ton. Pragmatics 13(3): 363379.
Burenhult, Niclas
To appear Distance undone: new light on the Jahai multi-term demonstrative system.
In: Michael Dunn and Srgio Meira (eds.). Demonstratives in cross-linguistic
perspective.
Calvo Perez, Julio
1999
Pragmatica y gramatica del Quechua Cuzqueo. Cuzco: Centro de Estudios
Regionales Andinos Bartolome de las Casas.
Chapman, Mackenzie
2009
A brief introduction to demonstratives in Mandarin Chinese. (Unpublished
paper, Summer Institute of Linguistics of the Linguistic Society of America,
7pp).
337
Cicourel, Aaron
1964
Method and Measurement in Sociology. New York: Free Press.
Cicourel, Aaron
1974
Cognitive Sociology: Language and meaning in social interaction. New York:
Free Press.
Cicourel, Aaron
1985
Text and discourse. Annual Reviews of Anthropology 14:159185.
Cicourel, Aaron
1992
The interpenetration of Communicative contexts. In: Alessandro Duranti and
Charles Goodwin (eds.), Rethinking Context: Language as an interactive phenomenon, 291310. Cambridge: Cambridge University Press.
Clark, Herbert H. and Catherine Marshall
1981
Definite reference and mutual knowledge. In: Aravind K. Joshi, Bonnie L.
Webber and Ivan A. Sag (eds.), Elements of Discourse Understanding, 1063.
Clark, Herbert H., Robert Schreuder and Samuel Buttrick
1983
Common ground and the understanding of demonstrative reference. Journal of
Verbal Learning and Verbal Behavior 22: 245258.
Cousins, William E.
1894
A Concise Introduction to the Study of the Malagasy Language as Spoken in
Imerina. 3rd ed. Antananarivo: The Press of I.M.S.
Davis, Philip W. and Ross Saunders
1975
Bella Coola Nominal Deixis. Language 51(4): 845858.
1976
Bella Coola Deictic Roots. International Journal of American Linguistics
42(4): 319330.
Denny, J. Peter
1978
Locating the universals in lexical systems for spatial deixis. In: Donka Farkas,
Wesley M. Jacobsen, and Koral W. Todrys (eds.), Papers from the Parasession
on the Lexicon, 7184. Chicago: Chicago Linguistic Society.
Denny, J. Peter
1982
Semantics of the Inuktitut (Eskimo) spatial deictics. International Journal of
Dez, Jacques
1980
Structures de la Language Malgache, lments de grammaire lusage des
francophones. Paris: Publications Orientalistes de France.
Diessel, Holgar
1999
Demonstratives: Form, function and grammaticalization. Amsterdam: Benjamins.
Diessel, Holger
2005
Distance contrasts in demonstratives. In: Martin Haspelmath, Matthew Dryer,
David Gil, and Bernard Comrie (eds.), World Atlas of Language Structures,
Dixon, Robert M.W.
1972
The Dyirbal Language of Northern Queensland. Cambridge: Cambridge University Press.
338
William F. Hanks
Dixon, Robert M.W.

1980
The Languages of Australia. Cambridge: Cambridge University Press.
Dunn, John A.
1979
A Reference Grammar for the Coast Tsimshian Language. Canadian Ethnology Service, Paper 55. Ottawa: National Museum of Man, Mercury Series.
Eco, Umberto, Marco Santambrogio and Patrizia Violi
1988
Meaning and mental representations. Bloomington, IN: Indiana University
Press.
Enfield, Nicholas J.
2003
Demonstratives in space and interaction: Data from Lao speakers and Implications for semantic analysis. Language 79(1): 82117.
2006
Social consequences of common ground. In: Nicholas Enfield and Stephen Levinson (eds.), Roots of Human Sociality, Culture, Cognition and Interaction,
399430. Oxford: Berg.
2009
The Anatomy of Meaning, speech, gesture and composite utterances. Cambridge: Cambridge University Press.
Enfield, Nicholas J. and Tanya Stivers (eds.)
2007
Person Reference in Interaction, Linguistic, Cultural and Social Perspectives.
Evans, Gareth
1982
The Varieties of Reference. New York: Oxford University Press.
Evans, Nicholas
1995
A grammar of Kayardild. Berlin: Mouton de Gruyter.
Evans, Nicholas
2003
Bininj Gun-wok: a pan-dialectal grammar of Mayali, Kunwinjku and Kune.
2 vols., xxix + 746. Canberra: Pacific Linguistics.
1971
Santa Cruz Lectures on Deixis. Bloomington, IN: Indiana University Linguistics Club.
Fillmore, Charles
1982
Towards a linguistic framework for spatial deixis. In: Robert Jarvella and Wolfgang Klein (eds.), Speech, Place and Action, 3159. Chichester: John Wiley.
Fillmore, Charles
1997
Lectures on Deixis. Stanford: CSLI publications.
Fortescue, Michael
1984
West Greenlandic. London: Croom Helm.
Frei, Henri
1944
Systmes de dictiques. Acta Linguistica 4: 111129.
Friedrich, Paul
1979
Language, Context and the Imagination. Stanford: Stanford University Press.
Gale, Richard M.
1968
Is it now now? Mind 73: 97105.
Gale, Richard M.
1968
Indexical expressions, egocentric particulars and token-reflexive words. In:
P. Edwards (ed.), Encyclopedia of Philosophy. Vol IV, 151155. New York:
Collier Macmillan.
339
Garfinkel, Harold
1967
Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice-Hall.
Garfinkel, Harold
1972
Studies in the routine grounds of everyday activities. In: David Sudnow (ed.),
Studies in Social Interaction, 130. New York: Free Press.
Garvey, Catherine
1964
Malagasy Introductory Course. Washington DC: Center for Applied Linguistics.
Goodwin, Charles
1981
Conversational Organization: Interaction between speakers and hearers. New
York: Academic Press.
Goodwin, Charles
1994
Professional Vision. American Anthropologist 96(3): 606633.
Goodwin, Charles
2000
Practices of Seeing, Visual Analysis: An Ethnomethodological Approach. In:
Theo van Leeuwen and Carey Jewitt (eds.), Handbook of Visual Analysis,
Goodwin, Charles
2003
Pointing as situated practice. In: S. Kita (ed.), Pointing: Where language, culture and cognition meet, 217242. Hillsdale, NJ: Erlbaum.
Goodwin, Charles
2006
Human sociality as mutual orientation in a rich interactive environment: Multimodal utterances and pointing in aphasia. In: Nicholas Enfield and Stephen Levinson (eds.), Roots of Human Sociality, Culture, Cognition and Interaction,
Goodwin, Charles and John Heritage
1990
Conversation Analysis. Annual Review of Anthropology 19: 283307.
Graczyk, Randolph
1986
Crow deixis. Unpublished Ms., Department of Linguistics, University of Chicago.
Green, Keith (ed.)
1995
New Essays in Deixis. Amsterdam: Rodopi.
Grice, H. Paul
1967
Logic and conversation. William James Lectures, Cambridge, MA: Harvard
University.
Grice, H. Paul
1978
Further notes on logic and conversation. In: Peter Cole (ed.), Syntax and
Semantics 9: Pragmatics, 113128. New York: Academic Press.
Gumperz, John
1982
1964
The Ethnography of Communication. American Anthropologist, Special Publication 66, Number 6, Part 2.
1976
Hanks, William F.
1983
Deixis and the Organization of Interactive Context. Unpublished PhD thesis.
University of Chicago, Department of Linguistics, Department of Anthropology.
340
William F. Hanks
Hanks, William F.
1984
The evidential core of deixis in Yucatec Maya. In: Papers from the Fifteenth
Regional Meeting of the Chicago Linguistic Society, 154173. Chicago: Chicago Linguistic Society.
Hanks, William F.
1986
Markedness and category interactions in the Malagasy deictic system. University of Chicago, Department of Linguistics Working Papers. 20pp.
Hanks, William F.
1990
Referential Practice: Language and lived space among the Maya. Chicago:
Hanks, William F.
1992
The indexical ground of deictic reference. In: Alessandro Duranti and Charles
Goodwin (eds.), Rethinking Context: Language as an interactive phenomenon,
4377. Cambridge: Cambridge University Press. [Reprinted from Papers from
the Parasession on Language in Context. Chicago Linguistic Society, 1989].
Hanks, William F.
1993a
Metalanguage and pragmatics of deixis. In: John Lucy (ed.), Reflexive language: Reported speech and metapragmatics, 127157. Cambridge: Cambridge University Press.
Hanks, William F.
1993b
Notes on semantics in linguistic practice. In: Craig Calhoun and Moise Postone (eds.), Towards a reflexive sociology: The social theory of Pierre Bourdieu, 139155. Oxford: Basil Blackwell.
Hanks, William F.
1995
Language and Communicative Practices. Critical Essays in Anthropology.
Edited by Maurice Bloch, Pierre Bourdieu, and John L. Comaroff, Boulder:
Westview Press.
Hanks, William F.
1996
Language form and communicative practices. In: John Gumperz and Stephen
Levinson (eds.), Rethinking Linguistic Relativity, 232270. Cambridge: Cambridge University Press.
Hanks, William F.
1999
Intertexts: Writings on Language, Utterance and Context. Denver: Rowman
and Littlefield.
Hanks, William F.
2001
Indexicality. In: Alessandro Duranti (ed.), Key terms in language and culture,
119121. Oxford: Blackwell.
Hanks, William F.
2005a. Explorations in the deictic field. Current Anthropology 46(2): 191220.
Hanks, William F.
2005b
Pierre Bourdieu and the Practices of Language. Annual Review of Anthropology 34: 6783.
Hanks, William F.
2006a
Context. In: Encyclopedia of Language and Linguistics, 2nd edition. Ed. by
Keith Brown. Oxford: Elsevier.
Hanks, William F.
2006b
Deixis. In: Encyclopedia of Literature. London: Routledge.
341
Hanks, William F.
2006c
Spatial frames of reference in language and thought. Review article of Space in
Language and Cognition, by Stephen Levinson. Language in Society 35(2):
285296.
Hanks, William F.
2007
Construal of persons in Maya conversation. In: Tanya Stivers and Nicholas Enfield (eds.), Person Reference in Interaction: Linguistic, cultural and social
perspectives, 149171. Cambridge: Cambridge University Press.
Hanks, William F.
2007b
The evidential core of deixis in Yucatec Maya. In: Zlatka Guerncheva (ed.),
Lnonciation mdiatise. Tome II, 311334. Paris-Louvain: Peeters.
Hanks, William F.
2008
Lngua como prtica social: das relaes entre lngua, cultura e sociedade a
partir de Bourdieu e Bakhtin. Cortez Editora: So Paulo, Brasil.
Hanks, William F.
2009
Fieldwork on Deixis. Journal of Pragmatics 41: 1024. Special issue on
Emancipatory Pragmatics. Ed. by William Hanks, Sachiko Ide, Yasuhiro Katagiri.
Haviland, John B.
1993
Anchoring, iconicity and orientation in Gugu Yimithirr pointing gestures.
Journal of Linguistic Anthropology 3(1): 345.
Haviland, John B.
1996
Projections, transpositions and relativity. In: John Gumperz and Stephen Levinson (eds.), Rethinking Linguistic Relativity, 271323. Cambridge: Cambridge University Press.
Haviland, John B.
2001
Gesture. In: Alessandro Duranti (ed.), Key Terms in Language and Culture,
8386. Malden, MA: Blackwell.
Haviland, John B.
2003
How to point in Zinacantn. In: Sotaro Kita (ed.), Pointing: Where language,
culture and cognition meet, 139170. Mahwah, NJ: Erlbaum.
Heath, Jeffrey
1978
Ngandi Grammar, Texts and Dictionary. Australia Institute of Aboriginal
Studies, Canberra. Atlantic Highlands, NJ: Humanities Press.
Heath, Jeffrey
1980
Nunggubuyu deixis, anaphora and culture. In: Jody Kreiman and Almerindo
Ojeda (eds.), Papers from the Parasession on Pronouns and Anaphora,
151165. Chicago: Chicago Linguistic Society.
Heath, Jeffrey
1981
Basic Materials in Mara: Grammar, Texts and Dictionary. (Pacific Linguistics,
Series C, No. 60.) Department of Linguistics, Research School of Pacific
Studies. The Australian National University.
Heath, Jeffrey
2004
Person. In: Geert Booij, Christian Lehmann, Joachim Mugdan and Stavros
Skopeteas (eds.), Morphologie: Ein internationales Handbuch zur Flexion und
Wortbildung / Morphology: An International Handbook on Inflection and
Word-Formation. Vol. 2, 9981015. Berlin: de Gruyter.
342
William F. Hanks
Heritage, John
1984
A change-of-state token and aspects of its sequential placement. In: Maxwell
Atkinson and John Heritage (eds.), Structures of social action: Studies in conversation analysis, 225246. Cambridge: Cambridge University Press.
Hill, Clifford
1982
Up/down, front/back, left/right: A contrastive study of Hausa and English. In:
Jrgen Weissenborn and Wolfgang Klein (eds.), Here and There: Cross-linguistic Studies on Deixis and Demonstration, 1342. Amsterdam: Benjamins.
Himmelmann, Nicholas P.
1996
Demonstratives in narrative discourse: A taxonomy of universal uses. In: Barbara Fox (ed.), Studies in Anaphora. 205254. Amsterdam: Benjamins.
Husserl, Edmund
1978
The origin of geometry. In: Thomas Luckmann (ed.), Phenomenology and
Sociology: Selected Readings, 4270. New York: Penguin.
Hymes, Dell
1974
University of Pennsylvania Press.
Jaggar, Phillip and Malami Buba
1994
The space and time adverbials NAN/CAN in Hausa: cracking the deictic code.
Language Sciences 16: 387421.
Jakobson, Roman
1960
Closing statement on linguistics and poetics. In: Thomas Sebeok, (ed.), Style in
Language, 350377. Cambridge, MA: MIT Press.
Jakobson, Roman
1971 [1957] Shifters, verbal categories and the Russian verb. In: Roman Jakobson: Selected Writings. Vol. 2: Word and Language, 130147. The Hague: Mouton.
Janssen, Theodorus
2000
Deixis and reference. In: Geert Booij, Christian Lehmann, Joachim Mugdan
and Stavros Skopeteas (eds.), Morphologie: Ein internationales Handbuch zur
Flexion und Wortbildung / Morphology: An International Handbook on Inflection and Word-Formation, 983998. Berlin: de Gruyter.
Janssen, Theodorus
2002
Deictic principles of pronominal, demonstrative and tense deictics. In: Frank
Brisard (ed.), Grounding: The epistemic footing of deixis and reference,
Jespersen, Otto
1965 [1924] The Philosophy of Grammar. New York: WW Norton.
Jungbluth, Konstanze
2005
Pragmatik der Demostrativpronomina in den iberoromanischen Sprachen.
Tbingen: Niemeyer.
Kendon, Adam
1992
The negotiation of context in face-to face interaction. In: Alessandro Duranti
and Charles Goodwin (eds.), Rethinking context, 323334. Cambridge: Cambridge University Press.
Kita, Sotaro (ed.)
2003
Pointing: Where language, culture, and cognition meet. Mahwah, NJ: Erlbaum.
343
Kita, Sotaro and James Essegbey

2001
Pointing left in Ghana: How a taboo on the use of the left hand influences gestural practice. Gesture 1(1): 7395.
Kuno, Susumo
1973
The Structure of the Japanese Language. Cambridge, MA: MIT Press.
Kurylowicz, Jerzy
1972
Universaux linguistiques. In: Luigi Heilman (ed.), Proceedings of the Eleventh
Congress of Linguists, 4773. Bologna.
Lasater, Ann
2009
Notes on Quechua deixis. (Unpublished Ms., Summer Institute of Linguistics
of the Linguistic Society of America, Berkeley, 9 pp.)
Leonard, Robert
1982
The semantic system of deixis in Swahili. PhD thesis, Columbia University.
Leonard, Robert
1985
Swahili demonstratives: Evaluating the validity of competing semantic hypotheses. Studies in African Linguistics 16(3): 281293.
Levinson, Stephen
1983
Levinson, Stephen
2003
Space in language and cognition: Explorations in cognitive diversity. Cambridge: Cambridge University Press.
Levinson, Stephen
To appear Demonstratives in Yl Dnye, the language of Rossel Island Papua New
Guinea. In: Michael Dunn and Srgio Meira (eds.), Demonstratives in crosslinguistic perspective. 29 pp.
Levinson, Stephen et al.
n.d
Demonstratives: Towards a semantic typology. MPI Nijmegen working paper
(Table, 1 page).
Linell, Per
2001
Approaching Dialogue, talk, interaction and contexts in dialogical perspectives. Philadelphia: Benjamins.
Lyons, John
1977
Semantics. Cambridge: Cambridge University Press.
Lyons, John
1982
Deixis and subjectivity: Loquor ergo sum? In: Robert Jarvella and Wolfgang
Klein (eds.), Speech, Place and Action: Studies in Deixis and Related Topics,
101124. New York: Wiley.
Manelis Klein, Harriet E.
1981
Location and Direction in Toba: Verbal morphology. International Journal of
Martinet, Andr
1960
Elements de Linguistique Gnrale. Paris: A. Colin.
Matisoff, James
1973
The Grammar of Lahu. Berkeley: University of California Press.
Meillet, Antoine
1926
Linguistique historique et Linguistique Gnrale. Paris: Champion.
344
William F. Hanks
Merlan, Francesca
1994
A Grammar of Wardaman: A Language of the Northern Territory of Australia.
Berlin: de Gruyter.
1967 [1945] The Phenomenology of Perception. Paris: Gallimard.
Mondana, Lorenza
2005
La constitution de lorigo dictique comme travail interactionnel des participants: une approche praxologique de la rfrence spatiale. Intellectica 4142:
75100.
Morel, Mary-Annick and Laurent Danon-Boileau (eds.)
1992
La Dixis, Colloque en Sorbonne (89 juin 1990). Paris: Presses Universitaires
de France.
Morris, Charles
1946
Signs, language, and behavior. New York: George Braziller.
Nunberg, Geoffrey
1993
Indexicality and deixis. Linguistics and Philosophy 16(1): 143.
Ochs, Elinor, Emanuel Schegloff and Sandra Thompson (eds.)
1996
Interaction and Grammar. Cambridge: Cambridge University Press.
Ozyrek, Asli and Kita Sotaro
n.d.
Joint attention and distance in the semantics of Turkish and Japanese demonstrative systems. (Ms)
Peirce, Charles S.
1955
Philosophical writings of Peirce. Ed. by Justus Buchler. New York: Dover
Publications.
Reichenbach, Hans
1947
Elements of Symbolic Logic. New York: Macmillan.
Rommetveit, Ragnar
1974
On Message Structure. London: Wiley.
Russell, Bertrand
1940
An Inquiry into Meaning and Truth. London: Allen and Unwin.
Sacks, Harvey
1992
Lectures on conversation. 2 volumes. Ed. by Gail Jefferson. Oxford: Blackwell.
Sacks, Harvey, Emanuel Schegloff and Gail Jefferson
1974
A Simplest Systematics for the Organization of Turn-Taking in Conversation.
Language 50(4): 696735.
Sadock, Jerrold
1978
Schegloff, Emanuel
1972
Note on a conversational practice: Formulating place. In: David Sudnow (ed.),
Schegloff, Emanuel
1982
Discourse as an interactional achievement. In: Deborah Tannen (ed.), Analyzing Discourse: Text and Talk. Georgetown University Round Table on Languages and Linguistics 1981, 7194. Washington: Georgetown University
Press.
345
Schegloff, Emanuel
1984
On some questions and ambiguities in conversation. In: J. Maxwell Atkinson
and John Heritage (eds.), Structures of social action: Studies in conversation
analysis. 2852. Cambridge: Cambridge University Press.
Schegloff, Emanuel
1987
Between micro and macro: Context and other connections. In: Jeffrey Alexander, Bernhard Giesen, Richard Mnch and Neil J. Smelser (eds.), The MicroMacro Link, 207234. Berkeley: University of California Press.
Schegloff, Emanuel
1992
In another context. In: Alessandro Duranti and Charles Goodwin (eds.), Rethinking context, 191288. Cambridge: Cambridge University Press.
Schutz, Alfred
1970
Phenomenology and Social Relations. Ed. by Helmut R. Wagner. Chicago:
Searle, John R.
1969
Speech Acts. Cambridge: University Press.
Senft, Gunter and Roel Smits (eds.)
2000
Annual report 2000. Nijmegen: Max Planck Institute for Psycholinguistics.
Sherzer, Joel
1983
Kuna Ways of Speaking. Austin: University of Texas Press.
Sherzer, Joel
1991
Verbal Art in San Blas. Cambridge: Cambridge University Press.
Sidnell, Jack
1997
Organizing Social and Spatial Location: Elicitations in Indo-Guyanese Village
Talk. Journal of Linguistic Anthropology 7(2): 143165.
Sidnell, Jack
2005
Talk and Practical Epistemology: The Social Life of Knowledge in a Caribbean Community. Amsterdam: Benjamins.
Sidnell, Jack
2010
Conversation Analysis, Comparative Perspectives. Cambridge: Cambridge
University Press.
1976
Shifters, linguistic categories, and cultural description. In: Keith Basso and
Henry Selby (eds.), Meaning in Anthropology, 1155. Albuquerque, NM: University of New Mexico Press.
1978
Deixis and deducibility in a Wasco-Wishram passive of evidence. In: Proceedings of the Fourth Annual Meeting of the Berkeley Linguistics Society,
238253. Berkeley, CA: Berkeley Linguistics Society.
Stivers, Tanya
2007
Alternative recognitionals in person reference. In: Tanya Stivers and Nicholas
Enfield (eds.), Person Reference in Interaction: Linguistic, cultural and social
Swanton, John
1911
Tlingit. In: Handbook of American Indian Languages. Smithsonian Institution,
Bureau of American Ethnology, Bulletin 40, ed. by Franz Boas. 2 vols., I:
346
William F. Hanks
Tao, Hongyin
1999
The grammar of demonstratives in Mandarin conversational discourse: A case
study. Journal of Chinese Linguistics 27(1): 69103
Weissenborn, Jrgen and Wolfgang Klein (eds.)
1982
Here and There: Crosslinguistic Studies on Deixis and Demonstration. Amsterdam: Benjamins.
Wichmann, Sren
1993
Spatial deixis in Azoy Tlapanec. Sprachtypologie und Universalienforschung
(STUF), Berlin, 46(2): 113132.
Wilkins, David
1999
Demonstrative questionnaire: THIS and THAT in comparative perspective. In: David Wilkins (ed.), Manual for the 1999 field session. Nijmegen:
Max-Planck Institute for Psycholinguistics.
Yang, Yuling
2006
A study on the asymmetry of individual zhe and individual na in contexts [in
Chinese]. Chinese Teaching in the World 4: 3341.
Zide, Norman
1972
A Munda demonstrative system: Santali. In: Jacqueline M.C. Thomas and Lucien Bernot (eds.), Langues et Techniques, Nature et Socits I. Approche Linguistique, 267274. Paris: Editions Klincksieck.
12.

Monika Schwarz-Friesel and Manfred Consten
Introduction
This chapter will provide an overview of pragmatic research on how people use
and comprehend referring expressions. A referring expression is typically defined
as one which is produced in order to identify or re-identify (in case of anaphora)
an object or a set of objects for a listener or reader in a relevant domain of text or
discourse. Thus, reference and anaphora are concepts inherently pragmatic in nature. Without considering pragmatic notions (such as focus of information, world
knowledge, contextual influence) their description in linguistic theory would remain insufficient. We will focus on past and current research on reference and anaphora. One line of this research is to delineate the mental processing of referring
expressions with the focus on text comprehension.
1.
Notions of reference
Reference (-theory) is the area of scientific research which deals with the ancient
problem of how language relates to the world. Reference is a term that has undergone major changes during the last five decades of research. In this section, the development from a static notion of reference as a relation independent from language users (1.1) to a dynamic notion of reference as a cognitive process (1.2) is
outlined.
1.1.
Static notions of reference
In the tradition of philosophical logics and formal semantics, reference has been
defined as a static relationship between expressions and the things they denote. The
basis for this notion is an objective view of the world. For many philosophers of
language, the abstract relationship between words and things is primary, irrespective of intentions, contextual influence and conceptual knowledge.
In this view, the function of language is to depict reality; each word stands for a
certain class of objects. Linguistics concentrated on the denotation of different
kinds of referential expressions. The distinction between sense and reference, formulated by Frege a century ago (1892), had been in the focus of semantic research
for quite some time. A semantic theory had to specify for each expression in a language what semantic information that expression conveyed. In the tradition of
348
Russell (1905), who has defined referring as asserting the existence and the
uniqueness of an object, the referential potential of definite descriptions and proper
names became a field of intensive linguistic research (see Thrane 1980). Apart
from the question whether reference is regarded as a static relation or as a process,
it has been discussed which kinds of noun phrase use are referential at all: Donnellan (1966) excluded attributive use of definite NPs from his notion of reference.
Thus, in his well-known example (1) the NP Smiths murderer is used referentially
only in case the speaker knows who Smiths murderer is and intends to identify
this individual for the hearer. Involving hearers roles in reference, Donnellan
approaches a dynamic notion of the term (see 1.2). Still, it crucially depends on
truth conditionality, and there is no notion of reference to mental concepts like
W HOEVER HAS MURDERED S MITH let alone hypothetical referents.
(1) Smiths murderer must be insane.
The main topics in reference theory were naming, truth and extension. By focussing on the semantic features of definite noun phrases and reference to real physical
objects, however, the actual use of referential expressions was largely neglected.
According to formal semantics, reference is not related to an intersubjective, mental or perceptual world, but to a formal model that functions as a substitute for reality (Dowty 1979). In model-theoretic semantics, the intension of an expression is a
function from possible worlds to truth values, and an extension (i.e. the set of possible referents) is the truth value for a particular world. Thus, reference theory in
modern linguistics has been dominated by a formal and not an empirical approach.
A reference theory of truth conditions, however, disregards the mental apparatus of
language users. This semantic, static notion of reference is still (albeit implicitly)
prevailing in some contemporary formal semantic approaches like Discourse Representation Theory.1
1.2.
Dynamic notions of reference
Dynamic notions of reference crucially arose in the framework of pragmatic approaches like the speech act theory; however, some semanticists made attempts to
integrate procedural aspects of reference: Strawson (1950) claimed (at least implicitly) that speakers have to account for the recipients capability of identifying a
referent and that an act of reference has to be accomplished by the recipient. Reference as a dynamic notion is thus a collaborative activity of speakers and hearers
who activate extra-linguistic mental representations by using language. It is based
on a common ground, i.e. knowledge shared by speakers and recipients (Clark and
Marshall 1981; Clark, Schreuder, and Buttrick 1983). Linguists have realized that
it is a mistake to think that the referring expression itself can guarantee uniqueness
in discourse. We assign referential distinctness not simply on the basis of lexical
or grammatical features such as definiteness but also on some vaguer notions of
349
plausibility and assumptions about the speakers knowledge of the distinctive

properties of the domain (Stenning 1978).
In the 1970s interest became centered on the pragmatic aspects of reference.
Reference, it has been argued, is not an abstract semantic relation, but part of a
speech act. This is reflected in Donnellans (1978) terms of speakers reference
versus semantic reference (the latter term being equivalent to meaning). With
Searle (1969), a reference act is part of a propositional act, which consists of the
reference act and a predication about the referent, propositional act and illocutionary act being the core aspects of a speech act. However, pragmatics did not from
the outset pay equal attention to the speakers and the recipients roles in reference
acts: reference is merely defined by the speakers intention to make the recipient
recognize a referent (Donnellan 1966, 1978); accordingly, a reference act is felicitous in case the referent exists, regardless of the hearers reception. Similarly,
Searle (1969) defines successful reference from the speakers point of view, but
he accounts for recipients who, in this model, will ask for more specification until a
state of fully consummated reference is achieved.
In the late 1970s, interest shifted towards a more cognitive based explanation
of reference. Emphasis was now placed on the mental representations and operations involved in referential communication. How recipients identify referents
became part of the broader question of how recipients extract the novel information conveyed by an utterance and incorporate this information into memory
(Clark and Sengul 1979). How do we process referential expressions in on-line
understanding? became one of the main questions. Important work has been done
in psycholinguistics and cognitive psychology dealing with the constructive nature
of referential interpretation. Bransford, Barclay, and Franks (1972) and others have
demonstrated that memory representations of a sentence combine information provided by the meaning of the sentence with information drawn from real-world
knowledge. Empirical studies have shown that language understanding is a constructive process. In on-line-processing studies it has been shown that recipients
instantaneously use pragmatic information (real world knowledge as well as cotextual information) in resolving referential ambiguities and in interpreting anaphors (Marslen-Wilson and Tyler 1982; McDonald and MacWhinney 1995). Accordingly, the structure of semantic memory and the representation of concepts as
part of world knowledge has become the focus especially of neuropsychological
studies relevant for reference (Damasio et al. 1996; Caramazza and Mahon 2006;
Caramazza and Mahon 2009).
In keeping with this mentalistic framework, the concept of world has changed
as well: It is no longer presumed that reference implies the existence of an object in
the physical world (see 1.1.). In reference theory, the traditional realistic approach
to perception is being challenged. We cannot make statements about the world independent of our perception and cognition. We can only talk about the world we
live in, i.e., the world structured by our perceptual and cognitive mechanisms. Ref-
350
erence is concerned with the projected world which is a mental construct of our
mind/brain. Thus, the notion of world is explained as a mental phenomenon
(Jackendoff 1983, 2002; Schwarz 1996).
Reference research has also benefited a lot from the development of cognitive
science, an interdisciplinary approach to the study of mental faculties which has so
strongly been influencing recent research. According to cognitive science, the
mind is too complex to be seen clearly from the perspective of a single discipline.
Thus, the scientific understanding of cognition requires a synthesis of psychology,
linguistics, AI, philosophy, anthropology (and most recently neuroscience), the result of which is an increased interdisciplinary awareness (Schwarz 2008; for methodological reflections and applications Derry, Gernsbacher, and Schunn 2005).
Thus, interesting questions concern the nature of our mental knowledge, how we
represent this knowledge in memory, and how we put this knowledge to use. The
following working hypothesis holds true for cognitive science: The human mind
uses mental representations to store knowledge about the world. Human beings are
viewed as language processors that make use of this body of stored knowledge to
operate on temporary structures.
To sum up, the area of reference theory has radically changed. The main
achievement is that reference is no longer seen as a static relation, but as the result
of a complex mental process which involves all types of knowledge (essentially
perceptual, conceptual and linguistic knowledge).
A modern theory of pragmatic/cognitive reference must take into account a
series of connected topics which until recently have been dealt with in separate
fields: pragmatics, language understanding, perceptual processing, and conceptual
categorization. The rules and principles which underlie referential communication
cannot be explained adequately without the knowledge of how we store and access
both meanings and concepts in memory.2 Thus, cognitive reference theory deals
with three main questions: What is the knowledge basis of referentiality in long
term memory? How do we activate this knowledge in the process of referring?
How does this activation interact with more general procedures of cognition (perception and conceptual categorization)?
1.3.
Text, mind, and world: basic assumptions on referential representation

and underspecification
Current reference theory does not merely describe the function of referential expressions but attempts to explain the production and comprehension of referential
items as part of natural language communication in text and discourse grounded in
pragmatic and cognitive principles. To fully understand a text, a recipient must integrate information in one sentence with information in another. Explaining how
people do so is one of the main tasks of modern text linguistics. On the grammatical surface of the text, particular expressions (discourse markers) or grammatical
351
patterns serve as explicit cohesive ties to relate different parts of the text.3 The text,
as a consequence, is perceived as a coherent structure, rather than as an assembly
of disjointed referential information.
One of the central assumptions in modern text linguistics, however, is that the
explicit grammatical structure of the text is to be regarded as just one part or level
of textual coherence. Accordingly, recent developments in text research suggest
that the implicit conceptual relations in text are of vital importance and deserve
particular attention.4 At the same time, many researchers have realized that no
serious analysis of coherence can be carried out without appealing to its cognitive
aspects.5 Completing the propositional form by cognitive operations in order to get
to the underlying referential representation becomes necessary quite frequently
and often involves some kind of semantic-thematic decomposition as can be seen
in (2):
(2) She dug a hole into the frozen ground.
In order to get to the complete conceptual representation, some slot-filling or referent-recreating operation involving WITH AN INSTRUMENT, TYPICALLY A SHOVEL
has to be performed.6 In (3) it can be seen that these default elaborations play an
important role in establishing and accepting coherence:
(3) She opened the door reluctantly. (a) When she turned the key / (b) ?? When
she turned the spoon / (c) ?? When she turned the shovel
The continuation the key in (a) is accepted as coherent without any cognitive effort,
whereas in (b) and (c) the recipient would have to make an extra inference to establish some kind of (strange or unusual) continuity relation.
Referential underspecification is an essential feature of the relation between
verbal expressions and the conceptual representations they are intended to express.
Therefore, underspecification and implicitness in text should by no means be regarded as some kind of textual deviance.7 On the contrary, systematic underspecification is to be regarded as default. It is the result of applying the principle of relevance to the production process by leaving out superfluous information (see
Sperber and Wilson 1986/21995). In accordance with some principle of cognitive
economy, underspecification is achieved by selecting only the really relevant information and thereby avoiding redundancy (Schwarz 2000a: 83; Sanders and
Spooren 2001: 4). In order to elaborate the linguistically encoded meaning of a text
(which is based on the grammatical and lexical text structure), recipients automatically construct a mental text-world model; i.e. recipients enrich the text base by incorporating both information from the text and information activated through conceptual instantiation and inferential processing.
352
1.4.
Text-world models and reference construction
The text-world model represents a referential constellation of states-of-affairs in

working and in episodic memory which is (in most cases) more complex and elaborated than the semantic text basis, i.e. a propositional level derivable from the text
surface. Accordingly, we have to distinguish three levels: the text surface comprising grammatical structure and cohesive means, the semantic level, which is a level
of lexical meaning and compositionally derivable from the text, and the text-world
model, which is not a text level but a mental level of referential structures
(Schwarz 2001: 18).
From a procedural perspective, semantic enrichment and conceptual elaboration (as part of text-world model building) occur spontaneously and have to be
considered as part of our textual competence. The recipients elaborative capacity
is based on their memory structures and the ability to activate parts of these knowledge structures or to draw inferences. Hence, text comprehension implies both decoding the linguistically encoded meaning of the text and at the same time constructing a mental text-world model which is a cognitive conceptualization of the
world depicted in the text. The term text-world model refers to the assumption that
recipients create a mental model of the world described in a specific text and store it
in episodic memory (Schwarz 2001). The general idea behind this is that verbal expressions serve as mental processing instructions for the recipient (Givon 1992,
1995; Fauconnier 1994). As already pointed out, the text-world model of a text is
based on the information of the text itself but elaborated by the recipients cognitive
processing. It is the result of applying knowledge representations to the text base.
Building up a text-world model is an automatic process which always takes place in
the comprehension process.8 Once a text-world model has been developed during
reading, all subsequently encoded information will be integrated into the configurations of the model. These configurations or mental files, which are stored in working memory, are constantly being updated as the reading process progresses. For
instance, the conceptual file of the referent introduced in the first sentence of
(4) includes the information (YOUNG MAN IN C OLOGNE ) in short-term memory.
(4) Kln. Ein junger Mann wurde gestern Abend Opfer eines brutalen berfalls.
Als der 25jhrige Vizedirektor des Kaufparks in Nippes die Tr ffnete Joachim H. sah pltzlich eine Kanone vor seinen Augen Der Vater von 2 kleinen
Kindern (Newspaper article, Klner Stadtanzeiger, 11 March 2004)
Cologne. A young man was brutally assaulted yesterday evening. When the
25-year-old assistant director of the shopping center in Nippes opened the
door Joachim H. saw a gun right before his eyes The father of 2 little
kids
With each sentence new information is added to the conceptual knot (which functions as a file address in working memory), and the end result of the referential
353
movement is a file which includes the following information: (Y OUNG M AN

J OACHIM H. IS 25 YEARS OLD , HAS 2 LITTLE CHILDREN , WORKS IN C OLOGNE AS ASSISTANT DIRECTOR ).
Building up coherence and resolving anaphoric expressions in text comprehension are two processes which go hand in hand. Focussing on the role of the
text-world model and the leading principle of conceptual plausibility in understanding and constructing coherent representations of texts, the interface between
verbal structures and memory representations has been illustrated. Coherence has
been described as conceptual continuity, in the sense that the states-of-affairs expressed by parts of the text can be connected and integrated into the text-world
model automatically and without effort on the base of our linguistic and conceptual competence. The maintenance of continuity can best be described as a process
of mapping referential representations onto conceptual representations evoked by
prior text.
1.5.
Accessibility of referents
One of the most important aspects of explaining referential coherence in text involves the division between old (or given) and new information. Different cognitive accounts concerning the information distribution and thematic development
have been proposed in recent years.9 Old or given information corresponds to
entities assumed to be accessible in text, new information corresponds to entities
to be assumed inaccessible to recipients.
Considering the resolution of (definite) referential expressions which connect
sentences and serve as signals for topic continuity leads to the question of how recipients access and recover the referents of these expressions in text comprehension and what information has to be available in order to establish continuity. The
notion of accessibility is of particular interest here. This notion refers to the degree
of activation of information in long- or short-term memory (see Givon 2005 on
grounding). It is assumed that highly accessible mental entities (those which are
most active in consciousness) require less processing effort to be retrieved than do
entities of low accessibility.10 Ariel (1990) mentions four factors that affect accessibility: recency of mention (the more recent the last mention of an entity, the
more accessible it will be); saliency (either physical or textual salience); competition (salience of an entity compared to other entities of the same type that may also
be present in the context); unity (whether an antecedent is within the same paragraph or point of view as an anaphor). One of the main functions of referring expressions is to indicate the degree of accessibility of the mental representations of
their referents in a given context (Ariel 1990: 16; see von Heusinger 2000, 2003 for
formal aspects of accessibility theory).
A problem with these approaches arises from the one-to-one correlation that is
assumed to hold between grammatical categories of referring expressions and the
354
accessibility of referents (such as zero-pronouns indicating highly accessible referents, i.e. in case of anaphoric chains, versus demonstratives indicating the lowest
degree of accessibility, i.e. in case of deixis.) Firstly, such a matching seems to be
empirically inadequate (i.e. for demonstrative anaphors expressing the speakers
emotive attitude to the referent, see Averintseva-Klisch and Consten 2007). Secondly, a mere relationship between grammatical forms and referent status, regardless of language users perspectives, runs the risk of being a circular explanation
(Consten 2004; Consten and Loll 2009): from the speakers point of view, properties of referents in a given context seem to be a precondition for using certain referential means. Speakers choose grammatical features (such as definite vs. indefinite, demonstrative vs. non-demonstrative, lexical vs. pronominal NP) according
to the status of the referent in discourse (such as identifiability, first-mention vs. resumption). From the recipients point of view, grammatical features of the utterance evoke expectations about the referent in question. Grammatical features of
referential means are the base for mental operations of reactivating referents retrievable from memory or establishing new referents. Linking accessibility and givenness within the account of our text-world model proposal, referential movement in text comprehension can best be described as a sequence of the following
three transitional states: Activation, re-activation and deactivation.11
Activation of referents takes place when a new mental file (formally represented as a knot in a network) is opened. The referent is activated and stored as a
conceptual label. Hence, incoming information about the same referent may be
filed under this label. This process is to be characterized as re-activation, since the
already established knot in the text-world model is once again activated in working
memory. If a new referent is mentioned in the text, the referent in current focus is
deactivated while at the same time a file for the new referent is opened and stored
as an additional conceptual knot in the text-world model (Schwarz 2000b;
Schwarz-Friesel 2007a). In this regard, given means that the referent is mentally
accessible in text-world model, new means that this is not the case. The distributive flow of information in text comprehension, then, is to be described as an interplay of retaining continuity and progressive development.
Traditionally, definite NPs serving as anaphors are regarded as the main grammatical means to indicate continuity by referring to accessible, given entities. In
the following sections, however, we shall present a number of definite anaphoric
NPs whose referents are seen as accessible even though they would procedurally
be viewed as constituting new instead of given information.
2.
Anaphora and coreference
Anaphorical functions are closely related to the grammatical feature definiteness.

Definiteness is characterized as a grammatical instruction to the recipient to access
355
or construct the referents needed in order to establish a coherent representation.12

Definiteness itself does not spell out the precise connections that have to be constructed in the progress of resolving any individual NP. A unified theory of anaphora has to specify the conditions under which the various uses of anaphors can
be resolved and how the resolution depends on the meaning of the antecedents/anchors.13 In order to provide a full account of how several kinds of anaphors are processed, a cognitive theory of reference resolution must include both representational and procedural factors of cognitive processing and pragmatic elaboration.
2.1.
Topic continuity and antecedent-search-models
The traditional view in text linguistics has been that anaphors are used to continue
a pre-established reference in text by pointing back to a specific antecedent,
usually an NP. Thereby, they sustain the current focus of attention and at least
with longer anaphoric chains make their referent discourse topic of a text segment. Thus, the main anaphoric function is topic continuity. The anaphoric procedure is a linguistic instrument for having the hearer continue (sustain) a previously established focus (Ehlich 1982: 325).14 In this respect, anaphoric
elements tell the recipient where to search for the topic in the existing storage
structure of episodic memory (Givon 1992: 5). Hence, the classical and prototypical anaphoric case can be seen in example (5):
(5) A man (referential expression 11) entered the room. Obviously, he/this man
(referential expression 21) was drunk.
The referential relationship between anaphor and antecedent is based on coreference (indicated by the subscript1), and the link between the expressions denoting
coreference can be described in terms of grammatical congruence and semantic
compatibility. In this respect, anaphors have been described as retrieval cues or
echoes of their antecedents (Hintzman 1987: 412; Gernsbacher 1989: 100). According to this view, anaphora resolution is seen as some kind of search procedure
which aims at discovering the best matching antecedent in textual structure. This
mental checking in resolving anaphors in text comprehension always leads to the
conclusion that the referential expression 2 refers to the same extra linguistic item
as the referential expression 1. In on-line processing, the recipient must decide in
each case of reference whether there is referential identity or not. Considering (6),
one can conclude that the referent of R2 in (a) is identical with the referent of R1,
whereas in (b) this is not the case (which becomes evident to the recipient only
when the relative clause in R2 is processed).
(6a) Peter G. (R11) has beaten his wife. The man (R21) is quite unscrupulous.
(6b) Peter G.(R11) has beaten his wife. The man who lives across the street (R22)
observed this.
356
Anaphors which establish topic continuity by merely reactivating the information

knot of their antecedent involve pronouns, NPs with the same head noun, hyperonyms, synonyms and certain general expressions. The analysis of anaphors, however, calls for a more flexible description. Establishing the relationship of identity
(R EFERENT OF R2 IS IDENTICAL WITH REFERENT OF R1) which is necessary for successful anaphoric processing very often demands more than just checking semantic
compatibility. Reference chains in natural language texts reveal a lot more types of
anaphora which do not fall into the category echo of their antecedent. The establishment of coreference and continuity relies much more on both the context and
constructive mental activity of the recipient than on the grammatical devices in the
text. Hence, establishing coherence in text is best to be described as the interaction
of linguistic indicators and cognitive processes. The text itself gives only some information potential which offers readers certain (necessary) clues to the mental
construction of conceptual representations of what is communicated.
2.2.
Progressive anaphors
We are going to discuss a number of anaphors with explicit antecedents which not
only reactivate old files in text-world model constellations, but activate at the same
time new information about the referent in focus. It should be noted that these
types of anaphors occur quite frequently in natural language texts and are by no
means to be evaluated as some kind of textual deviance. To illustrate these progressive or moving anaphors, consider example (7) which depicts specifying
anaphors.15 The G RANDDAUGHTER -referent, who functions as the protagonist, once
it is introduced with a lexical NP, is referred to with personal- or zero pronouns,
whereas the G RANDMOTHER -referent, as the main theme of the story, is more
closely specified with lexical anaphors.
(7) Rentnerin erwrgt
Hamburg Eine Frau aus Hamburg-Bergedorf hatte ihre Gromutter seit drei
Tagen nicht gesehen und machte sich Sorgen. Als sie die gehbehinderte alte
Dame besuchen wollte, entdeckte sie die 81-Jhrige tot in ihrer Wohnung.
(beginning of a newspaper article, Express, 30 June 2006)
Pensioner [fem.]1 strangled
Hamburg A woman from Hamburg-Bergedorf2 had not met her2 grandmother1 for three days and 2 worried about her1. When she2 was going to visit
the physically handicapped lady1, she2 found the 81-year-old1 dead in her1
flat.
In cases like (7), the anaphoric expressions are much more than a mere semantic
echo of their antecedent. Several new information chunks are filed successively in
the same conceptual node. Hence, continued elaboration through activation processes of the current open file is the case. In on-line comprehension, specifying an-
357
aphors include both re-activation and activation in working memory. In cognitive

terms, their informational status is both given and new, since the text-world model
constellation changes (see Schwarz 2000b). Further, anaphors (with explicit
antecedents) may not only introduce new information about a referent, they may
even trigger new referents into the text-world model.
In accordance with our conception of coherence, felicity of anaphors depends
crucially on the plausibility criterion which is dependent both on long-term memory knowledge and on the text-world model at hand. In (8) we see a combining anaphor (Kombinationsanapher in German, Schwarz 2000b: 122) combing entities into a new referent with different ontological status.16
(8) Give the powder1 into a glass of water2 and drink down the dilution1+2.
Here, a new referent (DILUTION ) is mentally constructed on the basis of old information. Resolving this anaphor means activating a new file for a new referent in
the text-world model. If the lexical anaphor was replaced by a pronoun, the result
would be awkward since lexical meaning has to be provided in order to specify the
new referent.
2.3.
Complex anaphors: condensing and evaluating abstract referents
Referent-creating mechanisms are a major part in the interpretation of so-called

complex anaphors as well, as in (9).
(9) Unbestritten ist, da die Zahl der Arbeitspltze in der Industrie geringer wird,
whrend gleichzeitig das Gewicht des Dienstleistungssektors zunimmt.
a) Dieser Proze ist auch noch lngst nicht abgeschlossen.
(Newspaper article cit. from TiGer Korpus 181381813917; b) and c) are constructed versions)
b) Diese Tatsache prgt die europische Volkswirtschaft seit Mitte des 20. Jahrhunderts.
c) Diese Behauptung ignoriert allerdings Statistiken, die noch Raum fr Zweifel
lassen.
[It is indisputable that [the number of jobs is decreasing in the industrial sector
with the importance of the service sector growing at the same time.] process ] state
a) This process process has not finished yet.
b) This fact fact has shaped European economy since the middle of the 20th century.
c) This claim proposition, however, ignores statistics still leaving doubts.
In these examples, it is not an NP functioning as antecedent, but a whole clause.
There are heterogeneous terms for these cases, see Halliday and Hasan (1976) for
extended reference and reference to fact, Fraurud (1992) and Dahl and Hellmann (1995) for situational anaphora, Schmid (2000) for shell nouns and
358
Asher (1993) for abstract object anaphora, Schwarz (2000a) and Consten,
Knees, and Schwarz-Friesel (2007) for complex anaphora. Furthermore, the
phenomenon is discussed as discourse deixis (Webber 1991), see Lenz (1997)
and Consten (2004) for differentiation between anaphora and discourse deixis.
Referents of the complex anaphoric NPs are not individual entities (although
coded by an NP) but second-order or third-order entities (see Lyons 1989 for aspects of semantic ascent) like events, processes, states, facts, and propositions (in
(9) indicated by the indices p = process, s = state, f = fact). In (9), the referential
structure denoted by the preceding that-clause is picked up as a mere process (i.e.,
concerning its temporal structure, see (a)) or with regard to its truth-conditional
factual state (b), or the whole sentence including the main clause it is indisputable
is regarded as the antecedent (thus, the respective referent is the STATE OF BEING INDISPUTABLE ) (see (c)). This referent becomes subject to a kind of metalinguistic
statement and, as a result, is ontologically changed into an abstract object unbound
to a truth value. With (b) and (c), a higher level in the semantic hierarchy of the text
is produced. Functioning as a paraphrase of a preceding part in the text, a hypostatization takes place denoted by the anaphoric expression. On the basis of old information, a new referent is created in the dynamics of the text. Further, complex
anaphors do not only include a categorization (e.g. the incident, this assumption)
but often evaluate the referent by their lexical content (e.g. this tragedy, this outrageous imputation), see Schwarz-Friesel (2007b: 215216, 2008: 199201).
Hence, two main functions can be made out: Complex anaphors give a classification (in the sense (X Y); THIS EVENT IS A DISASTER ) and a new conceptual file
node (DISASTER ) is established in the text-world model. In cognitive terms of the
given-new-distinction, both re-activation and activation take place at the same
time (for further discussion of complex anaphors see Consten, Knees, and
Schwarz-Friesel 2007, 2009; Consten and Knees 2008).
So far, we have seen that in many cases the processing of anaphoric devices is
not only constrained by their lexical content, but by the activation of conceptual
knowledge representations and the performance of referent-creating operations.
Besides, the examples discussed show that the resolution of many direct anaphors
involves much more than a simple search-and-match-procedure. Their full interpretation requires a more complex cognitive process. Resolving anaphors may involve specification, combination, classification, and/or evaluation mechanisms.
From a procedural perspective, many anaphors combine given and new information for the recipient and therefore can be characterized as given-and-newentities in the text.
2.4.
359
Reference and inference: indirect anaphora referring to mental

representations
Definite NPs may introduce new referents into the text-world model and at the
same time signal topic continuity, which can be most clearly illustrated with the
phenomenon of indirect anaphora which by no means is a marginal use of definite
NPs. A study by Fraurud (1990, 1992) has shown that in a corpus of written Swedish, over 60 % of all definite NPs occur as first mentions of a referent but are, in
fact, necessarily related to the preceding text (see Schwarz 2000a and Consten
2004 for a detailed account of indirect anaphora as a form of domain-bound reference). The passage in (10) is perfectly coherent although there is no textual
antecedent for the definite NPs the rails, the conductor and the engine.
(10) We arrived late since our train stood still at the rails for one hour. The conductor said, There are problems with the engine. (constructed example)
Indirect anaphora fall under the heading of anaphora because the interpretation of
the definite NPs depends on the existence of an adequate textual element in the
preceding text adopting the terminology of Fraurud (1990, 1992), we will call
such elements anchors and the process of linking indirect anaphors to these expressions anchoring18 (see (10) where train functions as anchor). Further, indirect anaphors establish referential continuity just like direct anaphora.
In text comprehension, the resolution of indirect anaphors confronts the cognitive system of the recipient with the problem of identifying and localizing the referent to which the definite description refers in the text. The definite article signals
accessibility of the referent, but in the text structure no such referent can be found.
Thus, we are concerned with the accessibility of implicit referents in the textworld. The prevailing assumption is that there is an associative relationship between trigger and anaphor which helps to establish some link between the two expressions. Thus, these referential instances are often called associative anaphors
(e.g. Hawkins 1978; Heim 1991). However, since only a small number of indirect
reference relations can be explained by the activation of associations, we prefer the
more general term indirect anaphor (see Erk and Gundel 1987).19 According to
most theories, the relation between anchor and indirect anaphor can best be explained in terms of associability, and the anchoring always involves some kind of
inferential linking or bridging (see Clark 1977; Fraurud 1992; Garrod and Sanford
1994; Cornish 1999).
According to another popular view, indirect anaphors are explained as frameevoked entities in implicit focus. In this respect, some authors distinguish direct
and indirect anaphors with the help of explicit and implicit focus.20 Explicit focus
is the current focus of attention, the highly activated part of workspace where incoming information is temporarily held, that is short-term memory which contains
representations of active referents explicitly introduced. Implicit focus is the rep-
360
resentation of entities evoked implicitly as they form part of a frame retrieved from
memory. Whereas the referents of direct anaphora are accessible in explicit focus,
the referents of indirect anaphora have to be accessible in implicit focus, that is as
default values of a frame or script.
However, not all indirect anaphors in a text refer to particular parts of a frame.
Besides, we have seen that the resolution of many direct anaphora demands more
than a search-and-match in short-term memory. Here, too, mental configurations
which exceed the explicit information in the text are built up (by the reader).
Therefore, we will argue for a more complex cognitive approach. Taking a procedural perspective, we want to demonstrate that there are different types of indirect anaphora. A distinction will be drawn between indirect anaphora which are
based on the activation of default knowledge in long-term memory and indirect anaphora which require a more refined cognitive process for their resolution. Further,
we are going to explain why all kinds of indirect anaphora have to be seen as
given-and-new-entities.
Different types of indirect anaphora and their resolution can be gradually distinguished. The criteria established for distinguishing these types are based on the
linguistic expressions used as anchor and on the different knowledge structures involved in interpreting the mental relationship between anchor and indirect anaphor. These conditions of anchoring do not sufficiently cover the whole range of
the complex phenomenon but show some fundamental procedures of interpreting
referring expressions which pragmatically depend on the preceding text. The
examples (which are taken from natural language texts in German) concern the
most frequent and prototypical types of textual indirect anaphora.
(11) Zu den Neuheiten gehrt eine elektronische Promille-Wegfahrsperre fr Berufskraftfahrer []. Mit der Zndung eines Autos verbunden, gibt die Barriere den Motor erst dann frei, wenn der Fahrer zuvor einen Alkoholtest per
Atemprobe abgegeben hat. (Newspaper article cit. from TiGer Korpus,
611561116, modified)
One of the novelties is an electronic blood alcohol immobilizer for professional drivers. Connected with the ignition circuit of a car, the barrier deblocks the engine only when the driver has absolved a field sobriety test by
breath check.
This type of indirect anaphor is established by meronymic relation between the anchor a car and the indirect anaphors the engine and, in a broader sense of meronymy, the driver. There is a part-whole-relationship between the referents ENGINE
and CAR, and an affiliation-relationship between the DRIVER-referent and the CARreferent.
Example (12) shows an indirect anaphor based on the thematic roles of the verb
park in the preceding sentence: the thematic structure of parking involves an agent
(that is addressed directly by you) and an object that is parked. Unlike (11), the
361
close anchoring allows for an pronominal indirect anaphor referring to the expected CAR-referent.
(12) Ich wre wunschlos glcklich, wenn Sie nicht immer auf dem Lehrerparkplatz parken wrden. Das nchste mal lasse ich ihn abschleppen. (Beverly
Hills 90210, German version of the TV-soap, cit. from Consten 2004: 93)
Id be completely happy if you didnt always park on the teachers parking
lot. Next time I will have it towed away.
In example (13), a script-based type is seen. The anchoring of the indirect anaphor
is based on the activation of the appropriate default values in long-term memory.
(13) In jenen letzten Augusttagen des Jahres 1961 erwacht in einem kleinen Krankenhauszimmer das Mdchen Rita Seidel Die Krankenschwester tritt an
das Bett als morgens der Arzt (Christa Wolf, 1973, Der geteilte Himmel, Mnchen: Deutscher Taschenbuchverlag. p.9).
In those last days of August 1961, a girl named Rita Seidel awakens in a hospital room. The nurse comes to the bed the doctor
Establishing a coherent link between the sentences in (13) is to treat the nurse and
the doctor as fulfilling typical roles or default values of the script H OSPITAL in
long-term memory which is activated in a top-down-process by the anchor-expression hospital room. Thus, resolving indirect anaphora involves an elaboration
which extends the explicit representation of the text. Comprehension depends on
the extension of the current representation, taking into account more than textual
information.
It is a common view that interpretation of all indirect anaphors is based on
some kind of inference or linking. This claim, however, neglects the difference between indirect anaphors which are based on a cognitive process involving the activation of semantic roles or default values and indirect anaphors which are truly
based on an inference process. The term inference should not be used for all kinds
of memory processes involved in text comprehension in the sense of any piece of
information that is not explicitly stated in the text (McKoon and Ratcliff 1992:
440), but only for problem-solving activities, that is the constructive use of world
knowledge. Otherwise, it is impossible to describe cognitive processes more precisely and in explicit detail. Lexical decomposition, slot-filling and spreading activation represent types of semantic activation processes which should not be
equated with using knowledge in inferential problem-solving activities. An indirect anaphor which is resolved not merely by activating a semantic role or default
value, but by drawing an inference is exemplified in (14), which is a quite macabre
example:
362
(14) Das Urteil ist somit rechtskrftig und wurde bereits vollstreckt. Die Urne
kann im Krematorium Dresden in Empfang genommen werden.
(From a letter the East German secret service addressed to a detainees wife
who had enquired about her missed husband. Published at a historical exhibition at Erfurt, 2000, cit. from Consten 2004: 98)
Thus, the judgement is non-appealable and has been enforced already. You
can collect the funeral urn at the Dresden crematorium.
The recipient did not yet know that her husband had been sentenced to death, and
this fact is not mentioned in the preceding text of (14) either. Thus, she had to draw
the inference THE ADJUDGEMENT WAS A DEATH SENTENCE in order to integrate the
definite NP the funeral urn coherently. The process of inferential anchoring does
not necessarily involve the activation of one specific frame or script. However, a
lot of inferences are based on script knowledge (see Schwarz 2008: 198f). Besides,
inferential processing does occur within the configuration (and constraints) of the
text-world model, that is, referents have to fulfill requirements of conceptual
plausibility.
The examples so far discussed show that within the group of indirect anaphors
different types can be made out which rely both on different knowledge structures
and processing operations for their anchoring (Schwarz 2000a and 2007a; Consten
2004). Indirect anaphors which rely on the activation of semantic or conceptual
knowledge are resolved within the immediate mental scope of the anchor expression comprising the prototypical information of the cognitive domain. Resolving inferable indirect anaphors involves the expansion of this scope. In general, anchoring conditions of referents for indirect anaphors can be formulated in the
following way: The referent of an indirect anaphor must be either an identifiable
part of the implicit semantic structure of the preceding sentences, or it must be a
default value of a specific frame or script activated by the anchor expression, or it
must be inferable on the basis of cognitive plausibility determined by prototypical
world knowledge.
The definite NP denoting an indirect anaphor refers to an entity being introduced into the text for the very first time (for the function of definiteness, see the
beginning of 2.). In this respect, indirect anaphors carry new information into the
text-world model. Even more, a new node for the new referent has to be established
in an activation process in working memory. On the other hand, at the semantic
level of the text the definite article signals that (given) information is easily accessible and, thus, demands a re-activation process (for indefinite indirect anaphors,
see Schwarz 2000a: 67). Seen as grammatical means with processing instructions,
indirect anaphors signal accessibility of the discourse referent while a new referent
is introduced to the text-world model.
From a given-new-perspective, thus, indirect anaphors combine both referential continuity and elaborative development in information processing. An acti-
363
vation process is taking place by establishing a new file in short-term memory,

while at the same time we have re-activation in long-term memory. More precisely,
the cognitive domain of the anchor expression has been in the status of semi-activation (for aspects of activation levels and the status of semi-activation in longterm memory representations of the anchors, see Schwarz 2000a: 136139).
Hence, re-activation of the semi-activated domain accounts for the ease of processing and the acceptability of indirect anaphors as continuity markers. In this respect,
indirect anaphors serve as means to establish coherence in the same way as direct
anaphors (Schwarz-Friesel 2007a).
To sum up: The anchoring of indirect anaphors is neither solely determined by
the principle of associability nor by the activation of a script. It turns out that both
associability and inferrability are restricted by the conceptual constraint of plausibility. A strong semantic link between anchor and indirect anaphor does not per se
guarantee proper and successful anchoring. Rather, the plausibility within the specific text-world model governs the anchoring process. The immediate resolution of
indirect anaphors comes about through an interaction between the information in
the anchor and the anaphor, the mental accessibility of the implied referent within a
specific context and the focus state of the prior and current textual representation.
From a perspective of referential movement, indirect anaphors have to be regarded
as given-and-new-entities. Combining both easy accessibility and the establishment of new nodes in text-world models, they serve as progressive continuity
markers.
To establish coherence, a mental representation for the textual referent must be
either available in the conceptual text representation or inferable from memory at
the time the anaphoric item is processed. Given this similarity between direct and
indirect anaphors, the interpretation of direct and indirect anaphors seems to obey
similar rules (see Schwarz 2000a; Consten 2003, 2004; Schwarz-Friesel 2007a). In
both cases, the semantics of these expressions only provides a minimal framework,
which needs to be complemented by different cognitive mechanisms at the conceptual level of the text-world model. In this respect, explaining anaphora means at the
same time finding an explanation for the interaction between language structures
and cognitive processes. Thus, anaphoric reference exemplifies prototypically the
pragmatic dimension of reference.
Notes
1. DRT; fundamentals by Kamp (1981) and Heim (1988); for an important application to the
theory of anaphora see Asher 1993.
2. In some theories of cognitive semantics a distinction is drawn between concepts (i.e.
mental representations that store all kinds of knowledge about entities of the world) and
meanings (i.e. concepts bound to linguistic forms). There is strong empirical evidence
from aphasia and lg. acquisition for such a distinction (see Schwarz 2008: 5974).
364
3. See Andersen and Fretheim (2000); Ariel (1998); Fraser (1999); Lenk (1998a and
1998b); Spooren and Risselada (1998).
4. See Gernsbacher and Givon (1995); Bublitz (1999); Spooren (1999); Sanders and
Spooren (2001).
5. See Hobbs (1979); Rickheit and Habel (1995); Sanford and Moxey (1995); Cornish
(1999); Schwarz (2000a); Zelinsky-Wibbelt (2003).
6. See Singer (1994); van den Broek (1994) for a detailed discussion of the procedural aspects of these elaborations; Recanati (2004) for problems that the necessity of conceptual elaborations entails for a notion of literal meaning.
7. This is implied in many accounts, i.e. with respect to underspecified anaphors Reinhart
(1980); Prince (1981); Heim (1988); Ariel (1990); see Schwarz (2000a: 78).
8. For slightly different conceptions of text-worlds or mental models, see de Beaugrande
and Dressler (1981); Van Dijk and Kintsch (1983); Johnson-Laird (1983); Garnham and
Oakhill (1992); Givon (1995); Cornish (1999).
9. See Prince (1981); Grosz (1981); Givon (1992); Gordon, Grosz, and Gilliom (1993);
Chafe (1994); Lambrecht (1994); Schwarz (2000a, b); Hasselgrd et al. (2002) for language specific phenomena at the grammar-discourse-interface.
10. See Gundel et al. (1993); Ariel (1988); Walker, Joshi, and Prince (1998); Byron, BrownSchmidt, and Tanenhaus (2008).
11. See Givon (1992) and Chafe (1994); similar states have been postulated by centering
theory, i.e. continuing, retaining and shifting; see Grosz (1981); Gordon, Grosz, and
Gilliom (1993); Maes (1997); Walker, Joshi, and Prince (1998); Kibble and Power
(1999).
12. For a survey of definite reference, see Consten (2004), for critical remarks on the interpretation of definite determiners, see Consten and Loll (2009).
13. See 2.4 for our notion of anchor.
14. See also Clark and Haviland 1977; Murphy 1985; Carter 1987; Bosch 1988; Givon
1992.
15. They are also called labelling anaphors, Francis (1986: 37); Spezifikationsanaphern
in German, see Schwarz (2000a: 6264.)
16. For similar examples, see Brown and Yule (1983); Stenning (1992); Freitas and Lopes
(1996).
17. The TiGer Korpus is a German newspaper corpus released in 2003, see
www.ims.uni-stuttgart.de/projekte/TIGER. The numbers indicate the sentence numbers
in the original corpus.
18. Hawkins (1978) calls them trigger, Cornish (1999) antecedent trigger.
19. Indirect anaphora has also been called bridging reference or inferrables. We prefer
the term indirect anaphora since it includes all types of anaphora that occur as firstmention definites in text and need some kind of anchoring but do not necessarily depend
on a bridging inference or rely on an associative relation; see Schwarz (2000: 4950).
20. Sanford and Garrod (1981, 1994a, b); Garrod (1995), for critical remarks, see Cornish
(2005) and Cornish et al. (2005).
365
References
Andersen, Gisle and Thorstein Fretheim (eds.)
2000
Pragmatic Markers and Propositional Attitude. Amsterdam: Benjamins.
Ariel, Mira
1988
Referring and accessibility. Journal of Linguistics 24: 6587.
Ariel, Mira
1990
Accessing noun-phrase antecedents. London: Routledge.
Ariel, Mira
1998
Discourse Markers and Form-Function Correlations. In: Andreas H. Jucker
and Yael Ziv (eds.), Discourse Markers: Descriptions and Theory, 223259.
Ariel, Mira
2001
Accessibility theory: an overview. In: Ted Sanders, Joost Schilperoord and
Wilbert Spooren (eds.), Text Representation: linguistic and psycholinguistic
processes, 2987. Amsterdam: Benjamins.
Asher, Nicholas
1993
Reference to Abstract Objects in Discourse. Dordrecht: Kluver.
Averintseva-Klisch, Maria and Manfred Consten
2007
The role of discourse topic and proximity for demonstratives in German and
Russian. In: Bergljot Behrens, Cathrine Fabricius-Hansen, Hilde Hasselgrd
and Stig Johansson (eds.), Language in Contrast Volume 7/Issue 2, 219240.
Beaugrande, Robert-Alain de and Wolfgang U. Dressler
1981
Einfhrung in die Textlinguistik. Tbingen: Niemeyer. English translation: Robert-Alain de Beaugrande and Wolfgang U. Dressler
1994
Introduction to Text Linguistics. London: Longman.
Bosch, Peter
1988
Representing and accessing focussed referents. Language and Cognitive Processes 3: 207231.
Bransford, John D., Richard R. Barclay and Jefferey J.Franks
1972
Sentence memory: a constructive versus interpretive approach. Cognitive Psychology 3: 193209
Broek, Paul van den
1994
Comprehension and memory of narrative texts. In: Morton Ann Gernsbacher
(ed.), Handbook of Psycholinguistics, 539588. San Diego: Academic Press.
1983
Discourse analysis. Cambridge: Cambridge University Press.
Bublitz, Wolfram
1999
Introduction: Views of Coherence. In: Wolfram Bublitz, Uta Lenk and Eija
Ventola (eds.), Coherence in Spoken and Written Discourse, 17. Amsterdam:
Benjamins.
Byron, Donna K., Sarah Brown-Schmidt and Michael K. Tanenhaus
2008
The overlapping distribution of personal and demonstrative pronouns. In:
Gundel, Jeanette K. and Nancy Hedberg, (eds.), Reference Interdisciplinary
Perspectives, 143175. Oxford: Oxford University Press.
366
Caramazza, Alfonso and Bradford Mahon

2006
The organization of conceptual knowledge in the brain: the futures past and
some future directions. Cognitive Neuropsychology 23: 1338.
Carter, David M.
1987
Interpreting Anaphors in Natural Language Texts. Chichester: Ellis Horwood.
Chafe, Wallace L.
1994
Discourse, consciousness, and time. The flow and displacement of conscious
experience in speaking and writing. Chicago: Chicago University Press.
Clark, Herbert H.
1977
Bridging. In: Philip Nicholas Johnson-Laird and Peter C. Wason (eds.), Thinking. Readings in cognitive science, 411420. Cambridge: Cambridge University Press.
Clark, Herbert H. and Susan E. Haviland
1977
Comprehension and the given-new contract. In: Roy O. Freedle (ed.), Discourse
production and comprehension, 140. Hillsdale, NJ: Lawrence Erlbaum.
1981
Webber and Ivan A. Sag (eds.), Elements of discourse understanding, 1063.
New York: Cambridge University Press.
1983
Common ground and the understanding of definite reference. Journal of Verbal
Learning and Verbal Behavior 22: 245258.
Clark, Herbert H. and C. J. Sengul
1979
In search of referents for nouns and pronouns. Memory and Cognition 7:
3541.
Consten, Manfred
2003
Towards a unified model of domain-bound reference. In: Friedrich Lenz (ed.),
Deictic Conceptualisation of Space, Time and Person, 223248. Amsterdam:
Benjamins.
Consten, Manfred
2004
Anaphorisch oder deiktisch? Zu einem integrativen Modell domnengebundener Referenz. Tbingen: Niemeyer.
Consten, Manfred and Mareile Knees
2008
Complex Anaphors in Discourse. In: Anton Benz and Peter Khnlein (eds.),
Constraints in Discourse, 181199. Amsterdam: Benjamins.
Consten, Manfred, Mareile Knees and Monika Schwarz-Friesel
2007
The Function of Complex Anaphors in Texts. In: Monika Schwarz-Friesel, Manfred Consten and Mareile Knees (eds.), Anaphors in Texts, 81102. Amsterdam: Benjamins.
2009
Complex Anaphors. The Impact of Ontology, Cotext and Conceptual Knowledge. In: Jordan Zlatev, Marlene Johansson Falck, Carita Lundmark and Mats
Andrn (eds.), Studies in Language and Cognition, 285302. Newcastle: Cambridge Scholars Publishing.
Consten, Manfred and Annegret Loll
2009
Indirekte Anaphern ein Zirkularitts-problem zwischen Grammatik und
Pragmatik. In: Attila Peteri (ed.), An der Grenze zwischen Grammatik und
Pragmatik, 325332. Frankfurt am Main: Peter Lang.
367
Cornish, Francis
1999
Anaphora, Discourse, and Understanding. Oxford: Clarendon.
Cornish, Francis
2005
Degrees of indirectness: two types of implicit referents and their retrieval via
unaccented pronouns. In: Antnio Branco, Tony McEnery and Ruslan Mitkov
(eds.), Anaphora Processing: Linguistic, cognitive and computational modelling, 199220. Amsterdam: Benjamins.
Cornish, Francis, Alan Garnham, H. Wind Cowles, Marion Fossard and Virginie Andr
2005
Indirect anaphora in English and French: A cross-linguistic study of pronoun
resolution. Journal of Memory and Language 52 (3): 363376.
Dahl, sten and Christina Hellmann
1995
What happens when we use an anaphor? Paper. Stockholm: University, Department of Linguistics.
Damasio, Hanna, Thomas J. Grabowski, Daniel Tranel, Richard D. Hichwa and Antonio R.
Damasio
1996
A neural basis for lexical retrieval. Nature 380: 499505.
Derry, Sharon J., Morton Ann Gernsbacher and Christian D. Schunn (eds.)
2005
Interdisciplinary Collaboration An emerging cognitive science. London:
Routledge.
Donnellan, Keith S.
1966
Donnellan, Keith S.
1978
Speaker reference, descriptions and anaphora. In: Peter Cole (ed.), Pragmatics, 4768. New York: Academic Press.
Dowty, David R.
1979
Word Meaning and Montague Grammar: the Semantics of Verbs and Times in
Generative Semantics and in Montagues PTQ. Dordrecht: Reidel.
Ehlich, Konrad
1982
Anaphora and deixis: Same, similar or different? In: Robert J. Jarvella and
Wolfgang Klein (eds.), Speech, Place and Action: Studies in Deixis and Related Topics, 315339. Chichester: Wiley.
Erk, Feride and Jeanette K. Gundel
1987
The pragmatics of indirect anaphors. In Jef Verschueren and Marcella Bertucelli-Papi (eds.), The pragmatic perspective, 533545. Amsterdam: Benjamins.
Fauconnier, Gilles
1994
Mental spaces: Aspects of meaning construction in natural language. 2nd ed.
Francis, Gill
1986
Anaphoric Nouns. Birmingham: University of Birmingham.
Fraser, Bruce
1999
What are Discourse Markers? Journal of Pragmatics 31: 931952.
Fraurud, Kari
1992
Processing noun phrases in natural discourse. Stockholm: University, Department of Linguistics.
Fraurud, Kari
1990
Definiteness and the processing of NPs in natural discourse. Journal of Semantics 7(4): 395434.
368
Frege, Gottlob
1892
ber Sinn und Bedeutung. Zeitschrift fr Philosophie und philosophische Kritik C: 2550.
Freitas, Sergio and Jos G. Lopes
1996
Solving the reference to mixable entities. Paper read at the Workshop Indirect
Anaphors in Lancaster.
Garnham, Alan and Jane Oakhill (eds.)
1992
Discourse representation and text processing. A special issue of Language and
Cognitiv processes. Hove: Lawrence Erlbaum.
Garrod, Simon
1995
Distinguishing between Explicit and Implicit Focus during Text Comprehension. In: Gert Rickheit and Christopher Habel (eds.), Focus and Coherence in
Discourse Processing, 317. Berlin/New York: de Gruyter.
Garrod, Simon C. and Anthony. J. Sanford
1994
Resolving sentences in a discourse context: How discourse representation affects language understanding. In: Morton Ann Gernsbacher (ed.), Handbook of
psycholinguistics, 675698. San Diego: Academic Press.
Gernsbacher, Morton Ann
1989
Mechanisms that improve referential access. Cognition 32: 99156.
Gernsbacher, Morton Ann and Talmy Givon (eds.)
1995
Coherence in spontaneous text. Amsterdam: Benjamins.
Givon, Talmy
1992
The Grammar of Referential Coherence as Mental Processing Instructions.
Givon, Talmy
1995
Coherence in text vs. coherence in mind. In: Morton Ann Gernsbacher and
Talmy Givon (eds.), Coherence in Spontaneous Text, 59115. Amsterdam:
Benjamins.
Givon, Talmy
2005
Context as Other Minds. The Pragmatics of Sociality, Cognition and Communication. Amsterdam: Benjamins.
Gordon, Peter C., Barbara J. Grosz and Laura A. Gilliom
1993
Pronouns, names, and the centering of attention in Discourse. Cognitive
Science 17: 311347.
Grosz, Barbara
1981
Focusing and description in natural language dialogues. In: Aravind K. Joshi,
Bonnie L. Webber and Ivan A. Sag (eds.), Elements of discourse understanding, 84105. Cambridge: Cambridge University Press.
Gundel, Jeanette K., Nancy Hedberg and Ron Zacharski
1993
Cognitive Status and the Form of Referring Expressions in Discourse. Language 69(2): 274307.
1976
Hasselgrd, Hilde, Stig Johansson, Bergljot Behrens, Cathrine Fabricius-Hansen (eds.)
2002
Information Structure in a Cross-linguistic Perspective. Amsterdam: Rodopi.
Hawkins, John
1978
Definiteness and Indefiniteness: A Study in Reference and Grammaticality Prediction. London: Groom Helm.
369
Heim, Irene
1991
Artikel und Definitheit. In: Arnim von Stechow and Dieter Wunderlich (eds.),
Semantik. Ein internationales Handbuch der zeitgenssischen Forschung,
487535. Berlin/New York: de Gruyter.
Heim, Irene
1988
The semantics of definite and indefinite noun phrases. New York: Garland.
Heusinger, Klaus von
2000
Accessibility, discourse anaphora, and descriptive content. In: Paul Baker
(ed.), UCREL Technical Papers Vol. 12: Special Issue Proceedings of the Third
International Conference on Discourse Anaphora and Anaphor Resolution
(DAARC2000. 173180. Lancaster: Lancaster University).
2003
The double dynamics of definite descriptions. In: Jaroslav Peregrin, (ed.),
Meaning in the Dynamic Turn, 150168. Amsterdam: Elsevier.
Hintzman, Douglas L.
1987
Schema abstraction in a multiple-trace model. Psychological Review 93:
411428.
Hobbs, Jerry R.
1979
Coherence and coreference. Cognitive Science 3: 6790.
Jackendoff, Ray
1983
Semantics and Cognition. Cambridge, MA: MIT-Press.
Jackendoff, Ray
2002
Foundations of language. Oxford: Oxford University Press.
Johnson-Laird, Philip Nicholas
1983
Mental Models. Towards a Cognitive Science of Language, Inference, and
Consciousness. Cambridge, MA: Harvard University Press.
Kamp, Hans
1981
A theory of truth and semantic representation. In: Jeroen Groenendijk, Theo
Janssen and Martin Stokkoff, M. (eds.), Formal methods in the Study of language, 277322. Amsterdam: Mathematical Theory.
Kibble, Rodger and Richard Power
1999
Using centering theory to plan coherent texts. Proceedings of the 12th Amsterdam
Colloquium, 187192. ftp://ftp.itri.bton.ac.uk/pub/reports/ITRI-9919.ps.gz
Lambrecht, Knud
1994
Information structure and sentence form: topic, focus and the mental representation of discourse referents. Cambridge: Cambridge University Press.
Lenk, Uta
1998a. Marking Discourse Coherence. Functions of Discourse Markers in Spoken English. Tbingen: Narr.
Lenk, Uta
1998b
Discourse Markers and Global Coherence in Conversation. Journal of Pragmatics 30: 245257.
Lenz, Friedrich
1997
Diskursdeixis im Englischen. Tbingen: Niemeyer.
Lyons, John
1989
Semantic Ascent: A Neglected Aspect of Syntactic Typology. In: Doug Arnold, Claire Grover and Louisa Sadler (eds.), Essays on grammatical theory
and universal grammer, 153186. Oxford: Clarendon Press.
370
Maes, Alfons
1997
Referent ontology and Centering in Discourse. Journal of Semantics, 14:
207235.
Mahon, Bradford and Alfonso Caramazza
2009
Concepts and Categories: A Cognitive Neuropsychological Perspective. Annual Review of Psychology 60: 2751.
Marslen-Wilson, William and Lorraine Komisarjevsky Tyler
1982
The resolution of discourse-anaphors. Text 2: 263291.
McDonald, Janet L. and Brian MacWhinney
1995
The time course of pronoun resolution: Effects of implicit verb causality and
gender. Journal of Memory and Language 34: 543566.
McKoon, Gail and Roger Ratcliff
1992
Inference during meaning. Psychological review 99: 440466.
Murphy, Gregory L.
1985
Processes of Understanding Anaphora. Journal of Memory and Language 24:
290303.
Prince, Ellen F.
1981
Toward a taxonomy of Given-New-information. In: Peter Cole (ed.), Radical
Recanati, Franois
2004
Reinhart, Tanya
1980
Conditions of coherence. Poetics Today 1: 161180.
Rickheit, Gert and Christopher Habel (eds.)
1995
Focus and Coherence in Discourse Processing. Berlin/New York: de Gruyter.
Russell, Bertrand
1905
On Denoting. Mind 30: 479493.
Sanders, Ted and Wilbert Spooren
2001
Text representation as an interface between language and its users. In: Ted
Sanders, Joost Schiperoord and Wilbert Spooren (eds.), Text Representation:
Linguistic and psycholinguistic aspects, 125. Amsterdam: Benjamins.
Sanford, Anthony J. and Simon C. Garrod
1994a
Resolving sentences in a discourse context. In: M. A. Gernsbacher (ed.),
Handbook of Psycholinguistics, 675698. San Diego: Academic Press.
1994b
Selective processing in text understanding. In: M. A. Gernsbacher (ed.), Handbook of Psycholinguistics, 699719. San Diego: Academic Press.
1981
Understanding Written Language: Explorations in Comprehension Beyond the
Sentence. Chichester: Wiley.
Sanford, Anthony J. and Linda M. Moxey
1995
Aspects of coherence in written language: A psychological perspective. In:
Morton A. Gernsbacher and Talmy Givon (eds.), Coherence in Spontaneous
Text, 161187. Amsterdam: Benjamins.
Schmid, Hans-Jrg
2000
English abstract nouns as conceptual shells. From corpus to cognition. Berlin:
Mouton de Gruyter.
371
Schwarz, Monika
1996
Reference as Process. In: Richard A. Geiger (ed.), Reference in a Multidisciplinary Perspective, 461470. Hildesheim: Olms.
Schwarz, Monika
2000a
Indirekte Anaphern in Texten. Studien zur domnengebundenen Referenz und
Kohrenz im Deutschen. Tbingen: Niemeyer.
Schwarz, Monika
2000b
Textuelle Progression durch Anaphern Aspekte einer prozeduralen ThemaRhema-Analyse. Linguistische Arbeitsberichte 74: 111126.
Schwarz, Monika
2001
Establishing Coherence in Text. Conceptual Continuity Within the Domain of
Text-world Models. In: Logos and Language 1523.
Schwarz, Monika
2008
Einfhrung in die Kognitive Linguistik. 3. Aufl.Tbingen: Francke.
Schwarz-Friesel, Monika
2007a
Indirect anaphora in text: a cognitive account. In: Monika Schwarz-Friesel,
Manfred Consten and Mareile Knees (eds.), Anaphors in Text, 320. Amsterdam: Benjamins.
2007b
Sprache und Emotion. Tbingen: Francke.
Schwarz-Friesel, Monika, Manfred Consten and Mareile Knees (eds.) 2007 Anaphors in
Text. Amsterdam: Benjamins.
Searle, John R.
1969
Singer, Murray
1994
Discourse inference processes. In: Morton Ann Gernsbacher (ed.), Handbook
of Psycholinguistics, 479515. San Diego: Academic Press.
1995 [1986]Relevance: Communication and Cognition.2nd ed. Oxford: Basil Blackwell.
Spooren, Wilbert
1999
Textlinguistik. In: Ralf Prings and Ulrich Schmitz (eds.). Sprache und
Sprachwissenschaft, 187210. Narr: Tbingen.
Spooren, Wilbert and Rodie Risselada
1998
Introduction: Discourse Markers and Coherence Relations. Journal of Pragmatics 30: 131133.
Stenning, Keith D.
1978
Anaphora as an approach to pragmatics. In: Morris Halle, Joan Bresnan, and
George A. Miller (eds.), Linguistic Theory and Psychological Reality, 162
Stenning, Keith D.
1992
Representations of individuals and the processing of reference change. Quaterly Journal of Experimental Psychology. Human Experimental Psychology
45A 3: 373398.
Strawson, Peter F.
1950
On Referring. Mind 59: 320344.
Thrane, Torben
1980
Referential-Semantic Analysis. Cambridge: Cambridge University Press.
372
Van Dijk, Teun A. and Walter Kintsch

1983
Strategies of discourse comprehension. New York: Academic Press.
Walker, Marilyn A., Aravind K. Joshi and Ellen F. Prince (eds.)
1998
Centering Theory in Discourse. Oxford: Clarendon Press.
Webber, Bonnie L.
1991
Structure and ostension in the interpretation of discourse deixis. Language and
Cognitive Processes 6: 107135.
Zelinsky-Wibbelt, Cornelia (ed.)
2003
Text, context, concepts. Berlin: de Gruyter.
13.
Speech acts
Elena Collavin
1.
Introduction
Speech act theory originated from the works of the Oxford philosopher of ordinary
language John Langshaw Austin (19111960). Austin first presented the main tenets of his theory in the lectures he gave at Oxford in the years 19521954 under
the title Words and deeds, and subsequently in the William James Lectures he
delivered at Harvard University in 1955. However, Austin pointed out that those
ideas were formed as early as 1939 (1962:iii). After Austins premature death the
notes of his William James Lectures were edited, supplemented with earlier notes
and tape-recorded lectures, and published in 1962 under the title How to do
Things with Words (Austin 1962). In 1969 one of Austins pupils, the American
philosopher John R. Searle, published his own version of the theory under the title
Speech acts An Essay in the Philosophy of Language, followed by several
more works dedicated to speech acts. Searles systematization and development of
Austins ideas has been very influential, to the point that Searles interpretation of
the theory is at times taken as the definitive view of speech acts.
Two main ideas are at the core of speech act theory: the first holds that the
meaning of an utterance is distinct from the function that the utterance performs
(what we shall call the force of an utterance); the second is that all utterances
amount to the execution of an act. Both these ideas well predate Austins theory
(for a detailed overview see Sbis 2009a). However, his is the first account that incorporates both in a radically innovative philosophical explanation of linguistic
communication. In a nutshell, speech act theory advances the fundamental claim
that speech is a form of action rather than a device for describing the world. The
theory compels us to see communication not simply as the passing of information
between a speaker and a hearer, but rather as the consequential and mutual acting
of participants upon each other. The relevance of the theory extends to the analysis
of linguistic communication in its many forms, and more fundamentally to the
understanding of human communication tout court. Discussion of the theory soon
crossed the borders of language philosophy. As a consequence the literature on the
subject is vast, and some references to speech acts are a staple in artificial intelligence, discourse analysis, second language acquisition, interpreting, education,
and social psychology.
374
2.
Elena Collavin
Origins
During the nineteen thirties Logical Positivism was a mainstream philosophical

movement. Born of polemic with metaphysical theories, the approach charged
philosophy with an apparently modest task: to elucidate, clarify and expose nonsense. According to this disciplinary frame, to understand a sentence amounts to
knowing what its truth conditions are. Only statements that are potentially empirically verifiable make sense. For instance one understands the sentence
(1) Giovanni owns a red car
if one knows under which conditions (1) would be true. Within this framework the
meaning of the sentence can be described as its truth conditions. In order to calculate the truth value of a sentence one must be able to unequivocally identify its referent, in this case, for example, Giovanni. As a consequence much debate at the
time was devoted to statements with non existing referents, like
(2) The unicorn owns a red car
According to some authors, such a statement would be nonsensical since in our
world, no referent could ever be found for the unicorn, and hence no truth value
could ever be bestowed upon it; according to others, (2) is simply false. By either
account a large number of perfectly common statements are rendered uninteresting
for the theory. Only certain kinds of descriptive sentences of ordinary language can
be dealt with by the theory. This theoretical approach is chiefly concerned with the
capacity of human language to depict or represent nature, very much like natural
sciences aim at doing in formalized ways. Although language is obviously used to
do many other things aside from describing, Logical Positivism only concerned
itself with the referential, descriptive faculty of language, other uses being deemed
marginal. Ludwig Wittgenstein, whose Tractatus (1922) had been a main source of
inspiration for Logical Positivism, famously changed his mind and went on to propose a radically different theory of linguistic meaning, which bears affinity with
Austins. In a nutshell, in his late works Wittgenstein identifies linguistic meaning
with whatever participants display they believe it to be, in conformity with the type
of social activity they are engaged in. Meaning in such a view becomes an actionoriented business of strictly local, interpersonal conventions, and is de facto detached from the linguistic forms of communication supporting it. While Wittgenstein believes that linguistic forms obey rules, contrary to Austin he does not focus
his interest on what these rule are.
Austin knew Wittgensteins work. Wittgensteins Philosophical Investigations
were discussed during some of the Saturday morning weekly meetings with colleagues that Austin held over the years in various colleges (Warnock 1969:14).
Austins proposal espouses Wittgensteins action-oriented view of linguistic meaning but strives to regiment it by painstakingly trying to pin down, categorize, and
Speech acts
375
peel apart the layers of actions performed linguistically on the one hand, and on the
other to correlate form and force of utterances. First, he starts with a rebuttal of the
received view of linguistic meaning.
3.
Austin on performative utterances
The pars destruens of Austins argument starts from a kind of utterance which
looks like a statement, and grammatically, I suppose, would be classed as a statement, which is not nonsensical, and yet is not true or false (Austin 1979:235).
Here are some examples:
(3)
(4)
(5)
(6)
(7)
I promise I will come tomorrow

I apologize
I will kill you
I award the plaintiff one million dollars in compensation
I bet you ten dollars that tomorrow it will not rain in San Diego
Coining a new word and an ugly word Austin calls these utterances performatives to indicate that issuing the utterance is performing an action (Austin
1962: 6). In the appropriate circumstances with (3) the speaker takes on an obligation to come tomorrow, with (4) the speaker issues an apology, with (5) the
speaker issues a threat (that is a promise to do something dreadful to the interlocutor), with (6) the speaker attributes an entitlement to the plaintiff, and with (7) the
speaker proposes to bet ten dollars that tomorrow it will not rain in San Diego.
Noticeably, the truth value of the above utterances does not seem to have much
role to play. For instance, to the utterance
(8) A: I promise I will come tomorrow
it would not be in order to respond
(9) B: Thats true
but rather something to the tune of ok, or I do not believe you, or even you always say that which all acknowledge the fact that the speaker was making a (perhaps unreliable) promise. Explicit performatives share a similar grammatical
structure characterized by the use of the present tense indicative and a first person
singular personal pronoun as subject. Such performatives are used to simultaneously name and perform the action denoted by the verb.
The following two utterances
(10) Susan bets you $10 that tomorrow it will not rain in San Diego
(11) I awarded the plaintiff $1,000,000 in compensation
376
Elena Collavin
are descriptive rather than performative due to the use of a third person subject and
(10) and the past tense (11); accordingly, uttering (10) does not result in Susan betting $10, and uttering (11) does not award any sum of money.
The type of verb involved also plays a part, as in
(12) I own a red car
which shares the grammatical structure of an explicit performative but lacks a verbum dicendi, which is another prerequisite for a verb being used as a performative.
Unlike performatives, (12) can thus be assessed on the truth value scale. Austin
calls these latter types of utterances constatives because contrary to performatives they are used to describe rather than to do something. While the linguistic form of utterances seems to be an essential aspect of their being performatives, in addition, a number of extra linguistic requirements must be fulfilled
for a speaker to perform the intended action. These conditions are of a social and
psychological nature:
(A.I) There must exist an accepted conventional procedure having a certain conventional effect, that procedure to include the uttering of certain words by certain persons in
certain circumstances, and further,
(A. 2) the particular persons and circumstances in a given case must be appropriate for
the invocation of the particular procedure invoked.
(B.1) The procedure must be executed by all participants both correctly and
(B.2) completely.
(C.1) Where, as often, the procedure is designed for use by persons having certain
thoughts or feelings, or for the inauguration of certain consequential conduct on the part
of any participant, then a person participating in it and so invoking the procedure must
in fact have those thoughts or feelings, and the participants must intend to so conduct
themselves, and further
(C.2) must actually so conduct themselves subsequently.
(Austin 1962: 1415)
When (5) is uttered by an actor on stage it does not result in the (real) speaker
threatening the (real) addressee, and (6) generates an entitlement for the plaintiff
only if it is pronounced during the appropriate judicial setting by the judge in
charge of deciding the case. The same judge uttering the very same words while
talking with her husband at the restaurant does not create any entitlement. Austin
suggests a way of testing if a verb is performative by putting hereby before the
verb conjugated in the first person singular of the indicative present. As expected
I hereby own a red car is not an acceptable (and thus not a performative) utterance.
Performatives are successful or unsuccessful, felicitous or infelicitous, rather
than true or false. Austin distinguishes two main ways in which a performative utterance can fail. Violations to the conditions under A and B produce misfires; the
act is performed purposefully but is void for lack of the required social convention,
Speech acts
377
persons, and circumstances. Violations of the conditions under C produce abuses;

the act is professed but it is hollow because of lack of the appropriate psychological
state or consequent behavior. In order to be executed some performatives may need
uptake on the part of the interlocutor. An individual person can issue an order or
permission on his own, but cannot make a bet without the cooperation of the interlocutor. When it comes to the psychological state of an individual, Austin makes
clear that while a promise uttered with no intention to honor it is problematic, for
all social purposes the promise has been issued and remains valid, our word is our
bond (Austin 1962: 10). There are plenty of cases in which it is unclear if a faulty
or incomplete performance of a performative utterance results in an infelicity.
While administering the constitutionally prescribed Presidential Oath of Office
to President Obama, Chief Justice John Roberts botched it by inverting the word
order of the oath. As a result the Presidential Swearing In was not perfectly
executed word by word. It was unlikely that the imperfect recitation of the oath
witnessed by millions of citizens would have had legal bearing for the Presidential
Office; nevertheless, later in the day the oath was administered a second time in
closed chambers, and apparently perfectly executed (Toobin 2009).
4.
Austins theory of the total speech act in the total situation
Austin claims that there is a large number of performative verbs in English which,
when put in explicit performative form, effect an action. Here are some: to
christen, to bequeath, to promise, to acquit, to sentence, to resign, to excommunicate, to vote for, to bet. However, Austin had other things in mind than sieving the
dictionary for special classes of verbs. His rhetorical edifice compels us to first see
the difference between two distinct classes of verbs and utterances, only to show
that the distinction does not hold. After having first created the constative/performative dichotomy, he ultimately erodes it and argues that all utterances are in
fact used to perform speech acts. His argument is twofold.
Firstly, there is the matter of linguistic form. We have seen that performatives
are a special class of utterances with a particular linguistic form, but this is not always the case. A sign on the wall of a room that says:
(13) No smoking
or the utterance
(14) Smoking is not allowed
are unequivocally performatives but do not have the explicit first person present indicative performative form. Furthermore,
(15) I most definitely will come tomorrow
378
Elena Collavin
amounts to assuming a commitment, but could not be made more explicit by adding hereby.
The same action, moreover, can be performed by different linguistic means.
Consider:
(16)
(17)
(18)
(19)
(20)
(21)
I compliment you on your wonderful performance!

Wonderful performance!
You were marvelous!
I have never heard a more beautifully executed performance in my life
Glinka himself would have loved it
Bravo!
All amount to complimenting the performer. Clearly, utterances can be performative without having the form of a performative. Austin suggests that explicit
performatives are just more unequivocal than implicit performatives in signaling
the action that is being carried out. Austins argument implies that underlying a
multiplicity of linguistic forms some aspect of the action performed (what we call
the force of the utterance) in (16) to (21) is the same and can be made explicit using
the performative form. The verb used might not be a good force indicator after all;
rather, the mood of the verb, the intonation, adverbs, connecting particles, and overall the circumstances of the utterance all provide indications of what kind of action
is performed (Austin 1962: 7377). Linguistic form, that is, can be deceiving, and
many utterances that do not look like explicit performatives can be performatives.
Hence the domain of performatives grows, while that of constatives shrinks.
Secondly, in Austins original distinction, using a performative utterance is an
act, i.e. is doing something, whereas using a constative utterance is just stating
something but not doing something (Austin 1979: 247). Statements are supposed
to be either true or false; performative utterances either felicitous or infelicitous,
which is unsatisfactory. First, the very concept of truth is problematic. Consider the
utterance
(22) France is hexagonal
which illustrates the difficulties that can arise if we try to use the concept of truth in
the domain of everyday language. If with the statement we meant that France is of
a perfect hexagonal shape than the statement would be false, but in many contexts
we would agree that the description is perfectly acceptable, or true. The truth value
of many descriptive statements is not easily calculable. Furthermore, statements
are liable to be assessed not only for their correspondence or failure to correspond
with facts; they are also liable to the same infelicities that can affect performative
utterances. Let us consider what is wrong with the following sentences:
(23) The cat is on the mat but I do not believe it is
and
Speech acts
379
(24) All Johns children are bald but John hasnt got any children (Austin 1979:
248)
In these examples of Moores paradox, Austin points out that there is avowal of
insincerity that makes the statements nonsensical or rather infelicitous for violation of the conditions under C (cf. above). Frege proposed that to assert a proposition is to present it as true. (But see Alston 2007: 11). With a shift from the domain of propositions to that of utterances we can say that asserting something is
expected to be associated with the belief that what is asserted is true.
Besides, the same criteria that determine the felicity of a performative (appropriate persons and circumstances, the existence of social conventions, correctness
of the execution, subsequent conduct, and the speakers psychological state (feelings, thoughts and intentions)) all come into play as much in uttering statements as
in uttering an order or a promise.
Furthermore, performatives can also be assessed for their correspondence to
fact. Both the following utterances
(25) I killed them
and its explicit performative version
(26) I confess that I killed them
constitute an admission of responsibility, but at the same time both proffer a statement that is either true or false.
We see then that stating something is performing an act just as much as giving an order
or giving a warning; and we see, on the other hand, that, when we give an order or a
warning or a piece of advice, there is a question about how this is related to fact which is
not perhaps so very different from the kind of question that arises when we discuss how
a statement is related to fact. Well, this seems to mean that in its original form our distinction between the performative and the statement is considerably weakened, and indeed breaks down. (Austin 1979: 251)
Austin rejects the constative/performative dichotomy in favor of a general theory

of Speech acts, arguing that all utterances (as actual occurrences of linguistic behavior) are the performance of an action. Utterances do not correspond to well
formed grammatical sentences, or to conversational turns (Sacks, Schegloff, and
Jefferson 1974). Rather, the speech act can be isolated in utterances as something
different and distinct from the meaning of the uttered sentence, something that
Austin, using Freges terminology, calls the force of the utterance.
[] besides the question that has been very much studied in the past as to what a certain
utterance means, there is a further question distinct from this as to what was the force, as
we may call it, of the utterance. (Austin 1979: 251)
What we need besides the old doctrine about meanings is a new doctrine about all the
possible forces of utterances [] (Austin 1979: 251)
380
Elena Collavin
For instance:
(27) Is the door shut?
and
(28) Shut the door
have the same propositional content but differ in the action that the utterance performs, hence in their force, the first utterance being a question, the second an order.
From now on the unit of interest for the theory is the speech act, intended as a verbally performed action. If the business of a speech act is for the speaker to do something with the words he or she utters, a theory of speech acts must explain in which
respect saying something may amount to doing something. After all, doing
something is a very vague expression (Austin 1962: 91). According to Austin
saying something amounts to simultaneously performing three types of acts:
i) A Locutionary act, which is the uttering of a sentence with sense and reference.
The locutionary act can be broken down into a phonetic act (the act of uttering
certain noises), a phatic act (the act of uttering sounds that have sense and reference), and a rhetic act (the act of uttering noises that belong to a certain language vocabulary and grammar). A speech act is always the utterance of a
phone, a pheme and a rheme.
ii) An Illocutionary act, which is the action performed by virtue of the force associated with a given linguistic expression. It is the performance of an act in
saying something, as opposed to the performance of an act of saying something
(Austin 1962: 99). This level of action depends on the social conventions that
allow us verbally to carry out clearly recognizable actions. The illocutionary
force of an utterance can be isolated by asking in which sense we were using a
given utterance. When we describe metapragmatically an utterance as an order,
a request, a suggestion, a statement, a promise, a threat, we refer to its illocutionary force. For instance, we could say: By saying the soup you cooked was
very poor she was criticizing him. We can say so fairly unequivocally because we usually can recognize a criticism (and an order, a promise, and so on)
when we hear or see one. Some verbs (in English, for instance, order, request, promise) are inherently illocutionary: when put in the first person
present indicative they provide an explicit performative, whereas others cannot
be used in the same way (I convince you or I offend you do not work the
same way as an explicit performative).
iii) A Perlocutionary act, which is the production of a consequence by the utterance. Contrary to what occurs at the illocutionary level, perlocutions are not directly achieved by the conventional force of an utterance. They can be intentional or unintentional, and they might involve unexpected effects, different
from those of an illocutionary act. Also, perlocutions occur at a further level, as
Speech acts
381
the interlocutors actual reaction to the speech act. A perlocutionary effect of

the utterance the soup you cooked was very poor might be that ones interlocutor is chagrined, gets angry, pours the soup down the sink, or all three.
The distinction between illocution and perlocution is not clear cut, and it is not at
all apparent that the perlocutionary level can be fully considered as a single level of
action embodied in a given utterance (Austin 1962: 108115). In order to identify
the difference it might help to distinguish between the effects that are built in to
given expression by virtue of its illocutionary force, and those that are not. For instance a (felicitous) order just by virtue of its illocutionary force will generate an
obligation in the person who receives it, but how the individual acts upon the obligation (complying, refusing, questioning, being offended by it, and so on) belongs to the realm of perlocutions.
The locutionary and the perlocutionary levels of the speech act have generated
much debate (Searle, 1969; Searle, 1973, Warnock, 1969, Davis, 1980) but the illocutionary level is the nucleus of the Austinian theory and its later developments.
According to Austin the illocutionary act is connected with the production of effects in the following sense:
(i) The force and the content of the utterance must be understood. I cannot say to
have warned an audience unless it hears what I say and takes what I say in a
certain sense (Austin 1962: 115).
(ii) The illocutionary act needs to take effect (see the above example of how giving an order generates an obligation).
(iii) In many instances there needs to be the manifested cooperation of the audience for the illocutionary act to come into being. The necessary uptake encompasses an interlocutors understanding of the content and force of an utterance, but it can also require a specific second part (or response) to the
utterance that ratifies the first part. While one may protest or give an order on
her own, one definitely cannot bet $100 that tomorrow it will not rain in San
Diego without someone taking up her bet.
Austins theory is rich with suggestions and loose ends, a number of which remain
unresolved. Scholars continue to debate how to distinguish clearly the three levels
of the speech act, how to classify speech acts, how to assess the role of force indicators, and how to apply the values of truth and falsity to speech acts.
5.
Grices theory of non natural meaning
After Austin, the development of speech act theory was greatly influenced by Paul
Grices notion of non natural meaning (Grice 1957; for an overview of Grices theory cf. Huang, this volume). In his work, Austin did not focus on the role that the
382
Elena Collavin
speakers intentions play in the production of speech acts. While he recognized

that the appropriate inner state of the speaker may be fundamental for carrying out
felicitous acts of a certain kind, Austin was chiefly concerned with how particular
linguistic expressions, social conventions, the interlocutors uptake, and allocation
of responsibility to the speaker are involved in carrying out verbally performed actions (Sbis 2009a). Grices theory of speakers meaning revolutionized the study
of language, including speech act theory, by bringing to the fore the communicative intentions of speakers. Grice distinguishes between the meaning of certain
natural events and the intentional communicative acts of humans, or natural signs
or symptoms on the one hand, versus signals (Clark 1996: 126) on the other.
The former have a natural relationship with what they stand for; they are what
Peirce calls indexes (Peirce 1989). The latter instead rely in part at least on the
speakers intentions to signal something. For instance, the characteristic red spots
on a childs face are natural signs that the child has measles, and have what Grice
calls a natural meaning (thus, the red spots on her face mean she has measles).
However, the doctors telling gesture after having examined the child, or her declaring that the child has measles manifests the doctors intention to communicate
what she believes to be the case (thus the doctors gesture meant that the child has
measles). The latter case in Gricean terms is an instance of non natural meaning. Linguistic meaning is always of a non natural kind, even if our talking is
likely also to provide a wealth of information that belongs to the realm of indexes
(indicating our gender, age, geographic origin, social background, and so forth).
Grice further distinguishes between speakers meaning and sentence meaning.
This is based on the notion that understanding someone amounts to understanding
what they intend to communicate to us. In a modified formulation that takes into
account some later arguments (proposed by, among others, Strawson, Searle, and
Sperber and Wilson), speakers meaning is defined as follows:
In presenting s to audience A, a speaker S means for A that p if and only if: (i) S intends
in presenting s to A that A recognize that p in part by recognizing that i. (Clark 1996:
130)
where i is the speakers intention. Speakers meaning is a type of reflexive intention, i contains reference to itself. A key part of the meaning resides in the interlocutors understanding of the effect that the speaker intends to achieve. A communicative act occurs through signaling and recognizing, and a key part of the
process involves understanding the communicative intentions of the speaker. Communication is a joint action, as becomes apparent in cases where the literal meaning of a communicative act appears to be lacking, unclear, incomplete, false or absurd, but communication still succeeds. Indeed, Grice later proposed a general
theory of meaning based precisely on what he calls the Cooperative Principle,
which alongside the notion of conversational implicature provides a universal explanation of why, with the cooperation of ones interlocutor and with mutual infer-
Speech acts
383
ential calculations, it is possible to communicate with the most eccentric linguistic

and non linguistic means.
6.
Searles theory of speech acts
Searle (1969) criticizes Grices account of meaning because it relies almost exclusively on the speakers intentions. A given meaning can be produced by any sentence given that the circumstances make possible the appropriate intentions
(Searle 1969: 45). Searle tries to strike a balance between intentionality and the
role of conventions in the formation of meaning.
In the performance of an illocutionary act in the literal utterance of a sentence, the
speaker intends to produce a certain effect by means of getting the hearer to recognize
his intention to produce that effect; and furthermore, if he is using words literally, he intends this recognition to be achieved in virtue of the fact that the rules for using the expressions he utters associate the expression with the production of that effect. It is this
combination of elements which we shall need to express in our analysis of the illocutionary act. (Searle 1969: 45)
Assuming that the illocutionary force is signaled by indicators of force (that Searle
calls Illocutionary Force Indicating Devices) the challenge is to explain exactly
how the two force and force indicators are linked. For Searle the force indicators are constitutive of the force. Rather than revealing it they create it. Whereas
for Austin felicity conditions needed to be fulfilled for the speech act to be carried
out successfully, according to Searle felicity conditions are constitutive of the act
itself. He distinguishes two elements in the syntactic structure of the sentence,
which he calls the propositional indicator, and the illocutionary force indicator.
(Searle 1969: 30) This is his version of the distinction between the illocutionary act
and the propositional content of the illocutionary act.
Searle rejects Austins threefold distinction within the locutionary level of the
speech act (phonetic, phatic, rhetic) in order to locate the locutionary act at the
level of enunciation (the utterance act). Others have also criticized Austins
tripartite distinction claiming that it is hard to keep separate the locutionary act
from its constituent levels (Warnock 1969). While some scholars have described
Searles work as largely a systematization of Austins (Levinson 1983: 238), according to Sbis (2009a) Searles theory sharply diverges from Austins. For Searle
the speech act coincides with the illocutionary act, a conclusion that stems from the
fact that Searle views speech acts as physical and psychological gestures of the
speaker, whereas for Austin the action corresponds with the (necessarily interpersonal) ascription of responsibility to the agent for the production of a given illocutionary effect.
Searles illocutionary act/speech act is divided between the illocutionary force
and the propositional content, schematized by: F(p), where F describes the illocu-
384
Elena Collavin
tionary force indicators and p is the proposition. The propositional content

determines the conditions of satisfaction, truth conditions in the case of belief,
obedience in the case of a promise. Searle proceeds by analyzing the conditions
sufficient for the successful performance of illocutionary acts.
A proposition is what is asserted in the act of asserting, what is stated in the act of stating
[] an assertion is a [] commitment to the truth of a proposition (Searle 1969: 29)
The illocutionary force indicator shows how the proposition is to be taken, or to put it
another way, what illocutionary force the utterance is to have; that is, what illocutionary
act the speaker is performing in the utterance of the sentence.
(Searle 1969: 30)
Some conditions are common to most illocutionary acts, while others are specific
to some types of acts. Austin proposed a classification based on a list of performative verbs, whereas Searles is based on the felicity condition of illocutionary
acts: all have a propositional content which specifies which propositional content
the speaker has to express, preparatory conditions which are contextual but linked
to the inner states of the speaker (what he thinks, what he believes, and so on), sincerity conditions and the essential condition which specifies which kind of illocutionary act a speaker intends to carry out. In his later work, Searle focuses on three
main criteria for the classification of speech acts: the purpose of the type of act
(what he calls the illocutionary point), the direction of fit of the type of act (from
words to world, or from world to words) and the expressed psychological state of
the speaker (Searle 1975).
In either case, in Searles writing the illocutionary effect that results from successfully carrying out an illocutionary act depends on the interlocutors understanding of the force and propositional content of the act, and coincides with the interlocutor recognizing the speakers intentions. Sbis (2009a, 2009b) notices that
Searles theory is problematic precisely because it makes the illocutionary effect of
the act coincide with the recognition of the speakers intentions by the hearer.
Rather than looking at the specific illocutionary effects of an act, Searles analysis
focuses on the types of intentions of the speaker, thus constructing action as the
psycho-physical gesture of the speaker. Sbis (2009b) contends that contrary to
Searle, to Austin illocutionary acts are conventional in the sense that they have
conventional effects and underlie the interpersonal rather than the mental aspect of
uptake of the illocutionary force. On the role of intentionality Searle explains that
he sees language as the natural extension of non linguistic biological capacities,
and linguistic meaning as an extension of fundamental pre linguistic intentionality
of a biologically nature like hunger and sexual desire. Perceptions and beliefs are
also forms of intentionality because to hold beliefs is to hold something that determines conditions of satisfaction, and to hold the capacity to recognize if they are
fulfilled or not (Searle 2007: 21). Indeed, to Searle linguistic action is in essence a
mental act of the speaker. What gets communicated in speech acts are intentional
Speech acts
385
states (Searle 2007: 28). Thus, the analysis focuses on how propositional content,
conditions of satisfaction, psychological mode and direction of fit function in the
manifestation of intentions. For instance, beliefs aim to represent how things are,
thus they have a mind-to-world direction of fit, while desires have a world-to-mind
direction of fit. Searle (2002: 7) conceives of the structure of mind and the structure
of speech acts as parallel because the structure of mental states and the structure of
intentional states are similar. While he concedes that at times our verbal actions do
not have the principal scope to transmit our mental states, he underlines that even
when we carry out speech acts that have not as principal role the function of expressing feelings, we still express a mental state in the form of a sincerity condition
(Searle 2002: 7).
7.
Searle on indirect speech acts
Searle points out that a speakers utterance meaning and the sentence meaning frequently diverge (Searle 1975: 59). This is the case most obviously with irony,
metaphorical uses, and insinuation. However, even in most ordinary circumstances
we can often see no correspondence between the three major sentence types in
English (declarative, imperative and interrogative) and the forces that prototypically correspond to them (stating, ordering and requesting). Consider
(29) I want you to do it (Searle 1975: 59)
which is a request but has the declarative form of a statement. This is a very common occurrence. In English, for instance, imperatives are routinely used to invite,
as in take a seat and come in. Utterances that have the illocutionary force indicators for one kind of illocutionary act can then be used to perform another kind
of illocutionary act. This is a serious problem for the theory, which is based on the
conventional nature of the illocutionary force, and on the existence of predictable
indicators of force. A way to treat utterances whose force differs from what
their force indicators suggest is to assume that they have a literal force, which is
associated by rule, and an indirect force, which is inferred by virtue or further information available to an interlocutor. Searle offers an explanation for the case of
directives based on his own felicity conditions. Let us consider the utterance can
you pass the salt? Given a Speaker S, a hearer H, and an action A (passing the salt)
the felicity conditions for such illocutionary act are:
Preparatory condition: H is able to perform A
Sincerity condition: S wants H to do A
Propositional content condition: S predicates a future act A of H
Essential condition: Counts as an attempt by S to get H to do A
(Searle 1975: 71)
386
Elena Collavin
The request is indirect because the speaker seeks to ascertain if the interlocutor is
capable of performing the action while in fact the speaker is actually requesting the
hearer to pass the salt. The ability of the hearer to pass the salt is a preparatory condition for requesting to pass the salt. Still, this does not explain why it is the case
that only certain requests are successfully made by asking if the interlocutor can
perform the wanted action. As a preparatory condition for any request the speaker
must hold the belief that the interlocutor can carry out the request, so why is it the
case that sometimes the speaker asks? The answer may be found in the rituals of
social interaction and the obligations of politeness (Brown and Levinson 1987), a
dimension that is not integrated in the notion of illocutionary force.
8.
Bach and Harnishs theory of speech acts
Although by far the most influential, Searles approach is not the only one to emphasize the role of speakers intention for explaining illocution. With the aim of
offering a philosophically inspired explanation of linguistic communication, Bach
and Harnish (1979) propose a different but equally Gricean-inspired, speaker-intention-focused classification of speech acts. They distinguish between conventions and rules, the conventions being actions that when performed in the appropriate circumstances count as something else, while rules are socially expected
forms of behavior (Bach and Harnish 1979: 121). Based on their categorical division between conventions and rules the authors draw a distinction between communicative and conventional illocutionary acts: Communicative illocutionary
acts succeed by means of recognition of intention, whereas conventional ones
succeed by satisfying a convention (Bach and Harnish, 1979: 110). Conventional
illocutionary acts are endemic to particular institutions. For an utterance to be a
conventional illocutionary act not only must it be the utterance that the convention requires, it must be issued by the right person under the right circumstances
(Bach and Harnish 1979: 110). Examples of conventional illocutionary acts are
christening, billing, sentencing, disqualifying. Conventional illocutionary acts all
belong to the class of what they call effectives and verdictives. These types of acts
are considered not essentially communicative because they are not implemented
by recognition of intention but by convention. Communicative illocutionary acts
instead are based not on conventions but on the recognition of the intentions of
the speaker (the illocutionary effect), and their classification is based on the
illocutionary intents of the speaker. Illocutionary intents correspond to types of
attitudes expressed by the speaker. The recognition of intentions occurs when the
hearer takes the speakers utterance as reason to believe that the speaker has the
expressed attitude. Although Bach and Harnish assign a different role to conventionality in the discussion of illocutionary force, they side with Searle in limiting
the illocutionary force to the recognition of intention of the speaker, and leaving
Speech acts
387
the conventional effects of illocutionary acts within the realm of perlocutions

(Sbisa 2009b).
9.
A plan based theory of speech acts
Philip Cohen and Raymond Perrault (1979) propose to describe speech acts as
planning operators. While espousing Searles recognition of an intention based notion of speech act, they envision a formalization that would allow, for any communicative act, to isolate which intentions are involved and on which bases the
speaker expects and intends those intentions to be recognized (Cohen and Perrault
1979: 178). They propose a theory that formally models the possible intentions
underlying speech acts. Intentions are conceived as plans and a link is shown between plans and non linguistic behavior.
This approach is focused on the individual who in the pursuit of a goal, plans
and enacts a sequence of actions, the communication of desires, intentions, and
emotional states, and which might include facilitating or making sure that others
recognize these plans. Actions are defined in terms of operators that are broken
down in preconditions, effects and bodies, the latter being the means by which effects are achieved (Cohen and Perrault 1979: 178). Operators are representations;
they are evaluated on the basis of ones representation of the world (Cohen and
Perrault call the speaker problem-solver). The idea is that part of ones model of
the world consists of a symbolic description of other peoples models of the world
(something akin to I know what you know or more modestly I believe that I
know what you believe). Speech acts act as operators on the models that speaker
and hearer maintain of each other.
This is not much different from Grices model for calculating conversational
implicatures (Grice 1975), which occur when what is said is different from what is
meant, so that the accurate comprehension of the literal meaning of what is being
said would not constitute understanding. Implicatures are pragmatic inferences
that require the hearer to make a mental calculation in order to decide what the
speaker really meant to say, for if taken literally what was said would not be fitting
in the current conversation. Grice bases his model for calculating implicatures on
the general Cooperative Principle, a descriptive rule stating that participants contributions to a conversation are relevant to the task at hand. The principle acknowledges that in order to comprehend non literal meaning we need to believe that our
interlocutor is being cooperative, and as a consequence that his or her contribution
must be relevant to the task we are involved in. A successful performance of a
speech act produces some modification of the speakers model of the hearer, and of
the hearers model of the speaker (Cohen and Perrault 1979: 179). The question is
which modifications? Modeling their proposal on Chomskys notion of the ideal
competent speaker, the authors focus on a competence theory of speech acts that
388
Elena Collavin
enables to state the conditions under which a competent participant can make the
determination necessary to classify ones utterances as amounting to the performance of a given speech act.
The model has been embodied in a computer program, and detailed for two
kinds of speech acts: making requests and giving information. But even though
Cohen and Perraults theory has been influential on artificial intelligence studies
applying speech acts theory, it suffers from the same flaws that plague Searles
model: The speech act occurs within the speakers mind as the expression of his or
her intentions, leaving no room for a constitutive contribution on the part of the interlocutor, whose role is only to decipher the speakers intentions.
10.
Speech acts and the analysis of discourse
To apply speech act theory to the analysis of discourse means to try and isolate the
actions that are being performed in ongoing interaction or in a text. No matter how
speech acts are conceived, the analysis will consist in mapping discourse onto categories of speech acts. As Levinson (1983: 278) points out, this could be either
trivial, if we were to wrongly assume that the force of an utterance is indicated literally by linguistic devices, or otherwise it could involve the attempt to predict the
force of a part of discourse in real circumstances. This proves to be largely an impossible task because the function of a given utterance is so essentially dependent
on the overall circumstances in which it occurs that there may be little left to do for
any theory of speech acts. Consider the following example given by Levinson,
which is quoted from Sinclair and Coulthards ethnographic work in the classroom:
(30) Teacher: What are you laughing at?
Child: Nothing
(Sinclair and Coulthard 1975: 30)
It is the second part of the exchange as much as the scholastic context and the social roles of the participants that provide the cue that the teachers utterance is an
order. The following example is taken from a 1993 televised political debate in
Italy:
(31) Casini: I want to ask you this question. Do you want to solve the problems of
the south of Italy, or is it the case that you do not give a damn about the problems of the south of Italy?
The closed question, which seems to syntactically anticipate either yes or no
as an answer, is most likely a challenge and an accusation, which as it has been
shown on the bases of corpus studies, is usually met with a denial (Schegloff, Jefferson and Sacks 1977). Further, an assertion like The door is open in the appro-
Speech acts
389
priate circumstances can function as a declarative, a request, a criticism, a question, a complaint, and possibly other kinds of actions as well. Because ultimately it
is not possible to associate form with function in the analysis of speech acts, the
usefulness of the theory to explain how actions are performed in discourse is limited.
As Levinson points out, failing the hypothesis of the literal force, we are left
with an infinite array of language games (Wittgenstein 1953: 1011) within which
utterances acquire their meaning (and their force, if we wish to keep the two separated) by virtue of local, situated agreements among participants. Within a theory
of speech acts it must be possible to unpack an expression and reveal the underlying illocutionary force of it, and assume that the original expression and the explicit one will display the same force. However, in the analysis of interaction we
often find that it is not possible to univocally isolate one particular illocutionary
force as it has been categorized within the theory. Further, some of the very reasons
why it is interesting to study how people discursively act with one another might be
seriously impeded if we are to assume that in the analysis of the speech act other elements, such as those indexing social relationship in terms of closeness and hierarchy, are irrelevant for the force of the utterance. For instance, the utterances
(32) Can you pass the salt?
(33) Would you be so kind as to pass the salt?
(34) Salt please.
cannot really be described as fully equivalent in terms of their illocutionary force
precisely because they are not equivalent in how they index social relations between participants. Arguably, this aspect, while it does contribute to the forcefulness and the compulsory strength of a request, is not clearly accounted for in terms
of illocutionary force. This is even more apparent in languages that use grammatical markers to distinguish formal versus informal forms of addressing. In Italian
for instance
(35) Mi passi il sale? Do you [+informal] pass me the salt?
and
(36) Mi potrebbe passare il sale? Could you [+formal] pass me the salt?
have their illocutionary force at least partially embodied in the very grammatical
markings that reveal the relationship between participants: the intimate or informal
relationship triggers a more forceful and direct request in (35), while the formal relationship in (36) triggers a weaker and less imposing request. Within the theory
there is no clear way to account for distinguishing between an illocutionary act of
request made between strangers and one made between intimate companions, but
to ignore the socially loaded deictic aspects of the exchange is to leave out elements that might actually weight on the force of the action performed. Further,
390
Elena Collavin
Levinson has shown that it is hard to associate a certain set of Searlian felicity conditions to actual elements of speech exchanges: questions, for instance, as a category are too variable to be tied by a given set of felicity conditions, and even
apologies do not all have the same requirements of felicity conditions (Levinson
1983: 281; Levinson 1979).
Responding to the objections moved by Levinson and Schegloff to the applicability of speech act theory to the analysis of discourse, Rees (1991) defends the
specific contribution that the theory can provide to the analysis of verbal interaction, which she suggests be used alongside Conversation Analysis. Van Rees argues that already in Searles work there is a recognition that a single utterance
might have multiple forces:
(37) Its really quite late
can be an objection, a suggestion, or even a request (Searle 1969: 70) (and a load of
further acts), and in interaction the illocutionary force might reside in supra- and
sub-sentential units which do not coincide with a sentence at all (Rees 1991: 34).
What counts in the attribution of illocutionary force to a part of discourse then is
not a simple, literal association between utterance and force but, as she claims, the
possibility to calculate in a replicable, non arbitrary way the force of a given stretch
of talk. This compromising approach leaves unresolved the key issue of linguistic
form in the matter of the calculability of the illocutionary force. After all, it just
shifts the problem in the realm of the competence of the speaker, but still profitably
uses the notions of performative and illocutionary force in the analysis of discourse, and has won over many scholars interested in the study of verbal interaction. While the difficulty to find a rule to relate form and function in a wide array of
real occurrences of speech acts remains unresolved, key concepts from the theory
have been widely adopted within the domain of discourse analysis.
11.
Beyond speech act theory: Action and the body
The speech act approach has largely been abandoned by those dedicated to the ethnographic study of linguistic action. Critics note that in both Austins and Searles
formulation, the unit of analysis is the surgically delimited, well-formed, grammatical sentence, and that despite Austins early intuitions the developments of the
theory devote no attention to discourse level phenomena which cannot be analyzed
without taking social interaction (Goffman 1983) into account. Echoing Austin,
Agha points out that
the utterance of a performative locution is only a fragment of the performative act; it is
merely the leading incident (to use Austins phrase) in a larger order of happenings
whose shape must be specified by further semiotic accompaniments in order for the utterance to have the effect in question (Agha 2007: 58).
Speech acts
391
In Austins formulation the total speech act is to be understood in the total situation, a claim that recognizes the fundamental role of participants other than
the speaker, and of social arrangements alongside syntactic rules in carrying out
speech acts. The division echoes the dichotomy between the meaning of a sentence
on the one hand and the role of context on the other, where context supplements,
clarifies or modifies the force of the verbal action. A distinction between language
as an autonomous and primordial code and additional information provided by
extra linguistic cues is rejected by those who study cognition in action and who
advocate an ethnographic approach, which considers different kinds of phenomena instantiated in diverse media (Goodwin 2000: 1490) to understand action as
jointly achieved by participants through talk, body movement and material affordances. Goodwin argues that a primordial site for the study of cognition and human
action is an actual situation where several people talk while engaged in a task
which involves their bodies in space, as well as tools and other physical objects
around which they are working. Goodwins approach relies on the detailed transcription and analysis of video recorded data of naturally occurring multi party interactions. The data can be girls enthralled in a game of hopscotch, archeologists
intent on unearthing an artifact, or surgeons performing a laparoscopic operation.
Methodologically it is a far cry from the study of linguistic action which starts with
a set of verbs culled from a languages lexicon, and proceeds with the analysis of
hypothetical utterances pronounced in imagined contexts the confines of which can
be drawn at leisure. While studying linguistic action, this view pays attention to the
multiplicity of semiotic resources available to participants and argues that
the construction of action through talk within a situated interaction is accomplished
through the temporally unfolding juxtaposition of quite different kinds of semiotic resources [] [T]hrough this process the human body is made publicly visible as the site
for a range of structurally different kinds of display implicated in the constitution of the
actions of the moment (Goodwin 2000: 1490).
Rather than being treated as distinct domains, language and material structure in
the environment are integrated components of a common process for the social
production of meaning and action (Goodwin 2000: 1490). This perspective proposes a radically different geography of cognition (Goodwin 2000: 1490), which
is inspired by Erving Goffmans performance oriented approach to complex interactional situations (Goffman 1959, 1983) by works of the Vygotskian circle and by
ethnographic studies of workplace practices.1
Further, Goodwins approach pays close attention to the linguistic nuances of
what is said, to the synchronic details of the interaction as they have been described
within Conversation Analysis (Sacks, Schegloff and Jefferson 1974; Sacks 1995),
and to the organization of face to face interaction. The human body is the locus for
the occurrence of publicly visible displays that are implicated in the cooperative
construction of actions and meanings. In these studies action is rarely achieved by
392
Elena Collavin
one individual, and meaning is a public, social phenomenon that does not somehow reside in the mind of a speaker. In his ethnographic work on the interactional
life of a man who suffered a stroke and was left with the use of only two words,
Goodwin shows how in such an extreme communicative situation meanings are
co-constructed, and actions are the ongoing joint product of the work of participants, always subject to renegotiation and redirection. While apparently distant
from some hypothetical model of interaction between competent participants, this
aphasic mans situation is shown to be exemplary of how actions are interactionally performed, and co-constituted through the cooperation of participants.
Predictably, Goodwin rejects Searles intention centered notion of speech acts.
Searle, like most speech acts theorists, focuses almost exclusively on linguistic
properties of an utterance, everything else being relegated to perhaps necessary but
largely intractable context. A fundamental flaw that Goodwin pinpoints is the role
given to participants in the performance of a speech act. In Searles theory the
hearer exists only in the speakers imagination; the role of the hearer is to understand the force of the utterance, not to co-construct the actions that are accomplished with talk (Goodwin 2000: 1491). In Goodwins work the analysis of action
moves away from speakers intentions and returns to something closer to Austins
notion of uptake. Social action is linked reflexively to its interpretation (Goodwin 2000: 1491). It can be accomplished only when both speaker and addressee
can systematically recognize what is being done not utterance by utterance but
within the evolving scenarios that are being projected into from moment to moment. The mutual understanding of what can be the relevant next move allows
coordinated chains of which are both publicly available and jointly produced.
Speech acts cannot rely on the mental life of a single participant (the speaker);
rather, the notion relies on the sequential analysis of talk in interaction including
such intersubjective mechanism as turn taking and repair (Goodwin 2000: 1492)
where meanings are publicly negotiated. In this context action is understood as
encompassing this interactively organized process of public recognition of meaningful events reflexively linked to the ongoing production of these same events
through the use of appropriate semiotic resources within an unfolding temporal
horizon (Goodwin 2000: 1492).
In 1983 Levinson foresaw that unless a correlation could be found between
form and force in everyday language use it would be hard to maintain a level of
illocutionary force distinct from other action oriented facets of an utterance (Levinson 1983: 282) including social deixis, perlocution, and inferentially available
meanings attachable to the utterance. So far, no version of the theory has offered
such a correlation. Nonetheless, the main tenets of Austins and Searles work
stand as a groundwork that any general theory of language use needs to confront.
Speech acts
393
Notes
1. In particular Ed Hutchins notion of cognition in the wild, which he developed during
his long lasting research of collaborative cognition in the domain of piloting military and
commercial airplanes and freight ships (Hutchins 1995; Hutchins and Palen 1997).
References
Agha, Asif
2007
Language and social relations. Cambridge: Cambridge University Press.
Alston, William P.
2007
Illocutionary acts and truth. In: Dirk Greimann and Geo Siegwart (eds.), Truth
and speech acts, 1030. New York: Routledge,
Austin, John L.
1962
How to do things with words. London: Oxford University Press.
Austin, John L.
1979 [1961] Philosophical papers. Ed. by James O. Urmson and Geoffrey J. Warnock,
3rd ed. Oxford: Oxford University Press. [First published by Clarendon Press].
1979
Linguistic communication and speech acts. Cambridge, MA: MIT Press.
Brown, Penelope and Stephen Levinson
1987
Politeness: Some universals in language usage. Cambridge: Cambridge University Press.
Clark, Herbert H.
1996
Using language. Cambridge: Cambridge University Press.
Cohen, Philip R. and C. Raymond Perrault
1979
Elements of a plan-based theory of speech acts. Cognitive Science 6: 177212.
Davis, Steven
1980
Perlocutions. In: John R. Searle, Manfred Bierwisch and Ferenc Kiefer (eds.),
Speech act theory and pragmatics, 3755. Dordrecht: Reidel.
Frege, Gottlob
1918
Der Gedanke. Eine logische Untersuchung. Beitrge zur Philosophie des deutschen Idealismus 1(2): 5877. [Translated in Gottlob Frege, 1988. Ricerche
logiche, 4374. Milano, Guerini,].
Goffman, Erving
1959
The presentation of self in everyday life. Garden City, NY: Doubleday, Anchor
Books.
Goffman, Erving
1983
The interaction order. American Sociological Review 48: 147.
Goodwin, Charles
2000
Action and embodiment within situated human interaction. Journal of Pragmatics 32: 14891522.
Grice, H. Paul
1957
Meaning. Philosophical Review 66: 377388. [Reprinted in Peter F. Strawson
(ed.), 1971. Philosophical logic. 3948. Oxford: Oxford University Press].
394
Elena Collavin
Grice, H. Paul
1975
Logic and conversation. In: Peter Cole and Jerrold Morgan (eds.), Syntax and
semantics 3: Speech acts, 4158. London: Academic Press.
Hutchins, Edwin
1995
Cognition in the wild. Cambridge, MA: MIT Press.
Hutchins, Edwin and Leysia Palen
1997
Constructing meaning from space, gesture and speech. In: Lauren Resnick,
Roger Saljio, Clotilde Pontecorvo and Barbara Burge (eds.), Discourse, tools
and reasoning: Essays on situated cognition, 2340. Berlin: Springer.
Levinson, Stephen
1979
Levinson, Stephen
1983
1989
What is a sign? In: Peirce Edition Project, The essential Peirce: selected philosophical writings, Volume 2 (18931913), 410. Bloomington, Indiana: University Press.
Rees, M. Agnes van
1991
The adequacy of speech act theory for explaining conversational phenomena:
A response to some conversation analytical critics. Journal of Pragmatics 17:
3147.
Sacks, Harvey
1995
Lectures on conversation: Volumes I and II. Ed. by Gail Jefferson, with an introduction by Emanuel A. Schegloff. Oxford: Blackwell.
1974
A simplest systematics for the organization of turn-taking for conversation.
Language 50: 696735.
Sbis, Marina
2009a
Speech act theory. In: Jef Verschueren and Jan-Ola stman (eds.), Key Notions
for Pragmatics, 229244. Amsterdam: Benjamins.
Sbis, Marina
2009b
Uptake and conventionality in illocution. Lodz Papers in Pragmatics 5: 3352.
1968
Sequencing in conversational openings. American Anthropologist 70: 1075
1095.
Schegloff, Emanuel. A., Gail Jefferson and Harvey Sacks
1977
The preference for self correction in the organization of repair in conversation.
Searle, John R.
1969
Searle, John R.
1973
Austin on locutionary and illocutionary acts. In: Isaiah Berlin, Lynd W. Forguson, David F. Pears, George Pitcher, John R. Searle, Peter F. Strawson and
Geoffrey J. Warnock (eds.), Essays on J.L. Austin, 141159. Oxford: Oxford
University Press.
Searle, John R.
1975
Indirect Speech acts. In: Peter Cole and Jerrold Morgan (eds.), Syntax and
Speech acts
395
Searle, John R.
1979
Expression and meaning. Cambridge: Cambridge University Press.
Searle, John R.
2002
Speech acts, minds, and social reality. In: Gnther Grewendorf and Georg
Meggle (eds.), Speech acts, minds, and social reality, 317. Berlin: Springer.
Searle, John R.
2007
Illocutionary acts and the concept of truth. In: Dirk Greimann and Geo Siegwart (eds.), Truth and speech acts, 3140. New York: Routledge.
1975
Towards an analysis of discourse: the English used by teachers and pupils.
London: Oxford University Press.
Toobin, Jeffrey
2009
Supreme Flub, The New Yorker, Blog Post, January 20, 2009. Available online
at: http://www.newyorker.com/online/blogs/tny/2009/01/supreme-flub.html
Warnock, Geoffrey J.
1969
John Langshaw Austin, a biographical sketch. In: Kuang Tih Fann (ed.), Symposium on J.L. Austin, 321. London: Routledge and Kegan Paul.
1973
Some types of performative utterances. In: Isaiah Berlin, Lynd W. Forguson,
David F. Pears, George Pitcher, John R. Searle, Peter F. Strawson and Geoffrey
J. Warnock (eds.), Essays on J.L. Austin, 6989. Oxford: Oxford University
Press.
1922
Tractatus logico-philosophicus. London: Routledge and Kegan Paul.
1953
14.
Types of inference: entailment, presupposition,

and implicature1
Yan Huang
This article is dedicated to Professor Nigel Vincent, my mentor at
Cambridge, on the occasion of his sixty-fifth birthday.
Introduction
As one of the basic forms of reasoning, inference can in general be defined as a process of accepting a statement or proposition (called the conclusion) on the basis of
the (possibly provisional) acceptance of one or more other statements or propositions (called the premises). Construed thus, it includes deduction, induction, and
abduction. It also includes entailment,2 presupposition, and implicature.
The aim of this article is to assess entailment, presupposition, and implicature
the three fundamental types of inference that are recognised in semantics and pragmatics (see e.g. Huang 2007, 2009g on the scope of pragmatics). Of the three categories of inference, entailment is essentially semantic in nature, though the notion
of pragmatic entailment will be discussed. Presupposition straddles the semanticspragmatics boundary, resulting in semantic and pragmatic presupposition. Finally,
with regard to the two breeds of implicature, while conversational implicature is
pragmatic in nature, conventional implicature can be categorised either as semantic or pragmatic, depending on how the distinction between semantics and pragmatics is drawn. Concepts parallel to conversational implicature such as explicature and impliciture will also be considered. Finally, a brief comparison among
entailment, presupposition, conversational implicature, and conventional implicature in terms of four parameters: defeasibility, negation, conventionality and semantics versus pragmatics will be made.
1.
Entailment
The term entailment is derived from formal logic.3 It refers to a semantic relation
that can be defined in terms of truth in (1).
(1) Entailment
A proposition (or sentence expressing a proposition) p entails
a proposition (or sentence expressing a proposition) q if and only
if the truth of p guarantees the truth of q.
398
Yan Huang
What (1) basically says is this: if p entails q, then q can be taken as following logically and inescapably from p. If p is true, then q is also true. By contrast, if p is
false, nothing is said about the truth value of q. Defined thus, entailment represents
a truth-functional relationship in the sense that its function is to predict the truth
value of a proposition from what is known of the truth value of another. By way of
illustration, let us consider (2). (I use | to stand for entail.)
(2) a. All of the universitys professors are hard-working.
b. | Some of the universitys professors are hard-working.
In (2), if the university in question has professors, then (2a) entails (2b), but not
vice versa. The reason is that if it is true that all of the universitys professors are
hard-working, then there is no way to avoid the conclusion that it is also true that
some of the universitys professors are hard-working. Notice that entailment proceeds from our knowledge of language. It depends on the constituents of relevant
sentences rather than context. Consequently, an entailment is not defeasible, that
is, it cannot evaporate in any linguistic or non-linguistic context. This is why it
must be seen as semantic in nature (see e.g. Huang 2007: 1617).
However, on Fauconniers (1975) and Israels (2004) view, there is also pragmatic entailment. By pragmatic entailment is meant an inference that is on the one
hand default, and on the other defeasible. Two examples are given in (3) and (4).
(3) a.
b.
(4) a.
b.
John can run 100m in 9.9 seconds.

| John can run a slower 100m.
John cant run 100m in 9.9 seconds.
| John cant run a faster 100m.
According to Fauconnier and Israel, the opposing, scalar inferences of (3b) and
(4b) are pragmatic entailments. They are entailments because the inferences are
default, i.e. automatically valid; they are pragmatic because the inferences can be
defeated. But how the inferences can be cancelled seems to me to be unclear. One
type of pragmatic entailment of this sort is scalar entailment stemming from what
Levinson (2000) called a Hirschberg-scale a contextually given ad hoc scale. The
relation between the ordered elements in such a scale does not need to be that of
semantic entailment. The scale can be based on any partially ordered contrast sets
in a contextually salient way. Examples include destinations ordered in terms of
their distance, autographs ordered in terms of their authors famousness, and problems ordered in terms of their complexity (see e.g. Huang 2009d for further discussion).
Semantic entailment plays a prominent role in semantics and especially in formal semantics. It is explored to define many of the other fundamental semantic relations. For example, two sentences may be said to have a propositional equivalence, to be paraphrases of each other, or to be synonymous if and only if they have
exactly the same set of entailments, as in (5). Next, two sentences may be con-
399
sidered to be contradictories if each entails the negation of the other so that whenever one is true the other must be false, as in (6). In a similar vein, a single sentence
may be said to be a contradiction if it contains contradictory entailments, as in (7).
While contradiction is a semantic relation in which one member of the pair must be
true and the other false, contrariety is a semantic relation in which only one term
may be true, though both terms may be simultaneously false. Consequently, two
sentences S1 and S2 may be taken to have contrary propositions if and only if S1 entails the negation of S2 and the negation of S2 does not entail S1. This is exemplified
in (8). Furthermore, semantic entailment is also used to define notions like analyticity and ambiguity.
(5) (Propositional equivalence)
a. The little hut was hidden by the trees.
b. The little hut was concealed by the trees.
(6) (Contradictory)
a. No one likes dark tourism.
b. At least someone likes dark tourism.
(7) (Contradiction)
?John isnt married, but his wife is a feminist.
(8) (Contrariety)
a. This skirt is blue.
b. This skirt is red.
In formal semantics, semantic entailment is employed to advance a theory of generalised quantifiers. In this theory, which is developed out of the work of Ladusaw
(1979), the concept of directional entailingness (also called monotonicity) is invoked. Directional entailingness provides an account of patterns of semantic entailment between sets and subsets. Upward entailing or entailment (also called
monotone increasing, usually represented as MONF) involves entailment from a
subset to a set. In other words, the direction of entailment is from more specific to
less specific. By contrast, downward entailing or entailment (also called monotone
decreasing, normally represented as MONG) is concerned with entailment from a
set to a subset. Another way of saying it is that the direction of downward entailment is from less specific to more specific. Now, contrast (9) and (10).
(9) a.
b.
(10) a.
b.
Every woman is cooking.

Every woman is roasting beef.
No woman is cooking.
No woman is roasting beef.
In (9), (9a) does not entail (9b), but (9b) does entail (9a). Consequently, the quantifier every triggers an upward entailment here.4 By contrast, in (10) the reverse is
true. Therefore, the quantifier no involves a downward entailment. The distinction
between upward and downward entailment is applied particularly to the study of
400
Yan Huang
negative polarity items (NPIs). It predicts correctly that NPIs like any and ever can
be licensed in the scope of a downward entailment operator but not in that of an upward entailment one (see e.g. Saeed 2009). The contrast is shown in (11).
(11) a. ?Every woman is ever roasting beef.
b. No woman is ever roasting beef.
But there has also been evidence that downward entailment is neither necessary
nor sufficient for the licensing of NPIs (e.g. Giannakidou 1998, Horn 2009, see
also Atlas 2007).
More recently, Chierchia, Crain and their associates (e.g. Chierchia 2004,
Crain and Pietroski 2002) have used the upward/downward entailment distinction
to mount a challenge to what Horn (2006) dubbed a Golden Age Pure Pragmatics
or GAPP-style analysis of scalar implicatures. According to Chierchia, Crain, and
their associates, while a standard upper bounding Q-scalar implicature, that is, an implicature from p to at most p , does arise from positive Horn-scales,
it is quite weak and even blocked in negative Horn-scales and other downward entailing environments. On the basis of this claim, Chierchia, Crain, and their associates argued that Q-scalar implicatures must be computed compositionally. Consequently, they fall under compositional semantics, hence part of grammar. But as
pointed out by Levinson (2000) and Horn (2006, 2009), the alleged blockage of
Q-scalar implicatures is due to the fact that a Horn-scale is reversed under negation
and other downward entailing operators, as in (12a), and consequently, a different
Q-scalar implicature is derived from the inverse scale, as in (12c) (see also Huang
2009b). (I use +> to signify conversationally implicate.) If this is the case, then
Chierchia and Crains argument may not be maintained (see also Sauerland 2004,
Russell 2006, and Geurts 2009 for various (neo)-Gricean analyses of Q-scalar implicature in downward entailment contexts).
(12) a. <not some, not many, not most, not all>
b. The earthquake didnt kill many of the villagers.
c. +> The earthquake killed some of the villagers.
There are, however, serious problems at the very heart of the (semantic) entailment
based approach to meaning. One such problem, which is well-known, is concerned
with the failure of a semantic entailment in a referentially opaque context. Consider first (13).
(13) a. Barack Obama is a popular politician.
b. The first African-American President of the United States of America is a
popular politician.
At the time of writing (April, 2009), the proper name Barack Obama in (13a) and
the definite description the first African-American President of the United States of
America in (13b) have the same extension, i.e. they refer to the same person. Given
401
Leibnizs Law, which is a law of intersubstitutability salva veritate (Latin for with
the truth unchanged), it states that the substitution of expressions with the same
extension does not affect the truth-conditions of the sentences. Consequently, if
(13a) expresses a true proposition, then so does (13b). In other words, (13a) and
(13b) are synonymous because they mutually entail each other. However, in a referentially opaque context, truth cannot be preserved when the co-referential expressions are substituted for each other. This is the case in (14).
(14) a. I want to dine with Barack Obama.
b. I want to dine with the first African-American President of the United
States of America.
Here, (14a) and (14b) have different truth conditions. This is because the expression the first African-American President of the United States of America is
given an intensional reading, that is, the reading that the speaker wants to dine with
whoever happens to be the first African-American President of the United States of
American and does not care and may not have known who that person is. The same
can be said of (15).
(15) a. John believes that Barack Obama is a popular politician.
b. John believes that the first African-American President of the United
States of America is a popular politician.
If John does not know or believe that Barack Obama is the first African-American
President of the United States of America, then (15b) means that John has a belief
about someone who is the first African-American President of the United States of
America without knowing or caring about who the actual holder of the office is.
Consequently, such a belief of Johns is said to be de dicto a belief about what is
said. The same is true of Freges classical example, usually referred to as the
Morning Star Paradox (see e.g. Cann 1993 for discussion about further problems of
the (semantic) entailment based analysis). Verbs that create a referentially opaque
context are commonly referred to as intensional verbs or verbs of propositional attitude. These include believe, know, doubt, want, hope, fear, look, seem, and seek.
Other words that establish a referential opaque context include adjectives like alleged, prepositions like about, connectives like because, and modal words of various categories like must, may, probably, obviously, and permissible.
2.
Presupposition
2.1.
Defining presupposition
Presupposition is defined as a proposition or inference whose truth is taken for

granted in the utterance of a sentence.5 Its main function is to act as a precondition
402
Yan Huang
of some sort for the appropriate use of that sentence. This background assumption
will remain in force when the sentence that contains it is negated. The German
mathematician and logician Gottlob Frege is generally recognised as the first
scholar in modern times who (re)introduced the philosophical study of presupposition (see especially Frege 1892), though the notion of presupposition may go back
at least as far as the mediaeval philosopher Petrus Hispanus (see e.g. Huang 2007).
Presupposition is usually engendered by the use of particular lexical items
and/or linguistic constructions. Lexical items and linguistic constructions that give
rise to presuppositions are called presupposition triggers. While presupposition
has in general been regarded as a heterogeneous collection of quite distinct and
different phenomena (Levinson 1983: 217) since Karttunen (1973), a list of its
representative examples may be given in (16) (23) (see e.g. Levinson 1983, Atlas
2005, Huang 2007). (I use the symbol >> to stand for presuppose. The positive
and negative versions of the examples are separated by /, and the lexical presupposition triggers are italicised.)
(16) Definite descriptions
The king of France is/isnt bald.
>> There is a king of France.
(17) Factive predicates
a. Epistemic or cognitive factives.
John knows/doesnt know that smoking is a dangerous pastime.
>> Smoking is a dangerous pastime.
b. Emotive factives
John regrets/doesnt regret that he has gone on the stage.
>> John has gone on the stage.
(18) Aspectual/change of state predicates
Mary has/hasnt stopped beating her boyfriend.
>> Mary has been beating her boyfriend.
(19) Iteratives
John returned/didnt return to Berlin.
>> John was in Berlin before.
(20) Implicative predicates
John managed/didnt manage to give up binge drinking.
>> John tried to give up binge drinking.
(21) Temporal clauses
After she left school, Jane worked/didnt work as a secretary.
>> Jane left school.
(22) Cleft sentences
a. Cleft
It was/wasnt the porter who called a taxi for John.
>> Someone called a taxi for John.
403
b. Pseudo-cleft
What the porter called/didnt call was a taxi.
>> The porter called something.
(23) Counterfactual conditionals
If an ant was as big as a human being, it could/couldnt run five times faster
than an Olympic sprinter.
>> An ant is not as big as a human being.
2.2.
Properties of presupposition
Presuppositions exhibit two main properties: (i) constancy under negation, and (ii)
defeasibility. In addition, certain cases of defeasibility give rise to what is known
as the projection problem of presupposition.
2.2.1.
Constancy under negation
By constancy under negation is meant that a presupposition generated by the use of

a lexical item or a syntactic structure remains the same when the sentence containing that lexical item or syntactic structure is negated (e.g. Strawson 1952). Using
constancy under negation as diagnostic, presuppositions can be more formally defined in (24).
(24) An utterance of a sentence S presupposes a proposition p if and only if
a. if S is true, then p is true;
b. if S is false, then p is (still) true.
What (24) basically says is this: for S to presuppose p, whenever S is true, p is also
true, and whenever S is false, p is still true.
There are, however, problems with constancy under negation. On the one hand,
constancy under negation may not be necessary. For example, there is a class of
sentences which are hard, if not impossible to negate, yet they bear presuppositions, as in (25). On the other hand, constancy under negation may not be sufficient. This is illustrated by (26). Although (26) satisfies constancy under negation,
the inference is standardly analysed as a felicity condition on performing the
speech act of requesting (e.g. Levinson 1983). This is because if the window is
open, there is no need for one to perform the speech act of requesting someone else,
for example, to open it.
(25) Long live the king of France!
>> There is a king of France
(26) Do/dont open the window.
?>> The window is not open
404
2.2.2.
Yan Huang
Defeasibility
Presuppositions are deafeasible. They are cancelled if they are inconsistent with (i)
background assumptions, (ii) conversational implicatures, and (iii) certain discourse contexts. Furthermore, they can also disappear in certain intrasentential
contexts, some of which give rise to the projection problem of presupposition. Defeasibility has in general been taken as the second most important property of presupposition.
In the first place, presuppositions can disappear in the face of inconsistency
with background assumptions or real-world knowledge. Contrast (27) and (28) (I
use ~ >> to stand for does not presuppose.)
(27) John got an assistant professorship before he finished his Ph.D.
>> John finished his Ph.D.
(28) John died before he finished his Ph.D.
~ >> John finished his Ph.D.
While (27) presupposes that John finished his Ph.D. by virtue of the temporal
clause, (28) does not carry that presupposition. This is because the putative presupposition conflicts with our real-world knowledge. Consequently, the unwanted presupposition vanishes.
Secondly, presuppositions can be cancelled by inconsistent conversational implicatures. By way of illustration, consider (29).
(29) If Susan has danced on the table, her mother will be unhappy that she has
done so.
+> perhaps Susan has danced on the table, perhaps she hasnt
~ >> Susan has danced on the table
The use of the factive predicate unhappy in (29) should give rise to the potential
presupposition that Susan has danced on the table. However, there is also a Q-clausal conversational implicature due to the use of the conditional in (29), namely,
perhaps Susan has not danced on the table. In the face of the contradictory conversational implicature, the putative presupposition is defeated.
Thirdly, presuppositions are contextually cancellable, that is, they can evaporate if they run contrary to what the immediate discourse context tells us. Witness
first (30), where the putative presupposition is defeated by the inconsistent proposition that has already been established in the immediate discourse background.
(30) There is no king of France. Therefore, the king of France isnt bald.
~ >> There is a king of France
In (30), the second sentence should presuppose that there is a king of France. The
reason it does not is because such a putative presupposition is inconsistent with the
preceding proposition that has already been established in the immediate discourse
405
background. As a result, the unwanted presupposition fails to survive. Next, presuppositions can be suspended by so-called reduction arguments arguments that
proceed by eliminating each of the possibilities in a discourse. Finally, presuppositions can disappear in a discourse where evidence for their truth is being
weighed and rejected.
In the fourth place, presuppositions can be blocked in certain intrasentential
contexts, that is, they can be defeated by using other clauses in the same complex
sentence to increment the local, intrasentential context. Three sub-cases are of particular interest. The first is that a presupposition of a sentence can be overtly denied
in a co-ordinate clause without any apparent contradiction. This is exemplified in
(31).
(31) John doesnt regret snoring atrociously because in fact he never did so!
~ >> John snored atrociously
Notice that in many cases, outright denial of presupposition is not possible with
positive sentences.
(32)*John regrets snoring atrociously because in fact he never did so!
There is thus, at least in these cases, an asymmetry between negative and positive
sentences with regard to defeasibility. This asymmetry has led to an entailment
analysis for positive sentences, namely, the argument that what is allegedly presupposed in these sentences is actually what is entailed (e.g. Atlas 2005). Since semantic entailments cannot be overtly denied without producing semantic anomaly,
the anomaly displayed in (32) is entirely expected. A second point to note is that
the negation involved in overt denial of presupposition in (31) is generally taken as
a metalinguistic negation (see e.g. Horn 2006). Next, a presupposition of a sentence can be explicitly suspended in an if clause that follows. Witness (33).
(33) Mary doesnt regret being a pet sitter, if she actually ever was one.
~ >> Mary was a pet sitter
Finally, presuppositions can disappear with certain verbs of saying such as say,
mention, tell, ask and announce, and certain verbs of propositional attitude such as
believe, think, imagine, dream and want (see e.g. Huang 2007: 7273).
2.2.3.
Theoretical issues and analyses
In general, there are three main theoretical issues in accounting for presupposition.
The first and oldest one is concerned with presupposition failure. If a statement
presupposes something which does not exist, what are the consequences for that
statement? As already mentioned, a much quoted example of statements of this
kind is the positive version of (16). Given that there is no king of France, the question boils down to this: what is the truth status of the sentence? Russells (1905)
406
Yan Huang
view was that since the sentence asserts both that there is a king of France and that
he is bald, if the king of France does not exist, the sentence is false. This analysis
was widely accepted and remained practically unchallenged for almost half a century. Then in 1950, Strawson published a seminal paper (Strawson 1950) in which
he presented an alternative account. According to this analysis, Russell failed to
distinguish sentences from the use of sentences. In using the sentence, a speaker
does not assert but merely presupposes that there is a king of France. If there is no
king of France, then there is a presupposition failure. Consequently, the sentence is
neither true nor false: there is simply a truth-value gap. Strawsons account has led
to the notion of semantic presupposition.
Secondly, there is the question of the origin of presupposition, namely, where
presuppositions come from. The standard, though imperfect answer to this question is that presuppositions are engendered by certain lexical and structural
triggers. Thirdly and finally, the most important issue is concerned with the projection problem of presupposition a special case of the Fregean compositionality,
namely, to state and explain the presuppositions of complex sentences (as
wholes) in terms of the presuppositions of their component simple sentences (as
parts). The projection problem manifests itself in two opposite directions. On the
one hand, the presuppositions of a component sentence may fail to be projected
onto, and hence inherited by, the whole complex sentence. This is the case for e.g.
(31) and (33). In addition, certain complex or compound sentences formed with
if then and either or also belong to this category. On the other hand, the presuppositions of a component sentence may be preserved when that constituent sentence becomes part of a more complex sentence. This is what happens when the
sentence containing a presupposition trigger is under negation, as in (16), when it
is embedded in modal operators such as its possible that, its likely that, ought to,
may be and should, and when it is embedded in the conditional and the disjunction
formed with if then and either or.
Three formal semantic-pragmatic models are particularly influential in tackling
the projection problem. The first is what I called the filtering-satisfaction (or Karttunen-Stalnaker) analysis in Huang (2007). The central idea of this approach is that
a presupposition is not cancellable. Rather it has to be satisfied (i.e. entailed) in its
local context, which is conceived mainly as a set of propositions. The local context
is constructed in a dynamic way and developing online. This makes it possible for
an unwanted presupposition to be filtered out during the derivation of a sentence in
a bottom-up manner (see e.g. Karttunen 1973, Stalnaker 1974). Stalnakers work
has resulted in the development of the concept of pragmatic presupposition. Next,
in contrast to the filtering-satisfaction analysis, underlying Gazdars (1979) cancellation analysis is the crucial assumption that a presupposition is cancellable. On
this theory, what a presupposition trigger engenders is merely a potential presupposition. A potential presupposition will become an actual one, unless it is defeated. With respect to the projection problem, what the cancellation analysis pre-
407
dicts is that each and every presupposition of the embedded clause will become an
actual presupposition of the complex sentence, unless it is nullified by certain linguistic and non-linguistic factors. Finally, in an attempt to combine the strengths of
both the filtering-satisfaction and cancellation theories, Heim (1983) developed a
particular version of the filtering-satisfaction model, couched in her dynamic semantic theory of context change. In Huang (2007), I dubbed this model the accommodation analysis. Central to this approach is Heims belief that the meaning of an
expression, including the presupposition of a sentence, is its context change potential. The context change potential of a sentence is the intersection of any context set
in a common ground together with the proposition expressed by the sentence. By
common ground is meant a background of beliefs or assumptions which are
shared by the speaker and his audience and which are recognised by them to be so
shared (Stalnaker 1974). Accommodation, then, provides a mechanism to increment the discourse context set with new, non-controversial assumptions.
Of these analyses, Karttunens and Gazdars are representatives of the twocomponent model view of presupposition, according to which, sentences including
those containing presuppositions have two types of content: semantic content and
presuppositional content. The latter is pragmatic in character. By contrast, Stalnaker and Heim take the one-component model view, operating within the general
framework of a theory of pragmatic presupposition. On this account, presuppositions are treated as conditions on whether a sentence can be admitted into a context
(see e.g. Portner 2005).
3.
Implicature
The concept of implicature (both conversational and conventional) has its origin in
the work of H. P. Grice, though some proto-Gricean ideas can be traced back at
least to the first century B.C. rhetorician Dionysius and the fourth century rhetoricians Servius and Donatus (e.g. de Jonge 2001). The ideas were later reiterated by
the nineteenth century English philosophers John Stuart Mill and Augustus De
Morgan (e.g. Horn 2006).
3.1.
Conversational implicature
A conversational implicature is any meaning implied or expressed by, and inferred

or understood from, the utterance of a sentence which is meant without being
part of what is strictly said6 (e.g. Huang 2009a, b, f). It is derived via Grices
(1975, 1989) cooperative principle and its attendant maxims of conversation. For
example, when one utters the sentence in (34a), he or she (ceteris paribus) conversationally implicates (34b).
408
Yan Huang
(34) a. Some of the tourists are admiring the view.

b. +> Not many/most/all of the tourists are admiring the view
Conversational implicatures are characterised by a number of distinctive properties (Grice 1975, 1989, Levinson 2000, Huang 2007: 3235, 2009a, b). In the first
place, there is the property of defeasibility conversational implicatures can
simply vanish in certain linguistic or non-linguistic contexts. How? They are cancelled if they are inconsistent with (i) semantic entailments, (ii) background or ontological assumptions, (iii) contexts, or (iv) priority conversational implicatures.
As an illustrating example, consider (35) and (36). (The symbol ~ +> is used to
stand for do not conversationally implicate.)
(35) John and Mary bought a villa in Auckland.
+> John and Mary bought a villa in Auckland together
(36) The Americans and the Russians tested an atom bomb in 1962.
~ +> The Americans and the Russians tested an atom bomb in 1962 together
The utterance in (36) has the potential mirror maxim, togetherness conversational implicature, as indicated in (35). However, this potential conversational implicature runs contrary to our background or ontological assumptions. Given our
knowledge about history, it was impossible for the USA and the USSR to test an
atom bomb together in 1962. Consequently, the potential togetherness conversational implicature is defeated by inconsistent real-world knowledge. Notice that
defeasibility is a necessary but not a sufficient condition for conversational implicature (Horn 2007).
A second property exhibited by conversational implicatures is non-detachability any linguistic expression with the same semantic content tends to carry
the same conversational implicature. (A principled exception is those conversational implicatures that arise via the maxim of Manner.) This is because conversational implicatures are attached to the semantic content, rather than the
linguistic form, of what is said. Therefore, they cannot be detached from an utterance simply by replacing the relevant linguistic expressions with their synonyms.
This is illustrated in (37), which indicates that the use of any linguistic expression
that is synonymous with almost will give rise to the same conversational implicature.
(37) Gordon almost/nearly lost /came close to losing his job.
+> Gordon did not (quite) lose his job
Thirdly, we have the property of calculability conversational implicatures can
transparently be derived via the cooperative principle and its component maxims.
A fourth property is non-conventionality conversational implicatures, though dependent on the saying of what is coded, are non-coded in nature. In other words,
they rely on the saying of what is said but they are not part of what is said. A fifth
409
property is reinforceability conversational implicatures can be made explicit

without producing too much of a sense of redundancy. This is because conversational implicatures are not part of the conventional import of an utterance. For instance, the conversational implicature in (34) is made explicit in (38). But (38) is
not judged to be semantically redundant.
(38) Some, but not many/most/all, of the tourists are admiring the view.
The sixth property of conversational implicatures is universality conversational
implicatures tend to be universal, because they are motivated rather than arbitrary.
For example, if a language has all and some, the use of the semantically weaker
some will universally carry the conversational implicature not all. In Huang
(2007: 35), data are drawn from Modern Standard Arabic, Catalan, Chinese, Modern Greek, Kashmiri and Malagasy to illustrate this point. Finally, there is the property of indeterminacy some conversational implicatures may be indeterminate.
They can be taken as conveying an open-ended range of implicaures relating to
matters in hand. Suppose one says (39). He or she may generate a range of indeterminate conversational implicatures.
(39) Our new professor is a machine.
+> Our new professor is cold, or/and
+> Our new professor is efficient, or/and
+> Our new professor is a workaholic, or/and
+>
A conversational implicature can be engendered in two distinct ways. On the one
hand, it can arise from strictly observing the maxims of conversation. In Huang
(2007: 27), I called conversational implicatures thus induced conversational implicaturesO. This is the case for e.g. (34), (35) and (37) above. On the other hand, a
conversational implicature can be generated by way of a speakers ostentatiously
flouting the maxims. In Huang (2007: 29), I dubbed conversational implicatures
thus engendered conversational implicaturesF. This is the case with the generation
of the conversational implicature in (39), which deliberately exploits Grices
maxim of Quality. There is thus the first Gricean dichotomy between conversational implicatureO and conversational implicatureF. Grices achievement here was
to have provided a unified analysis of both types.
A second Gricean dichotomy, independent of the first, is between those conversational implicatures which arise without requiring any particular contextual conditions and those which do require such conditions. Grice called the first kind generalised conversational implicatures (GCIs) and the second kind particularised
conversational implicatures (PCIs). By way of illustration, consider the two conversational implicatures in Marys utterance in (40).
410
Yan Huang
(40) John: How did yesterdays guest lecture go?

Mary: Some of the faculty left before it ended.
+> (a) Not many/most/all of the faculty left before the lecture ended
+> (b) The lecture didnt go well
The conversational implicature in (40a) has a very general currency. Any utterance
of the form Some x are Y will have the default interpretation Not many/most/all
x are Y. This interpretation will go through without needing any particular context, hence (40a) is a GCI. By contrast, the conversational implicature in (40b) depends crucially on context of some kind. Marys reply points to a possible connection, namely, if some of the faculty left a lecture before it ended, the lecture may
not have gone well. Without such a specific connection, we will not have the relevant conversational implicature, thus (40b) is a PCI. The theoretical importance
of this Gricean dichotomy has recently been subject to heated debates. Hirschberg
(1991), Welker (1994) and Carston (2002), for example, doubted whether such a
distinction can be maintained. On the other hand, Levinson (2000) put forward a
rigorous defence of it.
Since its inception, Grices classical theory of conversational implicature has
revolutionised pragmatic theorising, nourishing numerous neo- and post-Gricean
variants and reformulations. Horn developed a bipartite model calling it Manichaean pragmatics. In this account, there are two genera of conversational implicatures: namely Q[uantity]- and R[elation]-implicatures (see e.g. Horn 2009). Arguing for a clear separation between pragmatic principles governing an utterances
surface form and pragmatic principles governing its informational content, Levinson (2000) proposed that the original Gricean program (the maxim of Quality
apart) be reduced to a tripartite model with three genera of conversational implicatures: what he dubbed the Q-, I[nformativeness]- and M[anner] implicatures.
Q-implicatures can then be divided into three species: what I termed (i) Q-scalar , (ii)
Q-clausal and (iii) Q-alternate implicatures in Huang (2007: 4244). Furthermore, within
Q-alternate implicatures, we have two subtypes: Q-ordered alternate and Q-unordered alternate implicatures such as <bake, boil, grill, steam, stir-fry>. Notice that Q-ordered alternate implicatures can arise either from a non-entailment scale which is given by the lexicon without requiring any specific context (e.g. <succeed, try>) or from what
Levinson (2000) called a Hirschberg-scale or what Horn (2007, 2009) dubbed a
rank order, following in part work by Lehrer (1974). As already mentioned above,
a Hirschberg-scale is essentially a nonce scale, that is, a contextually given ad hoc
scale. Such a scale can be based on any partially ordered contrast sets in a contextually salient way. Examples include <divorce, separate>, <Barack Obamas autograph, Hillary Clintons autograph>, and <full professor, associate professor, assistant professor> (see e.g. Huang 2009d for further discussion). On the other
hand, in relevance theory (Sperber and Wilson 1995), the majority of classical and
neo-Gricean GCIs is refashioned as an explicature a proposition that is an infer-
411
ential development of one of the linguistically given incomplete conceptual representations or logical forms of the sentence uttered (but see e.g. Burton-Roberts
2007 for the comment that development in the definition is not defined in relevance theory). Explicature corresponds roughly to the intuitive notion of what is
said, though the Gricean notion of what is said is abandoned in relevance theory.
Conversational implicatures in the relevance-theoretic sense, called r-implicatures
in Huang (2007: 195), are largely PCIs in the classical and neo-Gricean sense. In a
similar vein, Recanati (2004), a radical contextualist in the current divide between
contextualism and (semantic) minimalism in the philosophy of language, reduced
some cases of conversational implicature to what he called the pragmatically enriched said. Also, Bach (2004) argued that certain aspects of speaker meaning are
neither part of what is said nor of what is conversationally implicated. Consequently, he proposed a third category of communicative content, intermediate between what is said and what is conversationally implicated, and dubbed the vehicle
of such a content conversational impliciture, because it is what is implicit in what
is said. In other words, on Bachs view, some cases of conversational implicature in
the classical and neo-Gricean sense are in fact (conversational) implicitures (see
also Bach 2006, Huang 2009e and Garret and Harnish 2007).
At this point, it is useful to discuss some of the current debates on conversational implicature. In the first place, the treatment of cardinals as scalar expressions
engendering Q-scalar implicatures is highly controversial (see e.g. Carston 2002,
Bultinick 2005, Hurewitz et al. 2006, Horn 2009 for arguments against, and Levinson 2000 for arguments for the scalar analysis). Secondly, there is the issue of the
epistemic strength of Q-scalar implicatures. It is concerned with the question of
what it is a speaker Q-scalar implicates against. Two neo-Gricean pragmatic positions can be identified here: the weak epistemic one represented by e.g. Hirschberg
(1991), Sauerland (2004), Geurts (2009), and Horn (2009), and the strong epistemic one advocated by e.g. Gazdar (1979) and Levinson (2000: 7779). In the
third place, a heated debate has been going on for the last two decades or so, focusing on the nature of Gricean and neo-Gricean GCIs in general and Q-scalar implicatures in particular. One view is that GCIs in general and Q-scalar implicatures
in particular convey default meanings, sans a conscious inferential process and irrespective of a particular context (e.g. see Levinson 2000 for strong defaultism and
Horn 2009 for weak defaultism). Another, relevance-theoretic position is that they
are essentially inferred contextually (e.g. Sperber and Wilson 1995). In addition, a
more recent, third view holds that the derivation of Q-scalar implicatures relies
heavily on structural factors (cf. Section 2 above and Chierchia 2004). Furthermore, all the three views have recently been subject to studies in experimental
pragmatics. While much of the relevance-theoretically oriented experimental work
favours the contextual inference approach (e.g. Noveck and Sperber 2007, Noveck
and Reboul 2008), there is also evidence in support of the default inference theory
(e.g. Grodner et al. 2007) and the structural inference view (e.g. Panizza and Chier-
412
Yan Huang
chia 2008). Next, in recent years, there has also been an intense debate on (i)
whether or not pragmatically enriched or inferred content can enter or intrude
upon the conventional, truth-conditional content of what is said, and (ii) if the
answer to (i) is positive, then what the pragmatic intrusion under consideration is.
Concerning the first question, a dividing line can generally be drawn between pragmaticists like Levinson (2000), Recanati (2004) and relevance theorists (Sperber
and Wilson 1995, Carston 2002), who argue for pragmatic intrusion (but see e.g.
Horn 2009 for reservations), and semanticists like Cappelen and Lepore (2005)
and King and Stanley (2005), who argue against it. Next, regarding the second
question, two current positions can roughly be identified. The first is that pragmatic
intrusion is of a special kind, which differs from conversational implicature.
Within this camp, three lines of argument are of particular interest. According to
Sperber and Wilson (1995), the pragmatic content is an explicature. Secondly,
there is the position taken by Recanati (2004) that it is a pragmatically enriched
part of what is said. A third argument is due to Bach (2004), who took the view that
the pragmatic inference under consideration is his impliciture. On the other hand,
the second position is represented by Levinson (2000). Within the neo-Gricean
framework, Levinson argued that these so-called explicatures/pragmatically enriched said/implicitures are in fact the same beast as conversational implicatures.
See also my neo-Gricean and revised neo-Gricean analyses of anaphora in e.g.
Huang (1991, 2000a, b, 2004, 2007 and 2009b), which in effect argue that the presemantic, pragmatic inference involved in the interpretation of anaphora is a conversational implicature.
Finally, it is worth mentioning that there have been various attempts to integrate the classical and neo-Gricean pragmatic theories of conversational implicature with other current linguistic theories. These linguistic theories include decision theory (Merin 1999), bidirectional optimality theory (Blutner 2004) and game
theory (Benz et al. 2006).
3.2.
Conventional implicature
We turn next to the second category of implicature postulated by Grice, namely,

conventional implicature. (In fact, Freges 1892, 19181919 analysis of the Andeutung relation is a direct precursor of Grices concept of conventional implicature.
See e.g. Bach 1999, Feng 2006, Horn 2007). An Andeutung or conventional implicature is a non-truth-conditional meaning which is not derivable from general considerations of cooperation and rationality, but arises solely from the conventional
features attached to particular lexical items and/or linguistic constructions involved. A few standard examples are given in (41)-(43). (I use +>> to stand for
conventionally implicate.)
413
(41) p therefore q +>> q follows from p

Lihua is a Chinese; she, therefore, knows how to use chopsticks.
(42) p but q +>> p contrasts with q
John is poor but he is honest.
(43) Even p +>> contrary to expectation
Even a child can scoff all the small bananas.
In (41), the conventional implicature triggered by the use of therefore is that being
a Chinese provides some good reason for knowing how to use chopsticks. In (42),
there is a conventional implicature of contrast between the proposition contained
in p and that contained in q. Finally in (43), even, being epistemic in nature, conventionally implicates some sort of unexpectedness, surprise or unlikeliness (Farncescotti 1995). Other representative lexical items that are considered to engender
conventional implicatures in English include actually, also, anyway, barely, besides, however, manage to, moreover, on the other hand, only, still, so, though, too
and yet.
What, then, are the essential properties of conventional implicature? From a
traditional point of view, conventional implicatures are considered to have the following properties (Grice 1989, Levinson 1983, Huang 2007, 2009c). Firstly, conventional implicatures are not derived from Grices cooperative principle and its
component maxims, but are attached by convention to particular lexical items
and/or linguistic constructions. They are therefore an arbitrary part of meaning,
and must be learned ad hoc. Secondly, conventional implicatures are not calculable
via any natural procedure, but are rather given by convention, thus they must be
stipulated. Thirdly, conventional implicatures are not defeasible, that is, they cannot be cancelled. Fourthly, conventional implicatures are detachable, because they
depend on the particular lexical expressions and/or linguistic constructions used.
Finally, conventional implicatures do not tend to be universal.
It should be pointed out that unlike the concept of conversational implicature,
the notion of conventional implicature is not taken to be very coherent. Even Grice
himself (1989: 46) warned that the nature of conventional implicature needs to be
examined before any free use of it, for explanatory purposes, can be indulged in.
Horn (2004: 6) has gone a step further by claiming that the role played by conventional implicature within the general theory of meaning is increasingly shaky.
Since its inception, conventional implicature has been subject to numerous attempts
to reduce it to semantic entailment, conversational implicature, and presupposition
(Levinson 1983), and more recently, to part of what is said (Bach 1999), part of tacit
performatives (Rieber 1997), vehicles for performing second-order speech acts
(Bach 1999), and procedural meaning in relevance theory (Blakemore 2004).
But recently, Potts (2005) has made a brave attempt to resurrect the concept of
conventional implicature. He retain[ed] Grices brand name but alter[ed] the product (Horn 2007) by focusing on expressives like epithets, attributive adjectives
414
Yan Huang
and honorifics and supplements like non-restrictive relatives, parentheticals and

appositives rather than lexical items such as but, therefore and even. He isolated
four essential properties of conventional implicature. The first of these properties
is conventionality conventional implicatures are part of the conventional meaning of the expressions involved. The second property is commitment conventional implicatures are commitments, and thus engenders entailments. The third
property is speaker orientation the commitments are made by the speaker of an
utterance. The final property is independence conventional implicatures are logically and compositionally independent of what is said (see also von Heusinger and
Turner 2006). Taking the view that conventional implicature is semantic in nature,
Potts developed a logic of the notion by modelling it with a type-driven multi-dimensional semantic translation language (see Feng 2006, Horn 2007 for criticisms
of this analysis). Feng (2006) presented another development of Grices notion of
conventional implicature. The properties extracted by him for conventional implicature are (i) non-truth-conditionality, (ii) speaker orientation, (iii) infallibility,
(iv) occurrency, (v) dependency, and (vi) context-sensitivity. He further argued that
properties (i) (iv) are intimately associated with subjectivity. Finally, contrary to
Pottss view, Feng maintained that conventional implicature has both a semantic
and pragmatic character. This is why a conventional implicature is so named by
Grice. [I]t is so named because it involves both linguistic and contextual information. It is conventional because it is associated with the conventional linguistic meaning of a certain expression It is implicated rather than said because
its full content requires contextual information, and does not affect the truth conditions of the utterance (Feng 2006: 184). This view is also echoed in Horn (2007:
50), who said that [conventional implicature] is semantic insofar as it involves an
aspect of the conventional meaning of a given expression rather than being computable from general principles of rational behavior or communicative competence, but it is pragmatic insofar as it involves considerations of appropriateness
rather than truth of the sentence in which it appears. Whether belonging to semantics or balancing on the edge between semantics and pragmatics, Pottss, Fengs
and Horns recent works have shown that the Fregeo-Gricean concept of conventional implicature is, after all, not that incoherent.
4.
A brief comparison
In this final section, I shall make a brief comparison among semantic entailment,
presupposition, conversational implicature and conventional implicature. I shall
compare them in terms of four parameters: (i) defeasiblility, (ii) negation, (iii) conventionality, and (iv) semantics versus pragmatics.
First, defeasibility. While semantic entailment and conventional implicature
are not defeasible, most cases of presupposition and all cases of conversational im-
415
plicature are defeasible. In the study of inference, a hierarchy is set up to rank various types of inference as defeasible, nonmonotonic, default, probabilistic, and
plausibilistic, with defeasible inference at the top. An inference is defeasible when
it has the possibility of error. Looked at in this way, nonmontonic inference is a
major type of defeasible inference. It is currently reasonable but its reasonability
can be cancelled upon the admittance of new information into common ground.
Default inference is then a particular kind of nonmonotonic inference. It is characterised by what computer scientists and artificial intelligence (AI) workers call negation as failure. An example might be that from the failure of the airport departures board to list a late-night flight to London Heathrow, one may infer that such a
flight has been cancelled. In this sense, both presupposition and conversational implicature are a case of presumptive reasoning like default inference. Probabilistic
inference is stronger than plausibilistic inference (see e.g. Woods 2009). Next, we
turn to negation. While presupposition can survive negation, entailment cannot. As
for conversational and conventional implicatures, their behaviour under negation
is rather complex and needs to be further studied. Finally, let us consider the question of conventionality and the related question of where in the linguistic terrain
the four types of inference are situated. Clearly, semantic entailment is the most
conventional and falls in the province of semantics. Presupposition is less conventional than conventional implicature but more conventional than conversational
implicature. Therefore, it straddles the boundary between semantics and pragmatics, but more on the side of pragmatics. Of the three types of implicature, although GCIs are more conventional than PCIs, both types are the pursuit of pragmatics. Finally, conventional implicature, which is more conventional than both
GCIs and PCIs, can be considered either as semantic or pragmatic, depending on
how the boundary between semantics and pragmatics is placed. If semantics is
taken to be concerned with those aspects of meaning that affect truth conditions,
then the investigation of conventional implicature falls on the pragmatic side of the
divide rather than on the semantic side, since, as noted above, it does not make any
contribution to truth conditions. On the other hand, if pragmatics is conceived of as
dealing with those inferences that are non-conventional, hence cancellable, then
conventional implicature falls within the province of semantics but outside that of
pragmatics, since it cannot be defeated.
Notes
1. I wish to thank the anonymous reviewers and Wolfram Bublitz and Neal Norrick, the two
editors of this volume for their stimulating comments on an earlier version of this chapter.
The work reported on here was partially supported by a grant from the University of
Auckland Faculty Research Development Funds, which is hereby gratefully acknowledged.
416
Yan Huang
2. On Saeeds (2009) view, entailment is not an inference. However, Cummings (2005:

8384) argued convincingly that such a view is too narrow.
3. The phrase p entails q appeared first in Moore (1922). The term entailment, according
to the second edition of the Oxford English Dictionary, was first used by L. S. Stebbing in
1933, who wrote: Professor Moores analysis of the distinction between material implication and entailment makes it possible.
4. Notice that whereas every is upward entailing with respect to its right, VP argument, as in
(9), it is downward entailing in its left, NP argument (e.g. Saeed 2009). For example,
Every woman is cooking unilaterally entails Every old woman is cooking. In terms of
monotone increasing/decreasing, the former is called right monotone increasing, and the
latter is labelled left monotone decreasing.
5 The term presupposition seems to be introduced into English by Strawson (1952).
6. Note that Saul (2002) was of the view that Grices main goal is to develop a theory
of speaker-meaning. Following Saul and biting the bullet, Horn (2009) now holds that
conversational implicature is a component of speaker-meaning rather than a pragmatic
inference. By contrast, Levinson (2000), Atlas (2005) and many others are still treating
conversational implicature as a pragmatic inference. My definition is applicable to both
sides.
References
Atlas, Jay D.
2005
Logic, Meaning, and Conversation. Oxford: Oxford University Press.
Atlas, Jay D.
2007
On a pragmatic explanation of negative polarity licensing. In: Noel BurtonRoberts (ed.), 1023.
Bach, Kent
1999
The myth of conventional implicature. Linguistics and Philosophy, 22:
32766.
Bach, Kent
2004
Pragmatics and the philosophy of language. In: Laurence R. Horn and Gregory
Ward (eds.), 46387.
Bach, Kent
2006
The top ten misconceptions about implicature. In: Betty. J. Birner and Gregory
Ward (eds.), Drawing the Boundaries of Meaning, 2130. Amsterdam: Benjamins.
Benz, Anton, Gerhard Jger and Robert van Rooy
2006
Game Theory and Pragmatics. London: Palgrave Macmillan.
Blakemore, Diane
2004
Discourse markers. In: Laurence R. Horn and Gregory Ward (eds.), 221240.
Blutner, Reinhard and Henk Zeevat (eds.)
2004
Optimality Theory and Pragmatics. London: Palgrave Macmillan.
Bultinck, Bert
2005
Numerous Meaning: The Meaning of English Cardinals and the Legacy of
Paul Grice. Oxford: Elsevier.
417
Burton-Roberts, Noel
2007
Varieties of semantics and encoding: Negation, narrowing/loosening and numericals. In: Noel Burton-Roberts (ed.), 90114.
Burton-Roberts, Noel (ed.)
2007
Pragmatics. London: Palgrave Macmillan.
Cann, Ronnie
1993
Formal Semantics. Cambridge: Cambridge University Press.
2005
Carston, Robyn
2002
Chierchia, Gennaro
2004
Scalar implicatures, polarity phenomena, and the syntax/pragmatics interface.
In: Adriana Belletti (ed.), Structures and Beyond, 39103. Oxford: Oxford
University Press.
Crain, Stephen and Paul Pietroski
2002
Why language acquisition is a snap? The Linguistic Review 19: 163183.
Cummings, Louise
2005
Pragmatics. Edinburgh: Edinburgh University Press.
Cummings, Louise (ed.)
2009
The Pragmatics Encyclopaedia. London: Routledge.
de Jonge, Casper C.
2001
Natura artis magistra: ancient rhetoricians, grammarians, and philosophers on
natural word order. In: Ton van der Wouden and Hans Broekhuis (eds.), Linguistics in the Netherlands 2001, 159166. Amsterdam: Benjamins.
Farncescotti, Robert Michael
1995
Even: the conventional implicature approach reconsidered. Linguistics and
Philosophy 18: 15373.
Fauconnier, Gilles
1975
Pragmatic scales and logical structure. Linguistic Inquiry 6: 353375.
Feng, Guangwu
2006
A theory of conventional implicature and pragmatic markers in Chinese. Ph.D.
dissertation. University of Reading.
Frege, Gottlob
1892
ber Sinn und Bedeutung. Zeitschrift fr Philosophie und Philosophisch Kritik 100: 2550. [Trans. On sense and reference. In: Peter Geach and Max Black
(eds.), 1952. Translations from the Philosophical Writings of Gottlob Frege.
Oxford: Blackwell.]
Frege, Gottlob
19181919 The Thought: A Logical Inquiry. Translation in Mind 55: 289311 [(1956)].
Garret, Merrill and Robert M. Harnish
2007
Experimental pragmatics: Testing for implicature. Pragmatics and Cognition
15: 6590.
Gazdar, Gerald
1979
418
Yan Huang
Geurts, Bart
2009
Scalar implicature and local pragmatics. Mind and Language 24: 5179.
Giannakidou, Anastasia
1998
Polarity Sensitivity as (Non-)Veridical Dependency. Amsterdam: Benjamins.
Grice, H. Paul
1975
Semantics 3: Speech Acts, 4158. London: Academic Press.
Grice, H. Paul
1989
Studies in the Way of Words, Cambridge, MA: Harvard University Press.
Grodner, Daniel, Natalie Klein, Kathleen M. Carbary and Michael Tanenhaus
2007
Experimental evidence for rapid interpretation of pragmatic some. Paper presented at XPRAG, Berlin.
Heim, Irene
1983
On the projection problem for presuppositions. Proceedings of WCCFL 2:
11425.
Heusinger, Klaus von and Ken Turner
2006
(By way of an) introduction: a first dialogue on the semantics-pragmatics interface. In: Klaus von Heusinger and Ken Turner (eds.), Where Semantics
Meets Pragmatics, 118. Oxford: Elsevier.
Hirschberg, Julia
1991
A Theory of Scalar Implicature. New York: Garland.
Horn, Laurence R.
2004
Implicature. In: Laurence R. Horn and Gregory Ward (eds.), 328.
Horn, Laurence R.
2006
The border wars: a neo-Gricean perspective. In: Klaus von Heusinger and KenTurner (eds.), Where Semantics Meets Pragmatics, 2148. Oxford: Elsevier.
Horn, Laurence R.
2007
Istvan Kecskes and Laurence R. Horn (eds.), Explorations in Pragmatics: Linguistic, Cognitive and Intercultural Aspects, 3969. Berlin: Mouton de
Gruyter.
Horn, Laurence R.
2009
1: 334.
2004
Huang, Yan
1991
A neo-Gricean pragmatic theory of anaphora. Journal of Linguistics 27:
301335.
Huang, Yan
1994/2007The Syntax and Pragmatics of Anaphora. Cambridge: Cambridge University
Press.
Huang, Yan
2000a
Anaphora: A Cross-linguistic Study. Oxford: Oxford University Press.
Huang, Yan
2000b
Discourse anaphora: Four theoretical models. Journal of Pragmatics 32:
151176.
419
Huang, Yan
2003
On neo-Gricean pragmatics. International Journal of Pragmatics 13: 87
110.
Huang, Yan
2004
Anaphora and the pragmatics-syntax interface. In: Laurence R. Horn and Gregory Ward (eds.), 288314.
Huang, Yan
2007
Huang, Yan
2009a
1: 118153.
Huang, Yan
2009b
Huang, Yan
2009c
Implicature. In: Louise Cummings (ed.), 23438.
Huang, Yan
2009d
Scalar implicature. In: Louise Cummings (ed.), 441444.
Huang, Yan
2009e
Impliciture. In: Louise Cummings (ed.), 23840.
Huang, Yan
2009f
What is said. In: Louise Cummings (ed.), 520522.
Huang, Yan
2009g
Pragmatics. In: Louise Cummings (ed.), 370375.
Hurewitz, Felicia, Anna Papafragou, Lila Gleitman and Roger Gelman
2006
Asymmetries in the acquisition of numbers and quantifiers. Language Learning and Development 2: 7796.
Israel, Michael
2004
The pragmatics of polarity. In: Laurence R. Horn and Gregory Ward (eds.),
701723.
Karttunen, Lauri
1973
King, Jeffrey and Jason Stanley
2005
Semantics, pragmatics, and the role of semantic content. In: Zoltan G. Szab
(ed.), Semantics vs. Pragmatics, 111164. Oxford: Clarendon Press.
Ladusaw, William A.
1979
Polarity sensitivity as inherent scope relations. Ph.D. dissertation. University
of Texas at Austin.
Lehrer, Adrienne
1974
Semantic Fields and Lexical Structure. Amsterdam: Benjamins.
1983
2000
Presumptive Meanings: The Theory of Generalized Conversational Implicature. Cambridge, MA: The MIT Press.
420
Yan Huang
Merin, Arthur
1999
Information, relevance, and social decision making: Some principles and
results of decision-theoretic semantics. In: Lawrence S. Moss, Jonathan
Ginzburg and Maarten de Rijke (eds.), Logic, Language, and Computation,
volume 2, 179221. Stanford: CSLI.
Moore, George Edward
1922
Philosophical Studies. London: Routledge and Kegan Paul.
Noveck, Ira and Anne Reboul
2008
Experimental pragmatics: A Gricean turn in the study of language. Trends in
Cognitive Sciences 12: 42531.
Noveck, Ira and Dan Sperber
2007
The why and how of experimental pragmatics: The case of scalar inference.
In: Noel Burton-Roberts (ed.), 184212.
Panizza, Daniele and Gennaro Chierchia
2008
Two experiments on the interpretation of numerals. Paper presented at Workshop on Experimental Pragmatics/Semantics, the 30th Annual Convention of
the German Society of Linguistics.
Portner, Paul H.
2005
What is Meaning? Oxford: Blackwell.
Potts, Christopher
2005
The Logic of Conventional Implicatures. Oxford: Oxford University Press.
Recanati, Franois
2004
Rieber, Stephen
1997
Conventional implicatures as tacit performatives. Linguistics and Philosophy
20: 5172.
Russell, Benjamin
2006
Against grammatical computation of scalar implicatures. Journal of Semantics
23: 36182.
Russell, Bertrand
1905
On denoting. Mind 14: 47993.
Saeed, John I.
2009
Sauerland, Uli
2004
Scalar implicatures in complex sentences. Linguistics and Philosophy 27:
367391.
Saul, Jennifer
2002
What is said and psychological reality: Grices project and relevance theorists
criticisms. Linguistics and Philosophy 25: 347372.
1995
1974
Pragmatic presupposition. In: Milton Munitz and Peter Unger (eds.), Semantics and Philosophy. New York: New York University Press, 197214.
Strawson, Peter F.
1950
On referring. Mind 59: 320344.
421
Strawson, Peter F.
1952
Welker, Katherine
1994
Plans in the common ground: Toward a generative account of conversational
implicature. Ph.D. dissertation. Ohio State University.
Woods, John
2009
Inference. In: Louise Cummings (ed.), 218220.
Part IV
The place of pragmatics in the description
of discourse
15.

Arnulf Deppermann
1.
Introduction
In this article, grammar is used as an overarching term, encompassing morphosyntax as well as the syntax of the sentential structures (phrase structure and topological structure). Grammar thus includes syntactic categories, functions, and
rules, grammatical constructions, and their realization in terms of forms and patterns of linearization. As grammatical constructions might be more or less lexically fixed and more or less restricted to, e.g., semantic classes of verbs, the boundary between lexicon and grammar is at best fuzzy, and may be better conceived of
as a continuum (Croft 2008). Pragmatics here is also used in a wide sense, referring to the use of language in context. In this way, it encompasses cognitive preconditions and functions of language, such as attention, memory, conceptualization, categorization, and inference, as well as interactional uses, such as transfer
of knowledge, turn-taking, speech acts, and rhetorical functions, displays of emotion and epistemic stance, and indexing social identities and relationships.
In order to set the scene, here are some basic questions about the relationship
between grammar and pragmatics, which will be taken up in this article:
a) The role of pragmatics for grammar
Which pragmatic factors have an impact on grammar? Candidate factors
include cognitive, interactional, situational, cultural, and bodily factors.
Which role do pragmatic factors play in defining grammatical structures?
Can or should grammatical structure be defined on a pure formal basis or
does it ultimately rest on functional categories?
How is pragmatics involved in parsing and interpreting grammatical constructions?
Which grammatical phenomena are governed by pragmatic factors both in
terms of their diachronic emergence and regarding their situated use?
How does usage shape grammar over time?
Which role do pragmatic uses and contexts play in the acquisition of
knowledge about grammar?
b) The role of grammar for pragmatics
How are grammatical structures adapted to pragmatic preferences and
functions?
How are they adapted to their interactive and situational context?
How is grammatical structure used as a flexible tool for speakers pragmatic concerns?
426
Arnulf Deppermann
Do grammatical structures as such have pragmatic meaning?

How are grammatical constructions coordinated with other orders of embodied action?
When asking questions about the relationship between grammar and pragmatics,
we are interested in a subset of issues of the more general question of how language
form relates to language function (see Newmeyer 1998).
2.
Theoretical positions: Formal vs. functional approaches to grammar
No linguist would deny that the choice and at least some aspects of the interpretation of grammatical structures are determined by pragmatic considerations.
However, it is a disputable matter how deeply grammar and pragmatics are intertwined, and consequently, if grammatical analysis needs to attend to pragmatic factors or if it should rather treat grammar as a self-contained module which is working and needs to be analyzed by its own self-sufficient principles. The latter would
be the Chomskian view, coined the autonomy of syntax-thesis (Chomsky 1965).
According to this view, pragmatic factors neither explain the basic parameters and
principles of grammar (e.g., universal organizing principles such as recursivity and
head-specifier-structures), nor do grammatical features arise from pragmatic functions and pressures: In general, it is not the case that language is readily usable or
designed for use (Chomsky and Lasnik 1993: 509). Grammar (which, in this
view, is language proper) is considered as a formal, algebraic apparatus for generating possible, well-formed linguistic structures and for deriving hypotheses about
linguistic structure from perceived speech. While this generativist stance is sometimes dubbed as a formalist approach, functionalists of various kinds (see below)
assign pragmatics an important role in explaining grammatical structure. Basically,
they view language as a means for the communication of thoughts and for social
interaction. They hold that pragmatic functions (such as information-transfer, turntaking, etc.), principles (such as iconicity, categorical organization around prototypes, etc.), and pressures (having to do with processing time, attention span, etc.)
motivate, constrain, and sometimes even fully explain the becoming, use, and interpretation of grammatical structures. There is, however, a broad variation between different functionalist approaches in terms of which grammatical phenomena and which pragmatic factors they focus on and how they conceive of the
impact pragmatics is said to have on grammar (see below and Newmeyer 1998).
2.1.
Pragmatic approaches to grammar
In what follows, we will shortly review approaches which see grammar rooted in
pragmatic motivations.
2.1.1.
427
Functional Grammar and Functional Discourse Grammar
The Dutch theory of Functional grammar (FG; Dik 1997) starts from the assumption that language is structured according to its function as a means of communication of knowledge. It assigns pragmatics the hierarchically dominant role in
determining the semantics and the grammar of clauses. In essence, however, pragmatics in this approach boils down to the information status of speaker and addressee in the ongoing discourse, which is said to determine the information structure of the clause in terms of choosing referring expressions, presuppositions, and
organizing focus vs. background. Expression rules determine how pragmatic and
semantic structures are mapped onto word order and morphosyntax. FG is a nonderivative, cognitivist theory, which requires typological, psychological, and pragmatic adequacy as standards for modelling linguistic structures in formalisms,
which provide for testable predictions (Dik 1997).
Functional Discourse Grammar (FDG) (Hengeveld and Mackenzie 2008) revises FG so as to become a discourse grammar, centering around the discourse act,
which may be larger or smaller than a sentence. Discourse acts are dominated by
speakers intentions (interpersonal level). In a top-down fashion, they govern the
representational (semantics), the structural (morphosyntax) and the phonological
level, which are each formally characterized in strict separation. In contrast to the
other approaches discussed here, (D)FG formalizes its assumptions in a predicate
calculus.
2.1.2.
Systemic-Functional Grammar
Since the 1960s, Systemic-Functional Grammar (SFG) has developed out of the
Firthian British functional linguistics. Its leading figure, M.A.K. Halliday (1985),
assumes that grammar is a system for the creation of meaning, which is structured
according to three metafunctions: the interpersonal (enacting relationships between speaker and addressee), the ideational (construing experience), and the
textual (organizing interpersonal and ideational meaning in the unfolding text or
interaction). Grammars are taken to consist of systems of options, which are available for realizing the metafunctions. Examples are the systems mood for realizing the interpersonal metafunction, transitivity (referring to types of processes
and the participants involved) for realizing the ideational metafunction, and
theme for realizing the textual metafunction. Systems are specific for ranks in the
constituency structure (clause, phrase, word, morpheme). An important place
within SFG is assigned to the study of registers, i.e., varieties of language which
are tied to certain types of situation (e.g., different degrees of formality, languages
for special purposes). SFG assumes that the context of the situation is to be categorized according to three domains, which roughly correspond to the metafunctions: tenor (social/interpersonal relationships), field (processes talked about), and
428
Arnulf Deppermann
mode (medium/channel). Some scholars add genre (relating to the context of culture) as a fourth important domain (Eggins 1994: Ch. 3).
2.1.3.
Functional Pragmatics
Functional Pragmatics is a German approach which sees language as a means designed for the transfer of knowledge and the achievement of mutual understanding
(Ehlich 1991; Redder 2008). Building on Speech Act theory and Karl Bhlers theory of linguistic fields (Bhler 1934; Daalder and Musolff this volume), Ehlich
(1991) distinguishes five linguistic procedures which are used to label five corresponding functional fields with respect to a specific addressee. Each procedure is
realized by specific linguistic means.
Field
Procedure
Linguistic means
Symbol
symbolic
lexemes of content words
Pointing
deictic
local, temporal, personal deictics; tense markers
Guiding
expeditive
interjections; intonation; imperative and vocative

mood
Operational
operational
anaphora; determiners; sub-/conjunctions;

prepositions, accent
Painting
expressive
imitation; exclamative intonation
Figure 1. Fields, procedures, and linguistic means in Functional Pragmatics

(adapted from Hoffmann 2001)
2.1.4.
Cognitive Linguistics and Construction Grammar
Most approaches of cognitive linguistics, such as Cognitive Linguistics (Lakoff

1987), Construction Grammar (Goldberg 1995; 2006), Cognitive Grammar (Langacker 1987; 2008), and Radical Construction Grammar (Croft 2001), consider
grammar to be a semiotic system. Grammatical constructions, which are seen as
the basic, irreducible building blocks of linguistic knowledge (see, e.g., Croft
2001; Goldberg 1995), are defined as form-meaning-pairings (Croft 2008). According to Langackers symbolic thesis, Grammar (or syntax) does not constitute
an autonomous formal level of representation. Instead, grammar is symbolic in nature, consisting in the conventional symbolization of semantic structure (Langacker 1987: 2). The meaning-pole of grammatical constructions includes not only
semantics in a narrow sense, but also discursive, interactional, textual, genre-related, rhetorical, and other pragmatic meanings.1 Besides, following Fillmore
(1985), Lakoff (1987), and others, Cognitive Linguistics denies that there is a clear
boundary between semantics and pragmatics, claiming that all linguistic knowl-
429
edge is in fact encyclopaedic and based on experience. It is only very recently that
cognitive linguistic approaches have turned to the analysis of text-corpora. When
dealing with pragmatic aspects of the interpretation of grammatical constructions,
they mainly focus on information structure2 and implicature (Kay 2004), while discursive phenomena are largely disregarded (but see Langacker 2001). Cognitive
linguists hold that grammar is basically organized in the same way as the lexicon,
linguistic units (Langacker 2000) or constructions (Croft 2008) being the
overarching term. All linguistic structures are subject to processes of categorization (Langacker 2000). Just like lexical items, grammatical constructions are
structured in terms of categories which exhibit a prototypical structure, resulting in
radial and sometimes fuzzy categories, where the individual members need not
share a common set of necessary and sufficient features, but may be related to one
another by family resemblances. Grammatical constructions extend over time
according to principles of metonymical and metaphorical extension, which supply
for motivated options of the change of categorical structures, which, however, cannot be predicted.
2.1.5.
Usage-based approaches
The usage-based approach (Barlow and Kemmer 2000; Bybee 2006, 2007) judges
frequency of usage to be the key factor for the becoming of grammatical constructions. Exemplar models predict that frequency of use is stored and determines the
cognitive entrenchment of linguistic units because it correlates with the probability
of storing a linguistic structure as an unanalyzed unit. Usage-based approaches resonate with connectionist models (MacWhinney and Bates 1989; Bod, Hay and
Jannedy 2003), which rely on neurobiological principles. They predict that high
frequency of reception and production of some linguistic structure strengthens its
activation pattern and facilitates automatic processing. Simultaneously, the occurrence of linguistic structures leads to an inhibition of activation and de-automatization of competing structures (e.g., prepositional genitive with of vs. morphosyntactic genitive with -s). In this view, grammaticality (and conventionality in
general) is a matter of degree and preference, depending on weighted strengths,
which are associated with supposed neural activation patterns and which are subject
to constant change. Synchronically, usage provides for prototype effects of grammatical structures, frequent structures becoming templates for categories and extensions organized around them. Diachronically, it is seen as a major cause of language change and, in particular, of grammaticalization (see 4.8). Such findings are
in line with the position of Emergent Grammar advocated by Hopper (1998, 2004;
Auer and Pfnder 2010). He rejects the assumption of any innate grammatical capacity and he denies that there was anything like the idea of a linguistic system. In
his view, language exists only as an ever changing process of linguistic usage. It is
informed by prior linguistic experiences, which provide the grounds for the routine
430
Arnulf Deppermann
use of (often schematically organized) fragments, which are creatively used as resources and adapted to the situated contingencies of local linguistic practice. In
Hoppers view, sense-making does not rest on rules for well-formed linguistic
structures, but emerges from hermeneutic interpretation, which is in need of interactional ratification.
2.1.6.
Interactional Linguistics
Interactional Linguistics (IL) is strongly based on the methodology of Conversation Analysis (see Domke and Holly this volume). It regards the sequential
organization of talk in oral interaction as the primary habitat of language.3 IL assumes reciprocal influences between linguistic and interactional structures: Linguistic structures are functionally adapted to discursive tasks, while patterns of discourse are sensitive to the linguistic options which individual languages afford and
which speakers choose. IL insists that a theory of the grammar of spoken language
must elucidate the principles of the online-production and -reception of talk-in-interaction (Auer 2009). Most important in this respect is the mechanism of projection of possible/required grammatical continuations and possible/expectable next
actions given the production of an ongoing turn so far (Auer 2005 and 2009). IL
studies put special emphasis on the interplay of grammar and prosody with respect
to action formation (see also Couper-Kuhlen this volume). In contrast to the other
approaches discussed here, IL is not a linguistic theory with a specific model of
grammar and a definite descriptive apparatus, but rather a methodological approach. IL refuses abstract theorizing and insists on the analysis of collections of
linguistic phenomena in talk-in-interaction based on audio-/video-recordings and
transcripts. During the last years, however, researchers in IL increasingly adopt
concepts from construction grammar as descriptive framework4 and positions from
usage-based approaches and emergent grammar (Auer and Pfnder 2010).
3.
The relevance of pragmatics for grammar
In what follows, the impact of pragmatic tasks, processes, and conditions on grammatical structure both in terms of its situated use and of properties of the linguistic
system will be discussed. Pragmatic key factors include temporality (online production and reception), interaction, cognition, action, the temporal and spatial
arrangement of the participants, the modes and materiality of embodied and mediated communication and frequency of usage. For grammatical phenomena relating to the pragmatics-semantics interface, like the interpretation of connectives
and scalar terms and the role of explicature, impliciture, enrichment, conventional
and generalized implicatures in relationship to propositional content and what is
said, see Saeed (this volume).
3.1.
431
Processing constraints and the grammatical packaging of information
The communication of thoughts and the performance of actions by linguistic

practice in real time is constrained by speakers and hearers capacities of cognitive processing. The temporal scope of attention and working memory delimits the
complexity of syntactic and propositional structures which can be processed as
units of planning and understanding. According to Chafe (1994: 109), there is a
one new idea constraint, meaning that neither the speaker nor the listener is able
to handle more than one new idea at a time. This constraint guides the packaging
of information in prosodic and syntactic units. Preferentially, speakers formulate
one new information per intonation unit, which has its (one) focus accent on the
item which expresses the new information. The intonation unit almost never comprises more than one clause. As a consequence, syntactic structures with multiple
recursions, which force speaker and hearer to keep track of two or even more syntactic and semantic structures simultaneously, are strongly dispreferred because of
the processing difficulties they cause.
An important syntactic repercussion of the one new idea constraint can also
be seen in the constraints of Preferred Argument Structure (Du Bois 2003), which
have been proved to hold for a large variety of typologically unrelated languages
(Du Bois, Kumpf and Ashby 2003). Du Bois (2003) claims that speakers avoid
more than one lexical core argument (which encodes the new information). For
transitive constructions, this amounts to the constraint to avoid lexical subjects
(and to express them pronominally), while direct and indirect objects can be lexicalized. The constraints of preferred argument structure explain, e.g., preferences
for cleft-constructions which can be observed in many languages, such as French
(cf. Lambrecht 1987). Instead of producing a transitive clause with both lexical
subject and object, speakers prefer a cleft-construction, which package the information in two clauses with only one lexical argument each.
(1) Transitive clause
Beaucoup de gens naiment plus cette faon de parler.
Many people do not like this way of talking anymore.
(2) il y a-cleft
Il y a beaucoup de gens qui naiment plus cette faon de parler.
There are a lot people who do not like this way of talking anymore.
Such cleft-constructions fulfil the same pragmatic function as left dislocation and
free topics: They are used to introduce a referent in a segment of its own, which, in
addition, is often produced as a separate intonation unit apart. Only after this, the
referent is topicalized and predicated upon.
(3) Left dislocation
aber ich mein ne gewisse neugier (.) die muss ja bleiben
but I mean a bit of curiosity (.) it must remain
432
Arnulf Deppermann
(4) Free topic

die polnische Wurst da (.) ist das so wie Salami?
the-FEM Polish sausage over there (.) is it-NEUTR like Salami?
Grammatical constructions as in (3) and (4) give evidence that in contrast to what
Speech Act Theory holds (cf. Searle 1969), an elementary speech act need not consist of reference plus predication. Rather, establishing a referent often is a fullblown action by itself. It is performed by clausal and phrasal constructions which
separate it from the following predication and serve as a starting point for it, which
first is grounded independently. Moreover, referential act and predicative act each
occupy an intonation unit of their own.
In a similar vein, speakers prefer right-branching constructions over leftbranching constructions because they allow for a cumulative processing of structure and meaning, which correlates with the temporal production of phrases. Leftbranching constructions, however, require language users to put elements on
hold in the working memory, before being able to assign them to their syntactic
heads, which are produced later. The need to keep track of several syntactic and
semantic representations in parallel also strongly disfavors central embedding, let
alone multiple recursion (such as multiple center-embedded relative clauses).
3.2.
Structuring information with respect to relevance and the recipients

knowledge
Text and talk is designed with respect to what speakers/writers assume to be common ground with their addressees. New information, which is either unknown or
not salient to the addressee at the moment, is introduced into the discourse by relating it to given information, which is judged to be salient, i.e., activated and retrievable with low processing effort for the hearer. In order to structure and encode
information accordingly, speakers keep track of the cognitive availability of referents (but also other denotata, such as states, actions, and events).
Given vs. new is not a simple dichotomy. Rather, there are different degrees of
accessibility of referents (e.g., Chafe 1994), which depend crucially on their identifiability and their state of prior activation. Denotata may be accessible because
they are situationally salient (via perception), established as parts of world knowledge (like the prime minister), have been introduced in prior discourse, or can be
inferred from it via bridging assumptions which, e.g., relate parts to wholes (Clark
1977).
(5) Bridging assumption (whole f part)
The couple moved on to the altar. The bride wore white.
f Bridging assumption about co-reference: The bride is part of the couple mentioned before.
433
The assumed familiarity of referents for the recipient thus can range from (brand-)
new over inferrable to (exophorically or anaphorically) evoked items (cf. Prince
1981). Accessibility of referents depends on several factors, such as its topical vs.
non-topical status, recency of last mention of the referent, or the degree of cohesion of the syntactic unit which contains the prior mention to the ongoing unit
(Ariel 1990).
Ariel (2008: 4452) proposes a referential marking scale. It orders referential
expressions according to the (assumed) degree of accessibility of the referent for
the recipient, which in turn depends on the degree of its activation in the recipients
memory that can be assumed when producing the referential expression (see also
Gundel, Hedberg and Zacharsky 1993). While new elements are introduced by indefinite NPs, accessible and given elements are referred to by definite NPs or (if
situationally salient or evoked) by pronouns. Starting with the most marked expression to be used for the least accessible referents, the referential marking scale
for person reference looks like this in more detail (Ariel 2008:44):
full name > long definite description > short definite description > last name > first name
> distal demonstrative > proximate demonstrative > stressed pronoun > unstressed pronoun > cliticized pronoun > verbal person inflections > zero
This scale shows how the two pragmatic preferences for person reference, namely
minimization and use of (recipient designed) recognitionals (Sacks and Schegloff
1979), translate into preferences for the use of grammatical resources. Their relevance can also be seen in cases of repair like (6), where more minimized forms are
used first and get replaced by (grammatically and semantically) more elaborate
forms if they fail to engender the recipients recognition.
(6) Preference for minimization and use of recognitionals (example from Sacks
and Schegloff 1979: 19)
A: I was the only other than the tch Fords? Uh Mrs. Holmes Ford? You know
uh the the the cellist?
B: Oh yes. She is the cellist.
Apart from morphosyntactic coding, word order is used as a syntactic device to
index the assumed familiarity of referents. Halliday (1985: 275) claims that there is
a natural ordering principle: the Given typically precedes the New, i.e., the
speaker organizes the message as a trajectory starting from the current state of consciousness of the recipient to an updated state (see also Chafe 1994). Therefore,
given information is expressed as theme in the prototypical case. In Germanic and
Romance languages, the theme is the first constituent of a sentence, whereas in
Japanese it is marked by the post-positioned theme marker -wa. New information
is presented in the rheme, i.e., the remainder of the sentence (Halliday 1985: Ch.3).
In addition to the assumed familiarity of referents, a second important dimension
of information structure is the organization of focus vs. background (Bring 2007;
434
Arnulf Deppermann
Krifka 2007). Focus refers to the part of an intonation unit which is framed as
being most relevant. Some of the means speakers can use to mark focus are the
same as for marking accessibility: word order and specific topological constructions like (pseudo-)clefts and passives. In addition, most important is the placement of the so-called focus accent (Uhmann 1991; Bring 2007), which marks
the focal word (but can project to higher-order nodes such as phrases or even the
whole clause). All other things being equal a Speaker will choose the Theme from
within what is Given and locate the focus, the climax of the New, somewhere
within the Rheme (Halliday 1985: 278). While there is a high correlation between
focus, new information and rheme, they need not coincide (Gundel and Fretheim
2004). The focus does not necessarily need to be put on new information, but it
can, e.g., be contrastive, i.e., marking an element which stands in contrast to a prior
turn or the recipients probable expectations. The focus can also comprise the
whole sentence, as in the case of offering requested information to open questions
like
(7) Focus on whole sentence (= requested answer)
A: Whats up?
B: The computer doesnt work.
or as in newspaper headlines
(8) Focus = whole headline (The New York Times, November 27, 2009)
Pilots Were Distracted, Transcripts Show
In elliptical responses, only the focus elements are realized.
(9) Adjacency ellipsis: Question-answer sequence
A: wieviel Lieder knnen wir spielen? was meint ihr?
How many songs can we play? What do you think?
M: Z::irka fuffzehn;
About fifteen.
In spoken language, the focus is often placed in thematic position, particularly, if it
is contrastive.
(10) Focus on theme
A: WANN ist er gekommen?
when did he come?
B: GEStern war er=s erste mal da.
yesterday he was there for the first time.
Another grammatical construction for focus marking by assigning it to the theme is
preposing a constituent (off he went; cf. Ward and Birner 2004). Focus preposing
needs to be distinguished from topicalization of a referent, which does not carry a
focus accent and contains neither focus nor new information.
435
As stated above, focus marking can be independent of accessibility status:

(11) Contrastive accent
Das war nicht ICH. Das warst DU.
It wasnt ME. It was YOU.
In fact, focus marking and marking of accessibility are not determined by identifiability or activation status of referents, but they can be used as rhetorical devices
to introduce new information by simply presupposing it as hearer-old (see also Edwards 1997:117122):
(12) Coding givenness as a rhetorical device (advertisement by Volkswagen)
Just when you thought our legendary hatchback couldnt get any better,
along comes the stunning new Golf.
In sum, signalling familiarity of referents and assigning focus vs. background results from an interaction between grammatical forms (such as definiteness, pronominalization, deictics, etc.), word order, and prosodic cues (accentuation),
which at times can be very complex.
3.3.
Projecting turn-construction and turn-taking
In spoken language, the clause is not only the preferred format for packaging ideas
(see 4.1), it is also used as the primary construction format for turns and actions
(Thompson and Couper-Kuhlen 2005). Participants use clausal structures for indexing and recognizing transition relevance places at which turn-transition might
be due (Selting 2000; Sacks, Schegloff and Jefferson 1974). A basic mechanism to
do this is temporal projection (Auer 2005 and 2009), which rests on the incremental production of linguistic structures in time. Linguistic structures produced in the
turn so far open up syntactic gestalts, i.e., probable or required continuations,
which have to be closed by producing the required missing elements. E.g., the apodosis of a conditional projects a protasis (and this is: a clause-combination) to follow, verbs project required arguments by virtue of their valence frame, determiners
project noun phrases, etc. Projection depends on typological features of the language. E.g., in German the verbal brace (Verbklammer) requires speakers to finish
clauses with an infinitival form (infinitive or participle) of the main verb (unless it
is used in present tense without a modal). Listeners can use this feature to generate
expectations about when the turn is possibly completed. SVO-languages like English permit an early projection of turn trajectories because the verb makes specific
choice of thematic roles expectable. An OV-language like Japanese, however, does
not allow for early projection, because constituents other than V do not project the
number of arguments and because Japanese allows for a much higher degree of ellipsis. Still, it marks endings more clearly in terms of syntax than SVO-languages
because of the regular positioning of the verb in sentence-final position and be-
436
Arnulf Deppermann
cause of the additional use of sentence-final particles (Thompson and CouperKuhlen 2005). Projectability of turn-endings makes itself felt not only in smooth
turn-transition, but also in cases of overlap, which tend to occur more often if elements are added after a possible clausal gestalt has been completed and thus a
possible transition relevance place has been reached (Jefferson 1973; Schegloff
1996). The collaborative production of sentences is another evidence for participants orientation to the exigencies of grammatical structure in terms of projectable trajectories of clauses (Hayashi 2003; Lerner 1991, 1996, 2004): After the
first speaker has produced the first clause of a complex sentence, second speakers
manage to deliver the second clause which was type-projected (e.g., a then-clause
after an if-clause has been completed) or they may add phrases or words at the
places where their occurrence is due.
Although clauses have more or less determinate endings, which may signal
turn-transition, they can be expanded by additional elements, so-called expansions or increments, which are syntactically related to the syntactic structure
constructed so far.5 Such turn-continuations occur regularly if a projected next
speaker does not take the turn when a transition relevance place is reached. They
address problems of understanding and affiliation with the turn so far, which might
account for an actual lack of response or an anticipated difficulty in responding.
They serve to repair the respective deficits of the turn and to renew the occasion for
turn-transition. Whether clauses project turn-transition and whether expansions are
a continuation of an ongoing turn-constructional unit or a new unit does not only
depend on syntax, but also on the prosodic contour of the clause and on the degree
of prosodic integration of the continuation. Transition relevance places are not
only determined by syntax, but they emerge from an interplay of syntactic, prosodic, semantic, and, at times, also pragmatic and genre-related considerations (as,
e.g., narrative structure).
3.4.
Social action formats
Grammar is a primary resource for performing recognizable social actions. Speech

act theory conceives of the relationship between speech acts and grammar in terms
of illocutionary force indicating devices, which index and sometimes even encode
types of speech acts. Most important indicators are sentence type (declarative,
imperative, interrogative), word-order, modal verbs and modal particles, performative formulae, punctuation, and sentence-final pitch contours (Searle and
Vanderveken 1985; Liedtke 1998). Empirical research into the relationship between action and grammar has refined this picture quite a lot: Instead of relying on
theoretically derived speech act taxonomies and trying to find linguistic correlates
for very broadly defined (universal) types of speech acts, corpus-based research
shows that linguistic action needs to be understood in terms of much more specific
emic types of actions to which speakers and recipients orient as such. At times,
437
they are tied to very specific sequential contexts, genres, and speech communities.
These actions are realized by idiomatic grammatical constructions, which have become closely associated with them by routine usage in specific pragmatic contexts
and genres, leading to the conventionalization of pragmatic meaning (Feilke 1996
and 2004). Examples of social action formats (Fox 2007) are constructions with
German nicht verstehen (do not understand; Deppermann 2010), which are used
for building reproaches:
(13) Social action format: [nicht verstehen (knnen) COMP] indexes reproach
B1: ich versteh auch andereseits NICHT wenn dieon the other hand I dont understand when the
klgerin zum arzt nach BRUCHtal MUSS JA?
plaintiff (=A1) needs to go to the doctor in Bruchtal
dass sie dann in de innenSTADT im maxim
that she then goes to Maxims in the city
KAFfee trinkt
to have a cup of coffee
(1.8)
A1: wieso is des verBOte?
why is that forbidden?
Deontic infinitives in German are another example. Deontic infinitives are not syntactically integrated, i.e., they are free infinitives, which are neither governed by
a matrix sentence nor are they part of a composite verbal phrase made up by an
auxiliary and an infinitive (cf. Fries 1983; Deppermann 2006 and 2007). With
deontic infinitives, the speaker expresses a normative stance towards some kind of
(future) action denoted by the verb. They are employed for a variety of deontic actions, which may vary in terms of the degree of obligation, such as commands, requests, recommendations, and suggestions, but also the expression of a wish (optative) or a complaint.
(14) Social action format: Deontic infinitive used for suggestion
A: also der Ablauf is grob (--)
so the procedure will be something like this
Anmoderation; (.) oder lAbern;
anchorage or talk-INF
M: vorstellen auf jeden fall
introduce-INF in any case
(15) Social action format: Deontic infinitive used for complaint
noch drei wochen arbeiten!
still three weeks of work-INF
arbeiten arbeiten immer nur arbeiten
work-INF work-INF always only work-INF
438
Arnulf Deppermann
An English example is the use of the grammatical construction [PRO COPULA

(Grading particles: really/so/quite/pretty) PRED-ADJ] for assessments, in particular, for compliments (Goodwin and Goodwin 1987):
(16) Social action format: [PRO COPULA (Grading particle) PRED-ADJ] used
for assessments
This is so nice.
Thats quite good.
Another example is the use of post-positioned causal adverbial clauses (starting
with because) as justifications dealing with actual or anticipated disaffiliation on
the recipients part (Ford 1993):
(17) Social action format: Post-positioned causal adverbial clause dealing with a
missing recipients reaction (example from Ford, Fox and Thompson 2003:
126)
A: well do you think its umm (0.2) ahm (0.2) stress? (.)
Cause a lot of back I know back pain (0.2) comes from stress.
Actions are rather contextually indexed than invariably coded by grammatical
constructions: Sequential, genre-related, and other contextual conditions, lexical
choice of the verb, and prosody determine which precise action is performed by a
deontic infinitive; causal adverbial phrases may also be used to secure understanding or to refute a statement, which an opponent had asserted before. Grammar only
very rarely stands in a context-free, one-to-one relationship to some type of action.
3.5.
Displaying stance and marking discursive coherence
Texts and interactions do not simply consist in the exchange of information and the
performance of actions. Speakers use various grammatical means to display their
epistemic and evaluative stance with regard to information and action, to convey
how they relate to recipients assumed understandings and to connect their utterance with prior and subsequent clauses and turns. Various kinds of discourse particles (see Fischer 2006) such as discourse markers (like you know, well, and in
some of its uses; Schiffrin 1987), response particles (like uhum, yes, okay; Gardner
2001; Sorjonen 2001), and modal particles (like just, germ. wohl, eben, halt, ruhig;
Diewald 2006a) are grammatical means which are specialized for displaying how
some action relates to prior (or presupposed) context in terms of coherence, epistemic stance, or affiliation with the partner. With the exception of modal particles,
discourse particles are not syntactically integrated. Mostly, they occupy turn-initial
positions in the so-called pre-front field (Auer 1996b), or, in the case of response
particles, they sometimes make up the whole turn. Pre-front field elements can also
be adverbials, address terms, and routine formulae.
439
(18) Pre-front field: Connective + response particle

also gut das haben sie verstanden
so okay you got this
(19) Pre-front field: Address term + discourse marker
herr meier ich mein wir haben uns ja gestern schon drber unterhalten
mister meier I mean we talked about this already yesterday
Pre-front field elements serve as pivotal elements which retrospectively tie the upcoming action to prior context and prospectively project the type of action to come
in various respects.
Projector constructions comprise another set of prefatory grammatical structures which serve for anticipatory framing of upcoming discursive action (Gnthner 2008; Hopper and Thompson 2008). Projector constructions can be realized in
various forms, which are in part language specific: as pseudo-clefts (wh-clefts;
Gnthner and Hopper 2010; Pekarek-Doehler 2010), it-extrapositions (Prince
1978), [N be that]-constructions (Gnthner 2008; Schmid 2001), il y a-clefts, or
the je veux dire que-construction (Pekarek-Doehler 2010).
(20) German pseudo-cleft (from Gnthner and Hopper 2010)
was so die sachn in natUrkostldn angeht,
as far as the things in health food shops are concerned
die sind ja dann meistens
they are mostly
erfahrungsgem auch relativ teuer
also quite expensive
(21) English pseudo-cleft (from Gnthner and Hopper 2010)
and so what happens is,
your mic runs into i:t,
and then it plugs into your receiver
(22) French pseudo-cleft (from Pekarek-Doehler 2010)
ce que je voulais dire
what I wanted to say
cest vrai qu il y a quand-mme tout ce phnomne,
its true that there is nevertheless this whole phenomenon
Examples (20)-(22) show that the two parts of the pseudo-clefts need not be linked
by a copula and a subordinating construction (what p is that q). Moreover, the second part often does not consist of a clearly delimited clause, but rather of some
more vaguely defined longer stretch of talk. The projector construction thus is a
very interesting case of the intersection of grammar and pragmatics (cf. Hopper
and Thompson 2008). The first parts (the wh-clauses etc.) do not project some spe-
440
Arnulf Deppermann
cific clausal structure to follow, but they rather serve as flexible fragments for holding the turn by projecting a multi-unit turn to follow. Depending on the metacommunicative framing or the announcement made by the wh-clause, it projects an
argumentation, a complex statement or a small story.
3.6.
Presupposing context
Ellipsis provides the most striking evidence for the claim that syntax is tied to the
here and now of talk (Bhler 1934) and that it is positionally sensitive (Schegloff
1996). According to traditional definitions, ellipsis applies to grammatical structures, which are acceptable in context (i.e., they are not anakoluths), but which are
phonologically reduced with respect to the exigencies of syntactic completeness
and of their semantic interpretation in situ. Ellipsis results from economical motivation (cf. Grices second maxim of quantity, see Grice 1975). It requires that
elided elements can be recovered from context (Halliday 1985: Ch.9.3). Context
can offer the missing elements in basically two ways: Ellipsis may be anaphorical
or situational.
a) Anaphorical ellipsis copies presupposed elements from prior text or talk (save
adaptations of speaker-listener deixis). Adjacent ellipsis is a most obvious case.
Typically, a second turn is restricted to the production of the (requested, problematic etc.) focus element, while the prior turn is presupposed as background:
(23) Adjacent ellipsis: Answer
IN is sexualitt fr sie total WEG?
is sexuality completely off for you?
IE als unterhAltungsthema (.) JA.
as a topic for talk (.) yes
(24) Adjacent ellipsis: Follow-up question
IE: das finde ich falsch
that is wrong in my view
IN: warum?
why
However, phonological reduction is not the full story. There are in cases where
there is no prior clausal structure which can simply be copied. Consider cases like
(25) and (26):
(25) Unattached NP increment (example from Ford, Fox and Thompson 2002)
She had it [a baby] yesterday.
Ten pounds.
441
(26) Free NP
CH: der abwasch ist FERtig (-) darf ich jetzt FERNsehen?
the dishes are cleaned (-) may I watch TV now?
PA: ERST hausaufgaben DANN fernsehen.
first homework then TV
While the subject can be copied from the preceding turn constructional unit in (26),
there is no verbal antecedent to be presupposed for the elliptical structure. Instead, it
has to be constructed on the basis of the representation of the conceptual content of
the discourse so far and by resorting to encyclopaedic knowledge and sequential expectations with regard to next actions. E.g., in (26), the request for permission makes
permission vs. refusal conditionally relevant and assigns normative power to the respondent. In addition, the construction [first NP/VP then NP/VP] is an elliptical social action format, which codes a normative order of actions. Thus contextual resources of interpretation can combine with the semantics and pragmatics which are
tied to elliptical grammatical constructions as such. Ellipsis may not only be motivated by striving for economy, but it may also have rhetorical functions. Deontic infinitives (cf. (14)-(15)) are a case in point: By neither encoding subject (i.e., an agent)
nor mood, the construction neither specifies who has to carry out the action nor the degree of obligation assigned to the agent. In some contexts, the lack of mood-coding
and the omission of arguments (agents, but also often patient, location, instrument,
etc.) is a politeness strategy (for requests and commands), while in others (like advertisements, see (27)) it allows the recipient to define for him/herself which degree
of normative obligation suits and if s/he feels addressed by the deontic infinitive at all.
(27) Deontic infinitives in advertisement
Jetzt mitspielen und gewinnen!
Join-INF the game and win-INF now!
In addition, these omissions help to profile (Langacker 2008: 6670) the action as
the most important element of the situational frame evoked by the infinitive clause.
These factors seem to account for the fact that in many contexts deontic infinitives
are preferred over imperatives and modal constructions, which require the speaker
to encode agents and the degree of normative obligation. Ellipsis can reach back
far beyond an immediate prior interactional context. Raymond (2004) shows how
stand-alone so in American conversation is used to prompt an action from a recipient as an upshot from the ongoing interactional sequence which had been due already (much) earlier in the interaction. The syntactic incompleteness of so is a case
of aposiopesis. It invites the recipients collaboration by indexing that the recipient
is supposed to know on behalf of the prior interactional sequence which kind of
action is warrantably expected from him/her. A similar case is sentence-final
but, which indexes a hanging objection which can be recovered from context
(Mulder and Thompson 2008).
442
Arnulf Deppermann
b) When situational ellipsis is used, it is the non-verbal context with respect to the
speakers origo (Bhler 1934), which supplies the missing information.
(28) Situational ellipsis: pragmatic routines (customer to waiter)
The bill, please.
(29) Situational ellipsis: pragmatic routine and local specification (announcement
at a railway station platform while train is approaching)
Zurckbleiben!
Keep away!
(28) and (29) are examples for routinized elliptical constructions which are tied to
specific social and (in the second case) spatial configurations which they require
for their interpretation. Note that there is a reflexive relationship between the construction and the context: While the situational context is necessary for their precise situated interpretation, the formulae themselves organize contexts by indexing
features which are (possibly) relevant.
3.7.
Multimodal coordination
Situational ellipsis mostly does not simply rest on some spatiotemporal context
which is taken for granted. Rather, the production of grammatical structure is coordinated in a fine-grained way with other modes of communication (Kress and
van Leeuwen 2001; Norris 2004) which make contexts selectively salient and
which cooperate in determining meaning. In oral interaction, the meanings of
spatial reference and situational ellipsis are established by a complex interplay of
linguistic structure, pointing, available visual clues, background knowledge, and
larger activity frames, which provide for putative relevance and possible semiotic importance of spatial structures (Goodwin 2003). E.g., pointing gestures
need to precede deictic terms used for spatial reference within or shortly before
the same turn in order to draw the recipients attention to the objects designated
and thus to constrain the situated meaning of the terms used. An example for the
multimodal embedding of ellipsis would be a director saying much too slow
while watching a film take on the monitor at the film set. Her visual orientation
to the screen, knowledge about some prior interaction with the camera woman,
and the routine practice of the director reviewing film takes in order to decide if
they are acceptable will be required in order to understand that the turn is taken
to mean that the camera moves much too slow and that the take needs to be redone. As in this example, coordination of gaze between speaker and hearer, visually available context, and gesture mostly provide for the thematic referential
aspects of communication, whereas linguistic communication is restricted to
rhematic predications expressing some stance or request for action. However,
gesture can also be used to elaborating a predication, as in the case of modal in-
443
dexicals like germ. so and engl. like (Streeck 2002). They need to be accompanied by an illustrative gesture in order to clarify which semantic aspect (like
size, form, direction of the referent, kind of movement) is indexed by the
modal expression.
3.8.
Grammatical variation relating to genre and medium
There are huge differences between spoken and written varieties in virtually
all areas of grammar, e.g., regarding the use of (complex) clause patterns (such
as extraposition, subjunction), tense and aspect marking, valency, or discourse
markers (Biber, Johansson and Leech 1999; Bublitz, Lenk and Ventola 1999). A
closer look, however, reveals that there is no simple dichotomy between spoken
and written language. Grammar is sensitive to fine-grained differences of register
(e.g. Biber 2006), genre and medium. During the last decades, micro grammars
related to genres of computer mediated communication (CMC) have developed.6
In genres and technical forms of communication with a high degree of interactivity like chats and SMS in particular, new grammatical conventions have emerged
for the display of emotions (multiplication of vowels and diacritics, expressive
use of capital letters, emoticons), the mimicry of nonverbal communication (root
inflectives known from comics such as germ. heul, grins), the adaptation of writing to oral interaction (eye dialect, dialectal and substandard spellings: boyz,
wuddya; use of discourse markers and non-standard grammar), and we-codes,
which are intransparent to out-groups (e.g. by using acronyms like lol, CTN, 4U;
Schlobinski 2006). In CMC in general, standards of linguistic correctness have
become liberalized, giving way to innovative playful, expressive, and economical uses.
In print and internet media, texts often combine with the visual design of pictures (Kress and van Leeuwen 2006; Stckl 2004) and other semiotic modes such as
sounds, animations, and movies (e.g., in vlogs). On websites, e.g., large parts of
texts are elliptical, consisting of free nominals which are not integrated into any
syntactic structure. However, their meaning is supplied by the visual line-up they
are embedded in and by the users pragmatic knowledge about how to use them on
behalf of there being links to other pages, forms to be filled out, etc. (Schmitz 2010).
3.9.
Pragmatic factors of grammaticalization
Language-internal and social factors have already been discussed for a long time as
determinants for linguistic change (see Aitchison 2001). Over the last three decades, inferential, discursive, and usage-based processes have increasingly become
recognized as major factors promoting linguistic change. A core area of change is
grammaticalization. The term refers to the emergence of new grammatical forms
(from old ones); more specifically, grammaticalization involves loss of lexical
444
Arnulf Deppermann
meaning (semantic bleaching), gain of functional meaning (e.g., word-class conversion), loss of phonological substance, restricted syntactic distribution or morphological incorporation, and may involve the development of a discursive pattern
into a grammatical convention. Grammaticalization has been found to be a probably unidirectional process, which runs along clines of increasing grammaticalization (Haspelmath 1999; Hopper and Traugott 2003: Ch.5). One such cline, which
can be attested for several languages, is the following diachronic path: content
word f function word (analytic) syntax (e.g., PRO) f agglutinative morphology
(clitization) f flexive morphology (e.g., subject marker) f zero.
Grammaticalization can be motivated by various strategies and preferences of
language users:
Articulatory economy leads to phenomena like reduction/attrition, clitization,
assimilation/co-articulation (Lass 1987).
Processing economy can be seen at work in heavy NP-shifts (Wasow 2002),
which correspond to the Gesetz der wachsenden Glieder (law of increasing
constituents, Behaghel 1909).
The enrichment of expressive power by innovations.
The desire to communicate unambiguously leads to, e.g, the decay of homonyms.
Iconicity (Haiman 1985) is a cognitive motivation which builds on principles
of naturalness (e.g., constructing superlatives by reduplication: chef chef;
again and again)
The desire to accrue social prestige, to gain the status of a member and to identify with a social group leads to the use of linguistic practices of prestigious ingroups (Labov 2001; LePage and Tabouret-Keller 1985). Innovations in youth
language provide paramount examples.
To be sure, linguistic change in general and grammaticalization in particular do
not only involve change of forms, but also semantic and pragmatic change. A common mechanism of change is metaphorical extension, which works by analogical
transfer of concepts from a source to a target domain. Metaphoric change can be
observed in many languages for deictics, where a metaphorical transfer of expressions for body parts and bodily orientation to spatial reference in general
comes about (Heine, Claudi and Hnnemeyer 1991; Heine 1997). A well attested
path of metaphorical change in the meaning of prepositions and connectives is spatial f temporal f causal f epistemic/speech act/discursive/textual (Sweetser
1990; Haspelmath 1997). The general dynamics of metaphorical extension consists in using anthropocentric, visually based, concrete concepts for more abstract,
discursive, and subjective entities. The diachronic change of, e.g., connectives
(like while, since, after (all)) thus shows increasing subjectification and abstraction
in terms of being used for displaying speakers attitudes, epistemic perspectives
and expectations (Sweetser 1990).
445
While metaphor operates on the paradigmatic level (supplanting one meaning

with another), metonymically based change rests on reanalysis of syntagmatic relationships, which in addition often become contracted over time (Hopper and
Traugott 2003: Ch.4). In metonymic change, a lexical item or (parts of) a construction adopt meanings from their collocational contexts. Metonymic change is more
gradual than the more abrupt and drastic metaphoric change, because what develops is rather informative strengthening or enrichment (i.e., an implicature is added;
Ariel 2008:163) and successive transformation. A much discussed example of metonymic change is the development of to be going to (e.g., Hopper and Traugott
2003: 8893; Bybee 2006): Originally, it denoted physical movement to some
place; later it became a purpose-construction, referring to the intention of an actor;
nowadays, it has become a generalized future construction, which may also be
used with inanimate, non-intentional subjects. Morphologically, there is a reanalysis of going to, which is no longer interpreted as a progressive form of a full verb,
but as a tense marker, and there is phonetic reduction (gonna), which attests to the
unit status of the construction. This reduction is absent in the progressive use (Im
going to a party vs. * Im gonna a party).
In many cases, implicatures (or invited inferences, see Traugott 2004) are at the
beginning of grammaticalization. The starting point are conversational implicatures which heavily depend on specific supportive contexts. Gradually, they become generalized implicatures, i.e., prototypical meanings (Traugott 2004). Grammaticalization has come about when the new meaning is firmly established, so that
there are (at least) two meanings, i.e., the item/structure has become ambiguous.
The next step then is reorganization and differentiation of the two meanings in
terms of the development of specialized distributions and collocational potentials,
differences in potentials for TMA-marking, etc. (Diewald 2006b).
While cognitive motivations and interpretations can be seen as important
driving forces for grammaticalization, cognitive factors as such are never directly
responsible for language change. They can leave their fingerprint on language
structure only via their discursive realization. Moreover, discourse is not only a
medium for the linguistic reflection of cognitive preferences. There are genuine
discursive motivations for the development of grammatical structures. Probably
all successful grammaticalization paths had previously led to salient discourse
patterns, themselves only necessary (but not sufficient) conditions for conventionalization (Ariel 2008:150). Except for language planning, coining of terminology, and some innovations promoted by the media and by subcultural actors,
people only rarely intend to promote language change. And even if there is some
definite intention of innovation, its success fully depends on how it is taken up
and spread by recipients, both in terms of the interpretation of the innovation and
the frequency of use. At least in the early stages of innovation, intentional strategies of people pursuing communicative goals often are responsible for language
change, which, however, as such mostly happens as an unintended by-product of
446
Arnulf Deppermann
these strategies. Keller (2004) terms such changes invisible hand-processes.

Sometimes, the long-term consequences of the factual collaboration of intentional strategies of a community of linguistic actors can even lead to a result
which is quite the opposite of what was originally intended. An example is the
dynamics of diachronic change of terms for women in German (Keller 2004):
The collective adherence to a strategy to select distinctive terms for addressing
and talking about women eventually leads to a pejorization of these terms (Weib,
Weibsbild, Frauenzimmer): Frequent use makes them lose their distinctive potential, thus speakers strive to replace them by more marked and highly valued
terms, which get increasingly used as a consequence and thus also lose their distinctive value, etc. (The same process can be observed in an accelerated pace
with innovations in youth language, e.g., regarding evaluative predicates). So, in
the long run politeness and distinction strategies have the inverse impact on the
terms used.
Other discursive preferences are conservatism (in a more macro view) and, in
terms of a micro view of dialogical sequences, priming and resonance (Pickering
and Garrod 2004; Brennan and Clark 1996). They account for the preference to follow precedents, which thus can become canonical. These strategies of repetition
are most promising in terms of securing understanding and building on intersubjectivity achieved before (Feilke 1996). They provide for a systematic interactive
motivation of the grammaticalization of options which initially may simply have
been chosen arbitrarily among a set of possible options (see Clark and WilkesGibbs 1986).
Patterns of use lead to the gradual emergence of new grammatical patterns
from non-innovative, grammatical usage. The emergence of discourse particles
(well, but, though) from connectives and other discourse markers (like yknow, verstehste do you understand, voil ) from mental verbs are examples of gradual
change. Precursor for the development of discourse markers from mental verbs is
the emergence of the construction [matrix sentence with mental verb + dependent
main clause] (Auer 1998), i.e., a semantically dependent clause follows which
does not exhibit any marker of subordination. The matrix sentence can increasingly change into a semantically bleached discourse marker, which then, in later
phases, may fully grammaticalize in terms of adopting a specific distribution (often
syntactically unembedded in turn-initial or turn-final position) and retaining only
light lexicosemantic content (if any) (Thompson 2002; Imo 2007; Karkkinen
2003). Dialogical patterns can also be a source of grammaticalization. A case in
point are clause combinations which arise from interactional sequences and transform them into grammatical formats to be used by individual speakers in single
turns (Givn 1979). An example is the emergence of conditionals from a questionanswer-response sequence of three turns:
447
(30) Grammaticalization of interactional sequences in complex sentences: Conditionals

A: Yes/No-Question (Will you help me?)
B: Agreement (Yes.)
A: Consequence (Then Ill give you two dollars.)
f A: Protasis (If youll help me),
Apodosis (then Ill give you two dollars.)
Discursive and cognitive factors are often intertwined in processes of grammaticalization. The category noun exemplifies how cognitive concerns are mediated
by discursive ones. Cognitive grammar claims that the grammatical category
noun has a schematic semantics, which consists in conceptualizing referents as
things which are defined as any product of grouping and reification (Langacker 2008:105). The analysis of the realization of nouns in discourse, however,
shows that prototypical representatives of nouns, which are count nouns and exhibit all grammatical features which are regarded as typical of nouns (determiners,
gender-, number-, and case-marking, etc.) introduce discourse topics which are
manipulable, i.e., available for subsequent co-reference (Hopper and Thompson
1984). Grammatical categoriality thus is achieved by discursive patterns, not
by mental conceptualization alone. Discursive considerations can even override
cognitive motivations. In the overwhelming majority of cases, causal adverbial
phrases (with because, weil, par-ce que etc.) follow their nucleus assertion, although adherence to an iconicity principle would predict that they precede because
causes are prior to effects (Ariel 2008: 175 f.). However, it is neither the way the
world is (temporal order) nor cognitive preference for iconic representation, but
the discursive function of dealing with the recipients actual or anticipated disaffiliation by supplying justifications, causes, etc. (see Ford 1993), which motivates
the syntactic distribution of causal adverbial phrases.
According to usage-based approaches to linguistic change, frequency of use of
linguistic forms is the driving force in grammaticalization (Bybee and Hopper
2001; Bybee 2006, 2007). Neither morphological nor semantic innovation needs to
be the starting point for change, but it is the preferential, salient discourse pattern
(Ariel 2008), which, via routinization, finally becomes entrenched as a cognitive
unit (cf. Langacker 2000). An example of this is the cline of PRO becoming clitic,
then morphosyntactic marking and finally zero. Another case in point is the increasing phonetic reduction of the French negative pre-verbal particle ne, so that
only the post-verbal pas is retained as marker of sentential negation. Entrenchment
has effects of its own, phonological reduction (leading to clitization, assimilation/
fusion, and elision; see above) being the most prominent one. Frequency has also
two other effects (Bybee 2006): conservation, and autonomy. The conservation effect is evidenced by the resistance of irregular forms to change because of their
high token frequencies. A case in point are irregular English past tense forms like
448
Arnulf Deppermann
was, came, had. They survive, while verbs with low token frequency successively
succumb to the high type-frequency of the regular, schematically derived past
tense morphology with -ed inflection (ex. wept f weeped). Autonomy (or emancipation) means the re-analysis of old and the emergence of new constructions,
which become used as unanalyzed units. They change and sometimes even cut
across syntactic categories, as could be seen in the case of gonna or the emergence
of discourse markers. Autonomy and unit status are also evidenced by increasing
speed of production and comprehension, prosodic segmentation from the discursive context and syntactic parsing as a unit. However, it is important to note that although frequency may play a major role in (the stabilization of) grammaticalization, this is not necessarily its cause. Rather, we need to ask why some pattern
becomes frequent. Motivations like economy (against redundancy of coding and
high predictability/inferrability in context) and others (see above) may cause or at
least support high frequency of use. Also, frequency does not automatically lead to
grammaticalization, because not all frequent patterns change, and the effect of frequency is mediated by factors like genre, speakers self-monitoring, linguistic
ideologies, etc.
3.10.
Acquiring grammatical constructions
Usage-based approaches show that the process of acquiring increasingly abstract

constructions rests on interactional experience and general cognitive abilities (Tomasello 2003; 2006). According to Tomasello (1999, 2003; Tomasello et al. 2005),
anthropological, social and cognitive prerequisites are the driving forces for the
childs construction of a language: the ascription of communicative (Gricean) intentions to interlocutors, the coordination of joint attention (Bruner 1983, Clark
2001), intersubjective cooperation (Tomasello 2008), and cognitive capabilities
like categorization, inference and induction of schemata, statistical comparison
and analogy, which are used for the identification of patterns of experience, action,
and linguistic practice. In contrast to Chomskys assumption about the poverty of
the stimulus, which is said to account for the need for a specialized language acquisition device, social interaction in terms of situated routines and modelling and
correction by care-givers in child-directed speech have also been shown to be of
major importance for language-learning (Cameron-Faulkner, Lieven and Tomasello 2003). Tomasello (2006: 258) questions the necessity to assume a languagespecific capacity, as Universal Grammar does, because children acquire the more
regular and rule-based constructions in the same way they acquire the more arbitrary and idiosyncratic constructions. Instead of applying abstract class-specific
rules to words according to their membership in syntactic categories (cf. Pinker
1999), knowledge about abstract syntactic structures emerges from lexical strings
by applying universal cognitive mechanisms of learning with the aid of the specific
pragmatic structures of context. The first schematic constructions to develop are
449
two-word pivot constructions, where a lexically fixed pivot is combined with various other items. The next step are item-based constructions and constructional islands (Tomasello 2003). These are constructional frames associated with specific
lexical items according to the frequency of their occurrence. They contain categorical slots for arguments and first syntactic markers (like word-order, morphosyntactic markers, prepositions). Children start to use and understand frequent
verbs in the context of various constructional frames, while less frequent verbs are
mastered only in a smaller number of constructional contexts (e.g., cut is only used
with direct object, while draw is also used with additional locative, instrumental or
beneficient prepositional phrases). So, syntactic rules do not seem to be categorically available in abstraction from lexicalization and context, but they are acquired
only incrementally by a process of increasing schematization. This process rests on
the discovery of analogies and variational patterns by comparison (see also Clark
1992 for the heuristics of conventionality and contrast). It lays the base for inferring slot-filling categories, which are increasingly abstracted and generalized to
fully schematic, abstract constructions via induction. The discovery of analogies
does not merely rest on perceptual features and cognitive operations over formal
categories, but more importantly on functionally based distributions, which cooccur with recurrent patterns of action (like object transfer or caused motion),
and interactional participation in action sequences (like referring and naming,
question-answer, narratives).
4.
Future prospects
As the paradigm of armchair linguistics gives way to empirical research,

scholars tend to conceive of pragmatics in a more comprehensive way, including
areas of research which were formerly assigned to text-, socio-, psycho- or discourse linguistics. Theoretical notions and methods originating from these areas
are increasingly integrated in pragmatic theorizing. In the last decade, a variety of
new empirical methods has emerged. They provoke changes in the view of the relationship between pragmatics and grammar and lead to new questions.
The increasing availability of large machine-readable (still overwhelmingly
written) corpora and the use of procedures of data-mining allows for the discovery of preferential patterns of co-occurrence and of contextual conditions of
use by statistical analyses (Stefanowitsch and Gries 2008). Such usage-based
studies strongly favor probabilistic and context-sensitive models of grammar
instead of universalistic rule-based approaches.
Video recordings of embodied multimodal interaction are prerequisites for
understanding the fine-grained sequential and simultaneous coordination between grammar, prosody, gesture, facial expression, and other body movements in real time. They also help to elucidate how the roots of linguistic rou-
450
Arnulf Deppermann
tines may lie in gestural routines, both ontogentically and phylogenetically

(Tomasello 2008).
Psycholinguistic experiments using parameters like latency of reaction are
used to test hypotheses about the psychological reality of alleged relationships
between (syntactically based) semantic vs. inferred meanings in terms of cognitive processing (Meibauer and Steinbach 2009).
Researchers have started to use physiological and neurobiological measures
like event-related potentials (Kutas and Federmeier 2007) and brain-imaging
(Bornkessel-Schlesewsky and Friederici 2007) as evidence for assumptions
about mental processing and cerebral localization of phenomena located on the
grammar-pragmatics interface.
The process of the acquisition of grammatical structures based on some basic
cognitive capabilities and discursive experience has started to be modelled by
learning robots, e.g., in the approach of Fluid Construction Grammar (Steels
and de Beule 2006).
Changes of the subject matter of linguistic pragmatics, which are due to sociocultural and technological developments, also call for revised models and raise
new questions. New internet and communication technologies, which are creatively adopted by users, lay the grounds for a seemingly endless rush of new
genres, multimodal configurations and modes of interactivity, in which new grammatical opportunities and routines in new communities of practice emerge. Processes of migration and globalization lead to revisions and re-formations of speech
communities, including the emergence of new linguistic varieties and inter-cultures of mixed codes which may become conventionalized.
Notes
1. See already the seminal paper by Fillmore, Kay and OConnor (1988).
2. Cf. ditransitive vs. to-dative (Goldberg 1995 and 2006) and argument omission (Goldberg 2005).
3. Ochs, Schegloff and Thompson (1996), Selting and Couper-Kuhlen (2001), Barth-Weingarten (2008).
4. Cf. e.g. Ono and Thompson (1995), Deppermann (2006, 2007), Gnthner (2006) and Imo
(2007).
5. Auer (1996a and 2007), Schegloff (1996), Ford, Fox and Thompson (2002), CouperKuhlen and Ono (2007a).
6. Crystal 2006; see also vol. 9 of the Handbook of Pragmatics handbook series: Pragmatics of Computer-mediated Communication, ed. by Susan C. Herring, Dieter Stein and
Tuija Virtanen.
451
References
Aitchison, Jean
2001
Language change: Progress or decay? 3rd ed. Cambridge: Cambridge University Press.
Ariel, Mira
1990
Accessing noun phrase antecedents. London: Routledge.
Ariel, Mira
2008
Pragmatics and grammar. Cambridge: Cambridge University Press.
Auer, Peter
1996a
On the prosody and syntax of turn-continuations. In: Couper-Kuhlen, Elizabeth and Selting, Margret (eds.), Prosody in conversation. Interactional
studies, 57100. Cambridge: Cambridge University Press.
Auer, Peter
1996b
The pre-front field in spoken German as a grammaticalization position. Pragmatics 6(3): 295322.
Auer, Peter
1998
Zwischen Parataxe und Hypotaxe: abhngige Hauptstze im Gesprochenen
und Geschriebenen Deutsch. Zeitschrift fr Germanistische Linguistik 26(3):
284307.
Auer, Peter
2005
Projection in interaction and projection in grammar. Text 25(1): 736.
Auer, Peter
2007
Why are increments such elusive objects? Pragmatics 17(4): 647658.
Auer, Peter
2009
On-line syntax: Thoughts on the temporality of spoken language. Language
Sciences 31: 113.
Auer, Peter and Stefan Pfnder (eds.)
2010
Emergent constructions. Berlin: de Gruyter.
Barlow, Michael and Suzanne Kemmer (eds.)
2000
Usage-based models of language. Stanford CA: Center for the Study of Language and Information.
Barth-Weingarten
2008
Interactional linguistics. In: Gerd Antos, Eija Ventola and Tilo Weber (eds.),
Handbook of Applied Linguistics. Vol. 2: Interpersonal Communication,
Barth-Weingarten, Dagmar and Couper-Kuhlen, Elizabeth
2002
On the development of final though: A case of grammaticalization? In: Ilse
Wischer and Gabriele Diewald (eds.), New reflections on grammaticalization,
Behaghel, Otto
1909
Beziehungen zwischen Umfang und Reihenfolge von Satzgliedern. Indogermanische Forschungen 25: 110142.
Biber, Douglas, Stig Johansson and Geoffrey Leech
1999
Longman grammar of spoken and written English. London: Longman.
Biber, Douglas
2006
University Language. Amsterdam: Benjamins.
452
Arnulf Deppermann
Bod, Rens, Jennifer Hay and Stefanie Jannedy (eds.)

2003
Probabilistic linguistics. Cambridge, MA: MIT Press.
Bornkessel-Schlesewsky, Ina and Angela D. Friederici
2007
Neuroimaging studies of sentence and discourse comprehension. In: Gareth
Gaskell (ed.), The Oxford handbook of psycholinguistics, 407424. Oxford:
Brennan, Susan E. and Herbert H. Clark
1996
Conceptual pacts and lexical choice in conversation. Journal of experimental
psychology: Learning, memory, and cognition 22: 14821493.
Bruner, Jerome
1983
Childs talk: Learning to use language. New York: Norton.
Bublitz, Wolfram, Uta Lenk, Uta and Eija Ventola (eds.)
1999
Coherence in spoken and written discourse. Amsterdam: Benjamins.
Bhler, Karl
1934
Sprachtheorie. Jena: Gustav Fischer.
Bring, Daniel
2007
Semantics, intonation and information structure. In: Gillian Ramchand and
Charles Reiss (eds.), The Oxford handbook of linguistic interfaces, 445474.
Oxford: Oxford University Press.
Bybee, Joan
2006
From usage to grammar: The minds response to repetition. In: Language
82(4): 711733.
Bybee, Joan
2007
Frequency and the organization of language. Oxford: Oxford University
Press.
Bybee, Joan and Paul J. Hopper (eds.)
2001
Frequency and the emergence of linguistic structure. Amsterdam: Benjamins.
Cameron-Faulkner, Thea, Elena Lieven and Michael Tomasello
2003
A construction based analysis of child directed speech. In: Cognitive Science
27: 843873.
Chafe, Wallace
1994
Discourse, consciousness, and time. Chicago IL: University of Chicago Press.
Chomsky, Noam
1965
Aspects of the theory of syntax. Cambridge, MA: MIT Press.
Chomsky, Noam and Howard Lasnik.
1993
The theory of principles and parameters. In: Jacobs, Joachim, von Stechow,
Arnim, Sternefeld, Wolfgang and Theo Vennemann (eds.), Syntax. Vol. 1,
Clark, Eve V.
1992
Conventionality and contrast: Pragmatic principles with lexical consequences.
In: Eva Feder Kittay and Adrienne Lehrer (eds.), Frames, fields, and contrasts:
New essays in semantic and lexical organization, 171188. Hillsdale, NJ:
Lawrence Erlbaum.
Clark, Eve
2001
Grounding an attention in the acquisition of language. In: M. Andronis,
C. Ball, H. Elston & S. Neuvel (eds.), Papers from the 37th meeting of Chicago
Linguistic Society. Vol. 1, 95116. Chicago: Chicago Linguistic Society.
453
Clark, Eve
2004
Pragmatics and language acquisition. In: Lawrence Horn and Gregory Ward
(eds.), Handbook of pragmatics, 562577. Oxford: Blackwell.
Clark, Herbert H.
1977
Bridging. In: Philip Johnson-Laird and Peter C. Wasow (eds.), Thinking,
Clark, Herbert H. and Deanna Wilkes-Gibbs
1986
Referring as a collaborative process. Cognition 22: 139.
Couper-Kuhlen, Elizabeth and Tsuyoshi Ono
2007a
Turn continuation in cross-linguistic perspective. Special Issue of Pragmatics
17(4).
2007b
Incrementing in conversation. A comparison of practices in English, German, and Japanese. Pragmatics 17(4): 513552.
Croft, William
2001
Radical construction grammar. Oxford: Oxford University Press.
Croft, William
2008
Construction Grammar. In: Dirk Geeraerts and Hubert Cuyckens (eds.), Handbook of Cognitive Linguistics, 463508. Oxford: Oxford University Press.
Croft, William and A. Cruse
2004
Cognitive linguistics. 225327. Cambridge: Cambridge University Press.
Crystal, David
2006
Language and the Internet. 2nd ed. Cambridge: Cambridge University.
Deppermann, Arnulf
2006
Construction Grammar Eine Grammatik fr die Interaktion? In: Arnulf Deppermann, Reinhard Fiehler and Thomas Spranz-Fogasy (eds.), Grammatik und
Interaktion, 4366. Radolfzell: Verlag fr Gesprchsforschung. Available at:
www.verlag-gespraechsforschung.de/2006/deppermann.htm.
Deppermann, Arnulf
2007
Grammatik und Semantik aus gesprchsanalytischer Sicht. Berlin: de Gruyter.
Deppermann, Arnulf
2010
Constructions vs. lexical items as sources of complex meanings. A comparative study of constructions with German verstehen. In: Peter Auer and Stefan Pfnder (eds.), Emergent constructions. Berlin: de Gruyter.
Diewald, Gabriele
2006a
Discourse particles and modal particles as grammatical elements. Approaches
to discourse particles. In: Kerstin Fischer (ed.), Approaches to discourse particles, 403425. Amsterdam: Elsevier.
Diewald, Gabriele
2006b
Context types in grammaticalization as constructions. In: Constructions HSV
1-9/2006 (Special Volume). Available at http://www.constructions-online.de/
articles/specvol1/686.
Dik, Simon C.
1997
The Theory of Functional Grammar. 2 vols. Berlin: de Gruyter.
Du Bois, John
2003
Discourse and Grammar. In: Michael Tomasello (ed.), The new psychology of
language. Vol. 2, 4787. Mahwah, NJ: Lawrence Erlbaum.
454
Arnulf Deppermann
Du Bois, John, Lorraine E. Kumpf and William J. Ashby (eds.)

2003
Preferred argument structure. Amsterdam: Benjamins.
Edwards, Derek
1997
Discourse and cognition. London: Sage.
Eggins, Suzanne
1994
An introduction into Systemic Functional Linguistics. London: Continuum.
Ehlich, Konrad
1991
Funktional-pragmatische Kommunikationsanalyse. In: Dieter Flader (ed.),
Verbale Interaktion, 127143. Stuttgart: Metzler.
1985
Frames and the semantics of understanding. Quaderni di Semantica 6(2):
222254.
Fillmore, Charles J., Paul Kay and Mary Catherine OConnor
1988
Regularity and idiomaticity in grammatical constructions: The case of let
alone. Language 64: 501538.
Feilke, Helmuth
1996
Sprache als soziale Gestalt. Frankfurt/Main: Suhrkamp.
Feilke, Helmuth
2004
Kontext Zeichen Kompetenz. In: Kathrin Steyer (ed.), Wortverbindungen
mehr oder weniger fest, 4164. Berlin: de Gruyter.
Fischer, Kerstin (ed.)
2006
Approaches to discourse particles. Amsterdam: Elsevier.
Ford, Cecilia E.
1993
Grammar in interaction. Adverbial clauses in American English conversations. Cambridge: Cambridge University Press.
Ford, Cecilia E., Barbara A. Fox and Sandra A. Thompson
2002
Constituency and the grammar of turn increments. In: Cecilia E. Ford, Barbara
A. Fox and Sandra A. Thompson (eds.), The language of turn and sequence,
2003
Social interaction and grammar. In: Michael Tomasello (ed.), The New Psychology of Language. Vol. 2, 119143. Mahwah NJ: Lawrence Erlbaum.
Fox, Barbara A.
2007
Principles shaping grammatical practices: an exploration. Discourse Studies
9(3): 299318.
Fries, Norbert
1983
Syntaktische und semantische Studien zum frei verwendeten Infinitiv und zu
verwandten Erscheinungen im Deutschen. Tbingen: Narr.
Gardner, Rod
2001
When listeners talk: Response tokens and listener stance. Amsterdam: Benjamins.
Givn, Talmy
1979
From discourse to syntax. Grammar as a processing strategy. In: Talmy Givn
(ed.), Syntax and semantics. Vol. 12: Discourse and syntax, 81112. London:
Academic.
Goldberg, Adele
1995
Constructions. A construction grammar approach to argument structure. Chicago IL: University of Chicago Press.
455
Goldberg, Adele
1998
Patterns of experience in patterns of language. In: Michael Tomasello (ed.),
The new psychology of language, 203220. Mahwah, NJ: Lawrence Erlbaum.
Goldberg, Adele
2004
Pragmatics and argument structure. In: Lawrence Horn and Gregory Ward
Goldberg, Adele
2005
Argument realization. The role of constructions, lexical semantics and discourse factors. In: Jan-Ola stman and Mirjam Fried (eds.), Construction
grammar, 1743. Amsterdam: Benjamins.
Goldberg, Adele
2006
Constructions at work: the nature of generalization in language. Oxford: Oxford University Press.
Goodwin, Charles
2003
Pointing as situated practice. In: Sotaro Kita (ed.), Pointing: Where language,
culture and cognition meet. 21741. Mahwah, NJ: Lawrence Erlbaum.
Goodwin, Charles and Marjorie Harness Goodwin
1987
Concurrent operations on talk: Notes on the interactive organization of assessments. IPRA Papers on Pragmatics 1(1): 154.
Green, Georgia M.
2004
Some interactions of pragmatics and grammar. In: Lawrence Horn and Gregory Ward (eds.), Handbook of pragmatics, 407426. Oxford: Blackwell.
Grice, H. Paul
1975
semantics. Vol. 3, 4158. New York: Academic Press.
Gundel, Jeanette and Thorstein Fretheim
2004
Topic and focus. In: Lawrence Horn and Gregory Ward (eds.), Handbook of
pragmatics, 175196. Oxford: Blackwell.
Gundel, Jeanette, Nancy Hedberg and Ron Zacharsky
1993
Givenness, implicature, and the form of referring expressions in discourse.
Gnthner, Susanne
2006
Von Konstruktionen zu kommunikativen Gattungen. Die Relevanz sedimentierter Muster fr die Ausfhrung kommunikativer Aufgaben. Deutsche
Sprache 1(2), 173190.
Gnthner, Susanne
2008
Projektorkonstruktionen im Gesprch: Pseudoclefts, die Sache ist-Konstruktionen und Extrapositionen mit es. Gesprchsforschung 9: 86114. Available
at: http://www.gespraechsforschung-ozs.de/heft2008/ga-guenthner.pdf.
Gnthner, Susanne and Paul J. Hopper
2010
Zeitlichkeit & sprachliche Strukturen: Pseudoclefts im Englischen und Deutschen. Gesprchsforschung 11: 128. Available at:
www.gespraechsforschung-ozs.de/heft2010/ga-guenthner.pdf
Haiman, John (ed.)
1985
Iconicity in syntax. Amsterdam: John Benjamins.
1985
An Introduction into Functional Grammar. London: Edward Arnold.
456
Arnulf Deppermann
Haspelmath, Martin
1997
From space to time. Temporal adverbials in the worlds languages. Mnchen:
Lincom.
Haspelmath, Martin
1999
Why is grammaticalization irreversible? In: Language 37: 10431068.
Hayashi, Makoto
2003
Joint utterance construction in Japanese conversation. Amsterdam: Benjamins.
Heine, Bernd
1997
Cognitive foundations of grammar. Oxford: Oxford University Press.
Heine, Bernd, Ulrike Claudi and Friederike Hnnemeyer
1991
Grammaticalization: A conceptual framework. Chicago: University of Chicago Press.
Hengeveld, Kees and J. Lachlan Mackenzie
2008
Functional Discourse Grammar. Oxford: Oxford University Press.
Hoffmann, Ludger
2001
Einleitung. In: Ludger Hoffmann (ed.), Funktionale Syntax, 117. Berlin: de
Gruyter.
Hopper, Paul J.
1998
Emergent grammar. In: Michael Tomasello (ed.), The new psychology of language, 155175. Mahwah, NJ: Erlbaum.
Hopper, Paul J.
2004
The openness of grammatical constructions. Papers from the 40th regional
meeting of the Chicago Linguistic Society 40: 239256.
Hopper, Paul J. and Elizabeth Closs Traugott
2003
Grammaticalization. 2nd ed. Cambridge: Cambridge University Press.
Hopper, Paul J. and Sandra A. Thompson
1984
The discourse basis for lexical categories in Universal Grammar. In: Language
60: 703752.
Hopper, Paul and Sandra A. Thompson
2008
Projectability and clause combining in interaction. In: Ritva Laury (ed.),
Crosslinguistic studies of clause combining, 99123. Amsterdam: Benjamins.
Imo, Wolfgang
2007
Konstruktionsgrammatik und Gesprochene Sprache-Forschung. Tbingen:
Niemeyer.
Jefferson, Gail
1973
A case of precision timing in ordinary conversation: Overlapped tag-positioned address terms in closing sequences. Semiotica 9: 4796.
Karkkinen, Elise
2003
Epistemic stance in English conversation. Amsterdam: Benjamins.
Kay, Paul
2004
Pragmatic aspects of grammatical constructions. In: Lawrence Horn and Gregory Ward (eds.), Handbook of pragmatics, 675700. Oxford: Blackwell.
Keller, Rudi
2004
On language change. London: Routledge.
Kress, Gunter and Theo van Leeuwen
2001
Multimodal discourse. London: Edward Arnold.
457

2006
Reading images: The grammar of visual design. London: Routledge.
Krifka, M.
2007
Basic notions of information structure. In: Caroline Fry, Gisbert Fanselow
and Manfred Krifka (eds.), Interdisciplinary studies on information structure
(ISIS) 6. Working Papers of the SFB 632, 1356. Potsdam: Universittsverlag.
Kutas, Marta and Kara D. Federmeier
2007
Event-related brain potential (ERP) studies of sentence processing. In: Gareth
Labov, William
2001
Principles of linguistic change. Vol. 2. Oxford: Blackwell.
Lakoff, George
1987
Women, fire and dangerous things. What categories reveal about the mind.
Chicago IL: University of Chicago Press.
Lambrecht, Knud
1987
On the status of SVO sentences in French discourse. In: Russell S. Tomlin
(ed.), Coherence and grounding in discourse, 217262. Amsterdam: Benjamins.
Langacker, Ronald W.
1987
Foundations of cognitive grammar. Vol. 1: Theoretical prerequisites. Stanford
CA: Stanford University Press.
2000
A dynamic usage-based model. In: Michael Barlow and Suzanne Kemmer
(eds.), Usage-based models of language, 160. Stanford, CA: Center for the
Study of Language and Information.
2001
Discourse in cognitive grammar. Cognitive Linguistics 12(2): 143188.
2008
Cognitive grammar. Oxford: Oxford University Press.
Lass, Roger
1987
Historical linguistics and language change. Cambridge: Cambridge University Press.
LePage, Robert B. and Andre Tabouret-Keller
1985
Acts of identity: Creole-based approaches to language and ethnicity. Cambridge: Cambridge University Press.
Lerner, Gene
1991
On the syntax of sentences in progress. Language in Society 30: 441458.
Lerner, Gene
1996
On the semi-permeable character of grammatical units in conversation: Conditional entry into the turn space of another speaker. In: Elinor Ochs, Emmanuel
A. Schegloff and Sandra A. Thompson (eds.), Interaction and grammar, 238
Lerner, Gene
2004
The place of linguistic resources in the organization of talk-in-interaction:
Grammar as action in prompting a speaker to elaborate. Research on language
and social interaction 37(2): 151184.
458
Arnulf Deppermann
Liedtke, Frank
1998
Grammatik der Illokution. Tbingen: Narr.
MacWhinney, Brian and Elizabeth Bates
1989
The crosslinguistic study of sentence processing. Cambridge: Cambridge University Press.
Meibauer, Jrg and Markus Steinbach (eds.)
2009
Experimental-pragmatics/semantics. Amsterdam: Benjamins.
Mulder, Jean and Sandra A. Thompson
2008
The grammaticalization of but as a final particle in English conversation. In:
Ritva Laury (ed.), Crosslinguistic studies of clause combining, 179204.
Newmeyer, Frederick J.
1998
Language form and language function. Cambridge MA: MIT Press.
2003
Grammar is grammar and usage is usage. Language 79: 682707.
Norris, Sigrid
2004
Analyzing multimodal interaction. London: Routledge.
Ochs, Elinor, Emmanuel A. Schegloff, and Sandra A Thompson (eds.)
1996
Ono, Tsuyoshi and Sandra A. Thompson
1995
What can conversation tell us about syntax? In: Philip W. Davis (ed.), Alternative linguistics, 213271. Amsterdam: Benjamins.
Pekarek-Doehler, Simona
2010
Clause-combining and the sequencing of actions: projector fragments in
French conversation. In: Ritva Laury and Ryoko Suzuki (eds.), Subordination
in conversation. Crosslinguistic analysis of form-function mappings, Amsterdam: Benjamins.
Pickering, Martin J. and Simon C. Garrod
2004
Towards a mechanistic psychology of dialogue. Behavioral and Brain
Sciences 27: 169226.
Pinker, Steven
1999
Words and rules. New York: Morrow.
Prince, Ellen F.
1978
A comparison of WH-clefts and IT-clefts in discourse. In: Language 54: 883906.
Prince, Ellen F.
1981
Toward a taxonomy of given/new information. In: Peter Cole (ed.), Radical
pragmatics, 223255. New York: Academic Press.
Raymond, Geoffrey
2004
The stand-alone so in ordinary conversation. In: Research on Language and
Social Interaction 37(2): 185218.
Redder, Angelika
2008
Functional pragmatics. In: Gerd Antos and Eija Ventola (eds.), Handbook of
applied linguistics. Vol 2. Interpersonal communication, 133178. Berlin: de
Gruyter.
Sacks, Harvey and Emmanuel A. Schegloff
1979
Two preferences in the organization of reference to persons in conversation
and their interaction. In: George Psathas (ed.), Everyday language, 1521.
New York: Irvington.
459
Sacks, Harvey, Emmanuel A. Schegloff and Gail Jefferson

1974
A simplest systematics for the organisation of turn-taking in conversation.
Language 50(4): 696735.
Schegloff, Emmanuel A.
1996
Turn organization: One intersection of grammar and interaction. In: Elinor
Ochs, Emmanuel A. Schegloff and Sandra A. Thompson (eds.), Interaction
and grammar, 52133. Cambridge: Cambridge University Press.
Schiffrin, Deborah
1987
Discourse markers. Cambridge: Cambridge University Press.
Schlobinski, Peter (ed.)
2006
Von *hdl* bis *cul8er*. Sprache und Kommunikation in den neuen Medien.
Mannheim: Duden.
Schmid, Hans-Joerg
2001
Presupposition can be a bluff: How abstract nouns can be used as presupposition triggers. Journal of Pragmatics 33(10): 15291555.
Schmitz, Ulrich
2010
Schrift an Bild im World Wide Web. In: Arnulf Deppermann and Angelika
Linke (eds.), Sprache intermedial, 383418. Berlin: de Gruyter.
Searle, John R.
1969
Searle, John R. and Daniel Vanderveken
1985
Foundations of illocutionary logic. Cambridge: Cambridge University Press.
Selting, Margret
2000
The construction of units in conversational talk. Language in Society 29:
477517.
Selting, Margret and Elizabeth Couper-Kuhlen (eds.)
2001
Studies in interactional linguistics. Amsterdam: Benjamins.
Sorjonen, Marja-Leena
2001
Responding in conversation: A study of response particles in Finnish. Amsterdam: Benjamins.
Steels, Luc and Joachim de Beule
2006
A (very) brief introduction to Fluid Construction Grammar. Proceedings of the
3rd International Workshop on Scalable Natural Language (ScaNaLu06):
7380. Available at: http://www3.isrl.illinois.edu/~junwang4/langev/localcopy/pdf/Steels06Intro_FCG.pdf.
Stefanowitsch, Anatol and Stefan Th. Gries
2008
Corpora and grammar. In: Anke Ldeling and Merja Kyt (eds.), Corpus linguistics. Vol. 2, 933951. Berlin, New York: de Gruyter.
Stckl, Hartmut
2004
Die Sprache im Bild Das Bild in der Sprache. Berlin: de Gruyter.
Streeck, Jrgen
2002
Grammars, words, and embodied meanings. On the evolution and uses of so
and like. Journal of Communication 52(3): 581596.
Sweetser, Eve E.
1990
From etymology to pragmatics. Cambridge: Cambridge University Press.
Thompson, Sandra A.
2002
Object complements and conversation. Towards a realistic account. Studies
in Language 26(1): 124164.
460
Arnulf Deppermann
Thompson, Sandra A. and Elizabeth Couper-Kuhlen

2005
The clause as a locus of grammar and interaction. Discourse Studies 7(45):
481505.
Tomasello, Michael
1999
The cultural origins of human cognition. New York: Harvard University Press.
Tomasello, Michael
2003
Constructing a language: A usage-based theory of language acquisition,
94281. New York: Harvard University Press.
Tomasello, Michael
2006
Acquiring linguistic constructions. In: Deanna Kuhn and Robert S. Siegler
(eds.), Handbook of child psychology. Vol. 2, 255291. New York: Wiley.
Tomasello, Michael
2008
Origins of human communication. Cambridge MA: MIT Press.
Tomasello, Michael, Malinda Carpenter, Josep Call, Tanya Behne and Henrike Moll
2005
Understanding and sharing intentions: The origins of cultural cognition. Behavioral and Brain Sciences 28: 675691.
2004
Historical pragmatics. In: Lawrence Horn and Gregory Ward (eds.), Handbook
of pragmatics, 538561. Oxford: Blackwell.
Uhmann, Susanne
1991
Fokusphonologie. Eine Analyse deutscher Intonationskonturen im Rahmen der
nicht-linearen Phonologie. Tbingen: Niemeyer.
Ward, Gregory and Betty Birner
2004
Information structure and non-canonical syntax. In: Lawrence Horn and Gregory Ward (eds.), Handbook of pragmatics, 153174. Oxford: Blackwell.
Wasow, Tom
2002
Postverbal behavior. Stanford: CSLI.
16.

John Saeed
1.
Introduction
In this article we focus on some relatively recent ideas and discussion about the
relationship between pragmatics and semantics. As a result we do not seek to characterize either enterprise comprehensively, or symmetrically, even if such were
possible in this limited space. Hence our discussion omits consideration of major
concerns of both: for pragmatics, for example, of the social and cultural dimensions of language use. In particular speech-act theorists are underrepresented, despite the importance to pragmatics of their concern with social institutions, conventions and language activities.1
Characterizing the relationship between pragmatics and semantics is made difficult by the differences in views of these fields in different linguistic theories.
Moreover, other disciplines are keen to explore the same territory. Philosophers
Lepore and Smith, for example, in prefacing a handbook of the philosophy of language, state aims that many linguists would recognize:
[] the philosophy of language now focuses on its primary concern: the nature of natural language and the extraordinary capacity of human beings to use it to express their
thoughts about the world and other subject matters. The way language works, how specific linguistic devices function to achieve their effects, how we come to know these
properties of expressions, and how we exploit them in our talk: all this is pursued by
contemporary philosophers of language. (2006: viii)
Looking the other way across the disciplinary fence, linguists may profess scepticism about the philosophical method, alleging a dependence on introspection, a
concentration on a few European languages and scant attention to other linguistic
levels such as phonology and syntax. However, the views of Frege, Russell, the
early Wittgenstein, Quine, Davidson and other philosophers have been of fundamental importance in many linguistic theories of semantics, in particular in pursuit
of the idea that languages are, or can be profitably viewed as, formal systems. As
will be clear from other contributions to this handbook, the ideas of the later
Wittgenstein, Grice, Austin, Strawson and Searle have been fundamental in the development of pragmatics.
More generally, the influence of philosophy on linguistic semantics can be seen
in the search for a role for propositions in the representation of sentence meaning.2
The idea that semantics is concerned with representations is widespread, even in
functional approaches, along with the associated idea that pragmatics is then concerned with interpersonal action.3 Problems with determining how semantic repre-
462
John Saeed
sentations relate to propositions have been crucial to the developing views of what
role pragmatics has in describing meaning. In the second half of the last century it
became clear to many scholars that if propositions are bearers of determinate truth
or falsity then the literal meanings of sentences cannot be equivalent to propositions. Typically when uttered, sentences dont provide enough information: they
leave a gap that has to be bridged by the listener. Increasingly the listeners interpretive activity has been characterized as pragmatic.
2.
Context and propositional meaning
Various types of gap have been identified. Perhaps the clearest case is deixis, such
as the italicized deictic expressions (or indexicals): Youre here now. Such expressions of course crucially require contextual information to be accessed, and
thus their linguistic meanings do not seem to be part of the content of the utterances
that contain them. They form part of what Gumperz (1992) calls contextualization
cues, which invoke the relevant contextual assumptions in the participants.
Some writers, following Kaplan (1989a, 1989b) have identified a split in the
meaning of such expressions between a context independent element, termed
character by Kaplan, and the context dependent meaning, which he termed content. The character determines how the context will determine the content: in
other words how the intended reference is identified and thus contributes to determining what proposition is expressed. A further distinction is then made between
two types of indexical: the first are pure indexicals like I and yesterday, for which
the contextual determination is thought to be straightforward or automatic, in the
terms of Perry (2001) and which are independent of speakers intentions. The
others (termed true demonstratives by Kaplan 1989a), like he and that (used deictically rather than anaphorically), require inferential work for the addressees to be
able to identify the referent. This latter set is called discretionary by Perry (2001).
This difference might be significant for setting the pragmatics-semantics boundary
since the automatic reference identification could be assumed to be a semantic process, akin to literal meaning, while the discretionary determination of reference
could be held to be pragmatic, involving inferences about speaker intentions (see
King and Stanley 2005 for discussion).
One problem is that the automatic set seems to shrink on examination. Traditionally English here and now have been considered pure indexicals. However
though we might want to say that here includes the position of the speaker and now
the time of utterance, in fact the determination of the zones these refer to seems to
depend on context-dependent inferences, sensitive to the speakers intent. Moreover, linguistic studies of richer deictic systems than we find in English have revealed referential grids that involve both literal and inferential elements of meaning, as in Yupik (Eskimo-Aleut; Alaska):
463
Yupik demonstratives (Anderson and Keenan 1985: 295)

Extended
Restricted
Obscured
mana
una
this (near SP)
tamana
tauna
that (near ADR)
imna
the aforementioned one
ukna
the one approaching the speaker
augna
ingna
amna
the one going away from the speaker
agna
ikna
akemna
the one across there
qaugna
kiugna
qamna
the one inland, inside, upriver
qagna
keggna
qakemna
the one outside
una
kana
camna
the one below, toward river
unegna
ugna
cakemna
the one downriver, by the exit
paugna
pingna
pamna
the one up there, away from river
pagna
pikna
pakemna
the one up above
This system incorporates a semantic classification of the objects to which the demonstratives refer: extended forms are for either large expanses of land or water,
or objects that are lengthy or moving; restricted applies to objects that are stationary, or moving within a confined area, and fairly small in extent, relatively near,
and visible; and obscured describes objects that are farther away and not clearly
in sight. At the same time the demonstratives crucially rely on the position of the
speaker and so are deictic (see Anderson and Keenan 1985 for details). Hanks
(1992) argues that in many languages deictic expressions, whose core function is to
identify the referent through different relations to elements of the speech act, are
typically fused with other information about the category of referents, the speakers attitude or stance and the intended speech act.
Another type of gap between sentences and propositions is caused by the need
for contextual disambiguation. Ambiguity, which may be structural as in (1), involve quantifier scope as in (2), or be lexical as in (3), results in a sentence fitting
two or more propositions and thus requires the audience to determine which fit is
more appropriate for the context:
1. He tracked the man with binoculars.
2. All reviewers watched one movie.
3. She held the note.
The ambiguity in (1) depends on whether the prepositional phrase with binoculars
is interpreted as modifying the verb track or the noun phrase the man. Both interpretations are possible but describe different situations. In (2) the ambiguity derives from interpretations of the relative scope of the quantifiers all and one: if all
is given wide scope over one, then there may be as many films as reviewers; if one
has scope over all then they all want to see the one film. While there have been attempts to determine such scope ambiguities by semantic or logical rules, other
writers (for example, Saba and Corriveau 2001: 276) suggest that quantifier scope
464
John Saeed
interpretation must involve a pragmatic level. One type of evidence is the influence
of real world knowledge. Thus while the sentence in (4) has the predicted scope
ambiguity (One and the same doctor read all the books / Each book was read by a
possibly different doctor), this typically doesnt arise in (5) because of what we
know about the world:
4. A doctor read every book.
5. A doctor lives in every village.
Lexical ambiguity as in (3), where the noun note can of course be a musical sound
or a written message and consequently different senses of held are invoked, arises
where the same phonological shape is shared by two or more distinct semantic
units (or lexemes). A distinction is often made between homonymy, like pump device for moving liquid or gas and pump shoe where the semantic units are clearly
distinct, and polysemy, where different but related senses are involved, like fault in
Whose fault was the banking crisis? and The San Andreas Fault doesnt go
through San Francisco. The decisions about relatedness are, however, not without
difficulty (Lyons 1977: 55069). The question of how the multiple senses are accessed and selected in context has been an important research issue in the psychology of language (see Simpson 1984, 1994 for discussion). It has also proved to be a
practical problem for natural language processing, for example in machine translation, and the investigation of the relevant contextual information for lexical disambiguation has been an important research strand in computational linguists, as
described in Agirre and Edmonds (2006).
Remaining with quantifiers it has been widely noted (e.g. Recanati 1989,
1993b, Bach 1994, Bring 1997) that in normal use the domain of a quantifier is
typically underdetermined and needs to be contextually enriched. So for the second
sentence in (6a) below, Recanati (2000: 678) suggests the representation in (6b),
where the leftmost square bracket identifies the situation or context to which the utterance relates:
6. a. Berkeley is a nice place. There are bookstores and coffee shops everywhere.
b. [Berkeley] << There are bookstores and coffee shops everywhere>>
Although in this case the domain is mentioned in the previous sentence, in commonly encountered examples like (7) below, the relevant domain is left for the
audience to supply from context.
7. Are you going to the party? Everyone will be there.
If quantifier restriction is freely supplied by contextual information then many
writers have concluded that the semantic content of the utterance underspecifies
the intended proposition.4
In a similar vein, several writers, including Sadock (1984), Horn (1992), and
Carston (1998), have argued about the extent of pragmatic enrichment needed to
465
explain uses of cardinal numbers, which may in different contexts for a cardinal
number x have the intended interpretations, exactly x as in (8) below, at least x in
(9), or at most x in (10):
8. He had six children but one died so now he has five.
9. She needs to reach thirty metres to make the javelin final.
10. Undergraduates may borrow six books from the library.
Another type of underspecification is shown by gradable adjectives (Kennedy
2007), for example, like tall, expensive, and fast, which require contextual fill in
to be interpreted. This is clear in implicit comparison uses like:
11. Isabel is tall.
12. Restaurants in Dublin are expensive.
Participants have to access contextually determined standards to interpret such
examples: Isabel might be tall for a seven year old child, for example. Context update models (Kyburg and Morreau 2000, Barker 2002) provide a presuppositional
account of this process. A related phenomenon is loose talk, argued by Lasersohn
(1999) to be distinct from vagueness. This is where speakers employ a pragmatically licensed degree of approximation in their use of words, even where the terms
are quite sharply defined. Thus a speaker might in casual conversations say (13)
below:
13. Ive a three am. flight tomorrow.
even if the flight is actually at 3.05 am. If the context is right, the inaccuracy is
treated as pragmatically irrelevant.
As pragmatic research has developed over the past decades more examples of
such under-determinacy have been identified. Typical examples include possessive
constructions as in (14); understood time references in (15); and the strengthening
of and in conjoined subjects as in (16):
14. The conversation turned to Johns book.
15. Ive had lunch.
16. John and Mary lifted the piano.
In (14) the relationship described by the possessive pronouns is contextually determined: the book he wrote, the one he owns, the one he was assigned to review, etc.
If (15) is a reply to an invitation, it communicates that the speaker has eaten lunch
within a contextually appropriate period of time (e.g. that day). In (16) there may
be a stereotypical or default interpretation that John and Mary lifted the piano together (Levinson 2000).
The increasing role of contextual inference in the determination of the truthdeterminate propositional meaning expressed by utterances has been called pragmatic intrusion (Levinson 2000). The growing emphasis on the context sensitivity
466
John Saeed
of language raises fundamental challenges to philosophical approaches which

base theories of meaning on theories of truth. The varying extent to which philosophers permit the intrusion of contextual information into semantics can be
viewed as grades of contextualism (Recanati 2004, 2005) as discussed in Preyer
and Peter (eds. 2005, 2007). Some philosophical defences against contextualism
have been proposed, for example by Borg (2004) and Cappelen and Lepore
(2005). These authors defend formal semantic approaches by proposing versions
of semantic minimalism, where context dependence is viewed as limited to cases
triggered by a limited set of linguistic forms. In Cappelen and Lepore (2005) this
set of forms is reminiscent of Kaplans (1989a, 1989b) indexicals, mentioned earlier. Their account pushes contextual variation into what they call speech act
pluralism. In this view, a sentence without one of the context-sensitive elements
always expresses the same proposition, regardless of context. However when uttered in context such a sentence may be expanded to express infinitely many propositions, only some of which are relevant to the context, and it is these expanded
propositions that motivate truth-value judgments. In other words semantic content
and speech act content can be different and, although the former is viewed as a
proposition, it is the latter that we have truth condition intuitions about. The cost
of this, from a linguistics perspective, is a weakening of the role of intuitions
about semantic content, which might risk the empirical underpinning of the study
of meaning.
Other philosophers of course are not troubled by a separation of semantic representations from truth conditions and propositions. The philosopher John Searle
adopts a contextualist position:
We know without hesitation what are the truth conditions of The fly is on the ceiling,
but not of The cat is on the ceiling, and this difference is not a matter of meaning, but a
matter of how our factual background information enables us to apply the meanings of
sentences. In general, one can say that in most cases a sentence only determines a set of
truth conditions relative to a set of assumptions that are not realized in the semantic content of the sentence. (Searle 1979: 86)
Searle argues that background assumptions have to be incorporated before a determinate proposition can be communicated. In particular, in this view, an addressee
cannot respond to the intended speech act, say an order, until this is done.
Noam Chomsky, whose view of a semantic representation concentrates on an
individuals mind-internal mapping between grammar and conceptual structure,
shows even greater scepticism about a role for truth conditional denotation in semantic analysis: the argument for a reference-based semantics (apart from an internalist syntactic version) seems to me weak. It is possible that natural language has
only syntax and pragmatics (1995: 26).
Chomskys view is that the meanings of sentences do not specify truth conditions even when relativized to context. This is reminiscent of Strawsons (1950)
view that although sentences can be used to assert things that are true or false, sen-
467
tences themselves are neither true nor false. Pietroski (2005: 256) characterizes
Chomskys position as follows:
Rather the meaning of S is a compositionally determined intrinsic property of S that
constrains and guides without determining how S can be used to make true or false assertions in various conversational situations making truth-evaluable assertions is one
of the things we can do with sentences, in contexts, though uses of this kind are highly
variable; while people refer to things, words dont; and sentence use may not be a theoretically tractable phenomenon.
Chomsky (1995) effectively proposes a break in the link assumed by many philosophers between the truth-conditional semantics of artificial languages and accounts of meaning for natural languages. In doing this he takes a use-based theory
deep into traditional semantic territory and makes sense of the dramatic suggestion
in the quotation above.
3.
Neo-Gricean inferential pragmatics
Within contemporary linguistics, debate about the relation between pragmatics and
semantics has followed from Paul Grices (1975, 1978, 1989) influential distinction between what is said in an utterance and what a speaker means by an utterance.5 This is often characterised (see for example Carston 2004) as a distinction
between what is said as propositional, or truth-conditional, content, and what is
intentionally implied by the speakers utterance. As noted by several writers (e.g.
Szab 2006: 386), Grice never quite seems to give us a tight definition of this notion of what is said, identifying it as closely related to the conventional meaning
of the words (the sentence) he has uttered (1989: 25). This leaves it somewhat
open as to how much resolution of context-sensitive expressions, such as indexicals, is necessary to reach what is said. The distinction in any case depends on his
notion of conversational implicature as implications that are not part of the conventional meaning of the utterance, are context-dependent and, crucially, intended
by the speaker, as in a typical (invented) example (17):
17. A: Would you like another drink?
B: Id better go home; Ive an early start in the morning.
Here Bs reply may be interpreted to implicate a refusal or simply no.
Grices work suggests a distinction between semantics and pragmatics: semantics is concerned with the proposition expressed while pragmatics is concerned with the implicatures of the utterance. His work opens up the prospect of
an inferential pragmatics that explores the gap between the linguistic content of
the speakers utterance, when viewed analytically, on the one hand, and what listeners seem to understand as speakers meaning. Pragmatic processing relies on
468
John Saeed
the predictability of inferential behaviour, seen by Grice as depending on co-operative principles underlying communication (Grice 1975, 1978). His cooperative
principle and maxims below predict that the listener assumes, and the speaker
relies on this, that the speaker is speaking truthfully, relevantly, and appropriately,
that is performing subtasks of assessing what the hearer knows, selecting the right
degree of clarity, etc. Clearly this is not a claim that speakers always behave in
this way; rather it is a claim that inferential strategies seem to rely on such assumptions.
Grices Co-operative Principle (see Grice 1989:26)
Make your contribution such as is required, at the stage at which it occurs, by
the accepted purpose or direction of the talk exchange in which you are engaged.
Grices maxims (adapted from Grice 1975: 46)
i. The Maxim of Quality
Try to make your contribution one that is true, i.e.
a) do not say what you believe is false
b) do not say that for which you lack adequate evidence.
ii. The Maxim of Quantity
Make your contribution as informative as is required for the current purposes
of the exchange (i.e. not more or less informative).
iii The Maxim of Relevance
Make your contributions relevant.
iv. The Maxim of Manner
Be perspicuous, and specifically:
a) avoid ambiguity
b) avoid obscurity
c) be brief
d) be orderly.
His approach moreover promises an account of pragmatic processing. Grice proposed that conversational implicatures must be capable of being worked out, i.e. a
reasoning process must be identifiable, even if it is not deduction. For every implicature one can predict a process of calculation involving what is said (literal
meaning), the cooperative principle and context. A characteristic Gricean formulation is:
Working out schema for conversational implicatures
(a) The speaker (S) has said that p.
(b) There is no reason to think that S is not observing the maxims.
(c) S could not be doing this unless he thought that q.
(d) S knows (and knows that the hearer (H) knows that he knows) that H can
see that he thinks that the supposition that he thinks q is required.
469
(e) S has done nothing to stop H from thinking that q.

(f) S intends H to think, or is at least willing to allow H to think, that q.
(g) And so, S has implicated that q. (adapted from Grice 1975: 50)
Grice used the term non-natural meaning (meaningnn) for interactional communication between mutually aware participants to distinguish it from simpler forms of
signification. He provided an initial characterization of this as below:
Figure 1: Grices Meaningnn (Levinson 2000: 13)
As suggested above, the first split in Figure 1 above might be taken to coincide
with the Gricean division between semantics and pragmatics, though Grice never
puts it in these terms, since what is said is presumably the truth-conditional content. However the notion of conventional implicature causes problems here since it
is conventionally marked or coded but inferentially determined. A typical example
is English but, which as shown below consistently carries a meaning of unexpectedness or contrast:
18. a. Hes a politician but hes honest.
b. Joan is tall but Libby is short.
c. But you know I dont drink coffee!
Despite this degree of conventionality, the meaning of but is excluded from what
is said because it does not contribute to the truth conditions of the proposition expressed. It also requires contextual inferences to determine the precise interpretation. Subsequent work in this general framework has described such behaviour
without this notion of conventional implicature and, as we shall see below, more
generally has suggested different borders between semantics and pragmatics.
The pragmatic inferences called conversational implicatures are claimed to
have specific features. They are described, for example, as cancellable, non-detachable and non-conventional. They are called cancellable because, for example,
we can add in (19) below another proposition to cancel the implicature without
causing anomaly:
470
John Saeed
19. a. John has three children, in fact ten.

b. John has three children, maybe more.
This contrasts with other types of inference like presupposition in (20), and entailment in (21), where anomaly results:
20. ?The King of France is bald and/but theres no King of France.
21. ?The president was assassinated yesterday but hes not dead.
Non-detachability is the claim that they are attached to a semantic level rather than
to surface linguistic form. Thus by changing the specific linguistic form, say via
synonyms, there is no change in the implicatures. Levinson (1983: 116117)
claims that if we know a context where (22a) implicates (22b) then (23a-d) will
also produce (22b):6
22. (a) Johns a genius.
(b) John is an idiot.
23. (a)
(b)
(c)
(d)
Johns a mental prodigy.

Johns an exceptionally clever human being.
Johns an enormous intellect.
Johns a big brain.
Finally non-conventionality captures the fact that the inference is not encoded in
the linguistic form of the utterance, as is clear from examples like (17) earlier.
To complete the Gricean taxonomy: particularized conversational implicatures
(PCIs) depend completely on specific context of utterance: they are once-off inferences. Generalized conversational implicatures (GCIs) are context dependent but
range over contexts, like a default. They only fail if blocked by inconsistency with
context. Carston (2004), following Levinson (2000), provides the following
example to illustrate the difference:
24. A: Did the childrens summer camp go well?
B: Some of them got stomach flu.
GCI: Not all the children got stomach flu.
PCI: The summer camp didnt go as well as hoped.
This Gricean conception of pragmatics has inspired an extensive literature on the
nature of both the maxims and the implicature themselves (cf. Huang this volume).
Horn (1984, 1989, 2004) for example argues for a reworking of the Gricean
maxims into two basic pragmatic principles, briefly outlined below:
Horns Q- and R-principles (Horn 1984, 1989)
(a) The Q-principle
Make your contribution sufficient;
Say as much as you can (given the R-principle)
471
(b) The R-principle

Make your contribution necessary;
Say no more than you must (given the Q-principle)
Horns Q-principle conflates Grices Quantity maxim and parts 1 and 2 of Grices
Manner maxims and can basically be paraphrased as say the most that you are
licensed to say. Alternatively, taking the hearers point of view in mind, it could be
paraphrased as maximize the informational content of what you say. On the other
hand the R-principle is a speaker-orientated economy principle. The two speaker
versus hearer economy principles are held to be in tension. The best known application of the Q-principle is the Q- or Horn scales of strength, schematized below:
Q-scale/Horn scale (Huang 2007: 38)
For <S, W> to form a Q- or Horn scale,
(i) A(S) entails A(W) for some arbitrary sentence frame A;
(ii) S and W are equally lexicalized, of the same word class, and from the same
register; and
(iii) S and W are about the same semantic relation, or from the same semantic
field.
Thus Horn scales are typically scales of alternates, ordered strong to weak, for
example:
25. (a) <all, some>
(b) <identical, similar>
The Q-maxim operating on scales ensures that the use of a weaker alternate implies
the negative of a stronger:
26. a. Not all the listeners enjoyed the show implicates some (i.e. not none)
did
b. Their results are similar implicates Their results are not identical
A similar but distinct revision is attempted by Levinson (1987a, 1987b, 1991,
2000), who, seeking to distinguish semantic content from surface form in relation
to economy, proposes three principles: the Q-, I-, and M-principles. The Q-principle, which corresponds to Grices Maxim of Quantity 1, is given in simplified
form below:
Levinsons Q-principle (simplified; Levinson 2000, Huang 2007)
Speaker: Do not say less than is required (bearing in mind the I-principle)
Addressee: What is not said is not the case.
This principle, like Horns Q-principle, gives rise to scales. In addition to the
simple scales as in (25) above, Levinson identifies more complex cases of alternate
sets. We can briefly mention two. The first are clausal, where the alternates are dis-
472
John Saeed
tinguished by the stronger entailing its subparts and the weaker not entailing its
components, as in (27):
27. Example of Levinson Q-clausal: conditionals
a. <since p then q>, <if p then q>
b. Since hes here, he can play entails He is here; and He can play
c. If hes here, he can play does not entail He is here, nor He can play
In this example, since p then q entails p, and entails q, as shown in (27b). However, if p then q does not entail p, and does not entail q, as shown in (27c). Because of this relationship, by uttering If hes here, he can play the speaker, by excluding the stronger, implicates: Maybe hes here; maybe not; maybe he can play;
maybe not i.e. uncertainty.
The second type is simple alternates, where the alternates are a non-entailing
contrast set. These may be ordered in strength (<stronger, weaker>) as in (28) or an
unordered taxonomy, like the colour terms in (29):
28. Q-alternate: <succeed, try>
29. <red, orange, yellow, green, blue, silver >
Hence from (29) a speaker uttering His car is blue (weakly) implicates His car
isnt silver, etc.
Levinsons I- and M-Principles concern economy. His I-principle, which corresponds to Grices Maxim of Quantity 2, is in simplified form:
Levinsons I-principle (simplified; Levinson 2000, Huang 2007)
Speaker: Do not say more than is required (bearing in mind the Q-principle)
Addressee: What is generally said is stereotypically and specifically exemplified.
This can be paraphrased as the use of a semantically general expression implicates
a semantically specific interpretation. This principle is used to explain how speakers employ or assume defaults and stereotypes, like those in (30):
30. A cheese board = for cutting cheese not made of cheese
A kitchen knife = used in kitchens not for cutting kitchens
It is also used for the examples of strengthening implications mentioned earlier, for
example the occasion when conjunctions like English and can imply temporal and
causal sequence as in (31); when if is interpreted as if and only if as in (32); and
the strengthening of negation as when (33a) implicates (33b):
31. Its always the same at parties: either you get drunk and no one will talk to
you or no one will talk to you and you get drunk (Blakemore 1992: 80)
32. If you pass your examinations, Ill buy you a car.
33. a. Susana doesnt like the Irish weather.
b. Susana dislikes the Irish weather.
473
Levinsons M-Principle relates to the form of utterances, and can be characterized

as:
Levinsons M-principle (simplified; Levinson 2000, Huang 2007)
Speaker: Do not use a marked expression without reason.
Addressee: What is said in a marked way is not unmarked.
This principle explains speakers use of unusual (marked) expressions to avoid defaults and stereotypes. It seeks to explain the pragmatic effects of negative asymmetry, where (34a) below will be interpreted differently from (34b); or where periphrasis carries distinctions of meaning as in (35):
34. a. I dont dislike Henry.
b. I like Henry.
35. a. Bob opened the door.
b. Bob caused the door to open.
Once again these principles are viewed as operating in tension; the interaction between them is a complicated and controversial issue: see Levinson (2000) and
Huang (2007) for discussion.
The work of neo-Gricean writers like Horn and Levinson has shed light on a
wide range of linguistic expressions, seeking to determine the role of semantic encoding and pragmatic inference in their interpretation. An important claim of Levinson (2000) is that generalized conversational implicatures represent a form of
conventional language use, characterized as default inferences, which can be
over-ruled in context. In this view, these linguistically-triggered default inferences contribute to propositional content. This then changes the Gricean relation
between semantics and pragmatics. Since pragmatic inferences contribute to
propositional content, then clearly if there is purely semantic representation it is
pre- or sub-propositional. Or, to put it another way: both semantic interpretation
and pragmatic inference combine in the formation of propositional truth-bearing
content.
4.
Relevance theory
This pragmatic intrusion, to use Levinsons term, is even more marked in relevance
theory (Sperber and Wilson 1995, Carston 2002), which is a more radical development from Gricean inferential pragmatics. Relevance theory is an approach to
communication based on the view that people are predisposed to pay attention to
phenomena in their environment when doing so is likely to bring about improvements in their belief system. In other words, we tend to pay attention to stimuli
which we expect will turn out to be relevant to us. Some stimuli, which include
474
John Saeed
pointing gestures and utterances, are designed to create an expectation that paying
attention to them will yield significant cognitive and other rewards. In other words,
they are designed to create expectations of their own relevance. This type of stimuli is called in this theory ostensive stimuli and the kind of communication which
involves the production and interpretation of ostensive stimuli is called ostensiveinferential communication.
In this view a phenomenon is relevant to an individual (a) to the extent that the
cognitive effects achieved when it is processed in context are large, and (b) to the
extent that the processing effort required for achieving the effects is small (Sperber
and Wilson 1995: 153).
This definition of relevance provides the basis for two principles of human cognition and communication, known as the cognitive and the communicative principles of relevance:
The Cognitive Principle of Relevance
Human cognition tends to be geared to the maximization of relevance. (Sperber
and Wilson 1995: 260)
The Communicative Principle of Relevance
Every act of ostensive communication communicates a presumption of its own
optimal relevance. (Sperber and Wilson 1995: 260)
The presumption can be described as follows:
Presumption of Optimal Relevance
(a) The ostensive stimulus is relevant enough for it to be worth the addressees
while to process it.
(b) The ostensive stimulus is the most relevant one compatible with the communicators abilities and preferences. (Sperber and Wilson 1995: 270)
The influence of Grice is clear here, with its assumption that the cost, so to speak,
of inference can only be motivated by a shared presumption of certain strategies on
both the speakers and hearers behalves.
Consequently, in this account, there are certain assumptions about economy, as
described below by Carston (2002):
Relevance-theoretic Comprehension Strategy
(a) construct interpretations in order of accessibility (i.e. follow a path of least
effort);
(b) stop when your expectation of relevance is satisfied. (Carston 2002: 380)
Of course expectations of relevance may vary across different situations of utterance and in practice adult hearers are expected to have developed strategies to account for speakers fallibility or deceptiveness; and other type of communicative
475
noise. Giving a schematic view of the sub-tasks in the comprehension process,

Wilson and Sperber summarize these as below:
Sub-tasks in the comprehension process (Wilson and Sperber 2004: 615).
a. Constructing an appropriate hypothesis about explicit content (in relevancetheoretic terms, explicatures) via decoding, disambiguation, reference resolution, and other pragmatic enrichment processes.
b. Constructing an appropriate hypothesis about the intended contextual assumptions (in relevance-theoretic terms, implicated premises).
c. Constructing an appropriate hypothesis about the intended contextual implications (in relevance-theoretic terms, implicated conclusions)
From a relevance theory point of view, the processes of recovery of explicatures
and implicatures are both equally inferential and interact freely. In distinction
to Levinsons (2000) position, relevance theorists argue that the pragmatic processes leading to explicatures (i.e. explicit propositional content) are not restricted to special linguistic constructions, nor are they implicatures. Like all
pragmatic processes though, they are motivated by the same presumption of optimal relevance. As we can see, the tasks in constructing a proposition involve context-dependent inference. In this way relevance theory reflects the influence of
context in the most basic of what has sometimes been seen as coded semantic
meaning.
Blakemore (2002) suggests that ostensive stimuli may contribute to the comprehension process in three main ways. First, they may focus the addressees attention in a particular direction (e.g. pointing gestures and acts of displaying in general). For example, a speaker might ask where the bus stop is and the addressee
might point to its location. In this case, the communicator, by displaying an object
to the addressee, suggests that it is relevant as evidence of his/her informative intention.
Secondly, verbal ostensive stimuli typically encode conceptual representations
which contribute to the truth-conditional content of the utterance directly. For
example, instead of pointing the addressee in this scene might say Its across
the road. In this case, the words used present some evidence of what the speaker
intends to communicate by virtue of the concepts that they encode. Thirdly, lexical items may encode procedures (i.e. constraints on or indications of the way
pragmatic inference should proceed). Blakemore (2002) discusses the English
examples in (3638):
36. Tom can open the safe. He knows the combination.
37. Tom can open the safe. So he knows the combination.
38. Tom can open the safe. After all he knows the combination.
476
John Saeed
Example (36) expresses two propositions. While the explicit content is unambiguous, the implicatures that can be retrieved are not. How the implicit content can be
constrained is exemplified in (37) and (38). In these examples, the discourse markers
so and after all guide the computational route that needs to be taken by the hearer. In
(37) the second statement is a conclusion drawn from the first. If someone can open
the safe then he knows the combination. So encodes a constraint on the way the utterance it introduces should be interpreted in context. In So p, p is relevant as a conclusion (which follows from some salient premises). In (38) the second proposition
strengthens the existing assumption that Tom can open the safe. A pronoun like he
in these examples encodes some procedural information which helps the hearer to
form a conceptual representation of the intended referent. So, there are two types of
procedural constraints: constraints on the derivation of implicatures (e.g. so, after
all, however) and constraints on pragmatic inferences involved in deriving the
propositional content of the utterance (e.g. pronouns, tense markers).
We can give a short example from Carston (2006) of how this approach applies
to a hypothetical conversational exchange in (39):
39. A conversational exchange (Carston 2006: 39)
Bob: How is Jane feeling after her first year at university?
Sue: She didnt get enough units and cant continue.
40. A possible interpretation of Sues reply (Carston 2006: 39)
a.
[JANE 1 DIDN T PASS ENOUGH UNIVERSITY COURSE CREDITS TO QUALIFY
FOR ADMISSION TO SECOND YEAR STUDY ]p & AS A RESULT OF P JANE 1
CANNOT CONTINUE WITH UNIVERSITY STUDY.
JANE IS NOT FEELING HAPPY.
b.
The interpretation in (40) contains two different types of propositional results. The
first, (40a) is an expanded or explicit version of the linguistically encoded material.
So although, as Carston notes, the word she doesnt conventionally contain the
meaning Jane or the word units contain an element of meaning university, it seems
that the linguistic items can be a trigger to identify the fuller version, by a process
of working out or inference. The second interpretation is somewhat different. It
provides an answer to Bobs question, but has to be seen as providing it indirectly
since the proposition cant be seen as directly connected to the linguistic content of
Sues answer. Though both parts of the interpretation in (40) seem to depend on
context they seem to do it in different ways. We can easily imagine a scenario
where (40a) is the same, but where Jane is quite happy not continuing her studies.
Perhaps she was forced to attend by her parents and this gives her the excuse to
leave. Thus the indirect answer would be the opposite. Thus the pragmatic processes in (40a) involve expanding the linguistic content while those in (40b) involve identifying intended implications of the expanded linguistic content. Both
however are seen as pragmatic processes.
477
This relevance theory view of the underspecification of encoded meaning is

clearly shown in their work on lexical meaning. The contextual variability of word
meaning is viewed as access to lexical concepts governed by relevance-based contextual inferences:
Quite generally the occurrence of a word in an utterance provides a piece of evidence, a
pointer to a concept involved in the speakers meaning. It may so happen that the intended concept is the very one encoded by the word, which is therefore used in its
strictly literal sense. However, we would argue that this is no more than a possibility, not
a preferred or default interpretation. Any interpretation, whether literal or not, results
from mutual adjustment of the explicit and implicit content of the utterance. (Sperber
and Wilson 1997:15)
The contextual variability of word meaning is seen as a product of a number of inferential processes, including narrowing, as in Wilson and Carstons (2006:409)
example:
41. All politicians drink.
Here the meaning of drink has been narrowed from the meaning drink liquid to
mean drink alcohol and possibly further, to drink alcohol in sufficient quantities
to be worth commenting on. In relevance theory terms, drinking liquid is clearly
necessary for staying alive and in most contexts the fact that a person drinks liquid
has no obvious relevance. Thus in many contexts the expected interpretation of
(42) below:
42. Gordon doesnt drink.
is that Gordon doesnt drink alcohol. Other processes include broadening or approximation (Wilson and Sperber 2004) where the sense of word is extended, as
the word square in:
43. His father has a square face too.
In this example it is unlikely that the speaker intends the usual dictionary meaning
of square. In this view, lexical narrowing and broadening are types of pragmatic
enrichment.
Relevance theorists treat such processes as general in lexical interpretation.
Wilson and Carston (2007) quote Murphys (1997) experimental work on evidence
for distinct word senses by respondents supplied antonyms that contribute to explicatures. For the English word fresh when used with different nouns these include those in Table 1:
478
John Saeed
Table 1.
Antonyms of fresh (Murphy 1997:2379)
fresh
ANTONYMS
shirt
dirty
vegetables
rotten
fish
frozen
sheets
recently slept-in
water
dirty/salt
bread
stale
air
polluted
outlook
tired
assistant
experienced
idea
old
They argue that this is a typical example where a general concept, here FRESH , gets
tailored to individual contexts.
Wilson and Carston (2007) argue that a whole range of what have been considered individual lexical processes, narrowing, broadening, hyperbole (as in 44
below), metaphor (45), category extension (46), etc. can be given a unified account
as inferential processes constrained by the encyclopaedic entries for concepts and
the expectations of relevance. In particular they suggest that this involves a process
of ad hoc concept construction, a once-off contextual interpretation, conventionally shown as starred concepts, i.e. FRESH * for the last example. One important aspect of this approach is the claim that processes like broadening, hyperbole and
metaphor are outcomes of essentially the same pragmatic processes, by different
balancing of encyclopaedic information, contextual assumptions and derived
contextual implications. Thus, their example (47) below could be intended and interpreted as literal, approximation, hyperbole or metaphor, depending on the contextual features (Wilson and Carston 2007: 2489):
44.
45.
46.
47.
5.
Nobody walks in Los Angeles.

Manchester United clawed their way back into the game.
New Orleans is the new Hollywood.
The water is boiling.
Cognitive semantics
Aspects of this relevance theory approach to lexical pragmatics suggest parallels to

that of cognitive semantics (Lakoff 1987, Langacker 1987, 1991, Talmy 2000 inter
479
alia). In cognitive semantics we also see a rejection of the view that literal encoded
meaning has a privileged position, and of the view that hyperbole, metonymy,
metaphor, etc. are special rhetorical devices or tropes. We also see a search for
common general principles of sense determination and extension in context; and
an emphasis on the role of encyclopaedic knowledge in accessing concepts. However, cognitive semanticists use these features of lexical meaning as an argument
against identifying a clearly-defined semantics/pragmatics distinction at all. Langacker (1987:154), for examples, states: The distinction between semantics and
pragmatics (or between linguistic and extralinguistic knowledge) is largely artefactual, and the only viable conception of linguistic semantics is one that avoids
such false dichotomies and is consequently encyclopedic in nature.
Langacker argues that it is impossible to distinguish between lexical and encyclopedic meaning, and, in common with other cognitive linguists, views the
principles involved in organizing concepts and lexically accessing them as part of
more general cognitive processes, rather than specialized linguistic operations.
However the similarities with the relevance theory position are clear. Cognitive
semanticists argue that the conventional meaning associated with a particular linguistic unit acts as a trigger for a process of meaning construction which involves
the determination of an appropriate interpretation for the context of the utterance.
Fauconnier and Turner (2002), for example, discuss interpretation of the English
word safe. They provide some examples of its range of meanings in the context of a
child playing on a beach:
48. a. The child is safe.
b. The beach is safe.
c. The shovel is safe.
As they note, in this context, the interpretation of (48a) is that the child will not
come to any harm. On the other hand, (48b) is not likely in this context to mean that
the beach will not come to harm, but rather that the beach is an environment in
which the risk of the child coming to harm is small. Finally, (48c) is not likely to
mean that the shovel will not come to harm, but that it will not cause harm to the
child. Their point is that there is no single fixed property that safe assigns to
the words child, beach and shovel. In order to understand what the speaker means,
the addressee employs encyclopaedic knowledge of children, beaches and shovels,
and knowledge relating to what it means to be safe. We can see a parallel with Wilson and Cartons (2007) picture of participants constructing a meaning by contextual adjustment based on encylopaedic knowledge.
Although there are individual differences among writers within this approach,
in general cognitive semanticists place no emphasis on a distinction between semantics and pragmatics, seeking instead to identify a range of cognitively motivated representations and computations that straddle the traditional border. An
important constituent of this approach is of course the rejection of truth conditions,
480
John Saeed
or correspondences to states of affairs, as a way of defining meaning. Lakoff

(1988:121) adopts a notion of experiential cognition, contrasting it with its objectivist counterpart:
Where objectivist cognition views human thought as fundamentally disembodied, experientialist cognition sees human thought as essentially involving the
kind of structured experience that comes from having human bodies, especially
from innate human sensory-motor capacities.
Where objectivist cognition sees meaning in terms of a correspondence theory, as the association of symbols with external objects, experientialist cognition sees meaning as essentially involving an imaginative projection, using
mechanisms of schematization, categorization, metaphor and metonymy to
move away from what we experience in a structured way with our bodies to abstract cognitive models.
Where objectivist cognition sees thought processes as the manipulation of abstract symbols by a great many highly-structured algorithms, experientialist
cognition posits a small number of general cognitive processes whose application to abstract highly-structured cognitive models constitutes reason.
Jackendoff (1990) makes a similar point, though couching his view in more
Chomskyan terms:
Hence a truth-conditional semantics in the Tarskian or Davidsonian sense requires a theory of E-language, of language as an abstract artefact extrinsic to speakers within a
theory of I-semantics, a statement in the Tarskian vein like Sentence S in Language L is
true if and only if condition C is met is taken as shorthand for something like Speakers
of Language L treat Sentence S as true if and only if their construal of the world meets
condition C, and it is subject to similar caveats about attentional and processing limitations. This is the basis of the approach of Conceptual Semantics, in which a level of
mental representation called conceptual structure is seen as the form in which speakers
encode their construal of the world. (Jackendoff 1990: 12)
This stance is reflected in the cognitive semantic approach to metonymy and metaphor. Though they are distinct processes for cognitive semanticists they both rely
on structured fields of real world knowledge, identified as conceptual or cognitive
models. In this view metonymy is a referential strategy that is based on links within
the same cognitive model while metaphor is a mapping across cognitive models.
Metaphor is essentially a way of understanding one conceptual model by relating it
to another, usually more concrete and familiar. There have been a number of proposals for such cognitive models including scripts (Schank and Abelson 1977),
frames (Fillmore 1982, 1985), and idealized cognitive models (Lakoff 1987).
The important fact for this literature is that these cognitive models are bodies of
real world knowledge, often culturally specific. Metaphor is then simply a special
type of the more general cognitive process of analogical mapping. Fauconnier
(1997: 19) uses the example of the coining of the term computer virus for a specific
481
type of harmful program, which he suggests is based on a conceptual model of biological viruses which is generalized or schematized away from the biological details:
Biological virus schema (Fauconnier 1997: 19)
x is present, but unwanted; it comes in, or is put in, from the outside; it does not
naturally belong;
x is able to replicate; new tokens of x appear that have the same undesirable
properties as the original x;
x disrupts the standard function of the system;
the system should be protected against x; this might be achieved if the system
were such that x could not come into it, or if other elements were added to the
system that would counteract the effects of x, or eject x, or destroy x
This schema is transferred to the general aspects of the computer situation; it provides a way of characterizing the new domain. The schema is itself based on lower
level conceptual schemas called image schemas, especially characterizing path and
force dynamics: entry, resistance etc. This analogical mapping between a health
schema and a computer domain licenses a whole system of lexical innovations so
that files can be said to be infected and can be placed in quarantine.
The conclusions cognitive linguists draw from such analyses include the following: (a) that the supposedly linguistic process is actually part of a more general
cognitive process; and (b) that such processes are based on real-world experience
and the resulting conceptual models.
These two points may be used as arguments against what are characterized as
assumptions of formal approaches to language: that language is different from
other cognitive processes and that linguistic knowledge is separable from general
knowledge.
We can take as an example of the resulting blurring of traditional semantic and
pragmatic processes Fauconniers (1994, 1997) work on reference. Fauconnier has
in his theory of mental spaces sought to integrate traditional topics such as referential opacity, modality, presupposition projection, counterfactuals, time deixis,
and indirect reference, together with metaphorical uses of language. Fauconnier
suggests that speakers and hearers construct mental models of situations and it is
through these mental models and operations upon them that reference is possible.
These mental spaces in effect can be seen as a cognitive version of possible world
semantics (Lewis 1973), with the difference that more linguistic behavior is
brought under the analysis. For Fauconnier the creation of discourse consists of the
creation and manipulation of series of mental spaces in which individuals and situations are identified, and between which connections are made, as negotiated by
the participants. He identifies linguistic markers, which he calls spacebuilders, as
invitations to the hearers to construct mental spaces and the relevant referential
links:
482
John Saeed
49. spacebuilders (Fauconnier 1994):

adverbials of location and time like in your new novel, in Peters painting,
when she was a child, after we find the crash site, etc.; modals like possibly
and really; connectives like if then; etc.
As well as these overt markers, spacebuilders, the use of mental spaces is also evident in the normal use of names, which can be shifted. As in the following example
where the same name can be used for a real, historical person; a character in a film;
and the actor playing the character:
50. (a) Truman Capote wrote In Cold Blood.
(b) In the film Capote, Truman Capote develops a close relationship with
Perry Smith, one of the killers.
(c) I didnt like the film: Capote was too handsome. (i.e. the actor Philip
Hoffman)
This shifting of referential expressions like names occurs naturally with deictic
pronouns of course:
51. If I were you, Id sell those shares now.
Such connections can be built up by a number of devices, the creation of fictional
characters, by representation in images, etc. Thus a speaker might say, looking at a
television screen:
52. Ronaldo looks a bit out of focus.
Here, in Fauconniers terminology, the name of the Ronaldo is the trigger and the
target is the image. Real Ronaldo (a) and screen Ronaldo (b) are linked by the
pragmatic function IMAGE , so a description of real Ronaldo (his name, da) can be
used to identify his television image (b). In other words, we assume that a speaker
can use a persons name to refer to an image of that person. This is clearly related to
metonymy: here an electronic pattern created within a machine is given the name
of a person whose image it relays. This relates to the traditional view that metonymy derives from a notion of contiguity or connectedness. The approach extends
naturally of course to more traditional cases:
53. a. The blue suit at the end of the bar is about to throw up.
b. The trains are on strike.
Fauconnier uses a very general statement to describe all of this referential behaviour:
The Identification (ID) Principle (Fauconnier 1994:3)
If two objects (in the most general sense), a and b, are linked by a pragmatic
function F (b = F(a)), a description of a, da, may be used to identify its counterpart b.
483
The important points here are that this is a pragmatic process and that any link is
possible: the versions of connectedness can be very idiosyncratic and culturally
specific, though some of course, like the use with images in paintings, photographs, film and video, quickly become conventionalised across cultures sharing
the technology. We might note once again that one major implication of this is that
it overtly recognises that reference depends on conceptual systems, including cultural knowledge, contextual knowledge etc. The cognitive processes freely access
all kinds of information.
Fauconnier has developed this notion of mental spaces further into conceptual
integration theory (Fauconnier and Turner 2002). In this theory knowledge from
multiple mental spaces can be blended to form a new space, forming an emergent
meaning structure that is distinct from the information provided by each input
space. This model seeks to represent speakers abilities to create blended metaphors and scenarios, as in the following newspaper example:
54. On Wednesday, he faces a fast-paced tango with Gordon Brown during the
pre-Budget report debate.7
Here we see the blend integrating cultural knowledge about British politics and
ballroom dancing.
Finally, an important element in the cognitive linguistics rejection of a clear
distinction between pragmatic and semantics is the notion of construal (Langacker 1987), which identifies the dynamic processes by which the speaker conceptualizes a situation for communicative purposes. This involves, for example, a
speaker conceptualizing entities at various levels of specificity, selecting a perspective on a scene, relating it to background assumptions and expectations and
profiling some entities as more prominent than others. In this view the speaker
imposes a structure on a scene, which is conditioned by and in turn conditions the
linguistic form, by selecting grammatical structures, lexical items, etc. This process is also in part governed by the communicative context, including the speakers communicative goals. This approach therefore dissolves any hard and fast
distinction between representation and communicative acts since construal works
at all levels of the grammar and lexicon and involves access to encylopaedic and
contextual knowledge.
6.
Conclusion
From our discussion it is clear that we can conclude with no neat characterization
of the distinction between pragmatic and semantics. There is a wide range of positions in the literature. We can however suggest some trends. More types of information and processes involved in the interpretation of utterances have come to
be viewed as extra-linguistic, and therefore seen by scholars as pragmatic. For the
484
John Saeed
many scholars who relate sentence meaning to truth conditions or propositions,

this can be seen as an encroachment, or intrusion to use Levinsons term, of pragmatics into semantics. From this perspective, the task is to establish what information about meaning is encoded in the linguistic material uttered, that is, by
the grammatical rules of the language. Or to put it another way: what is the nature
of the underspecified semantic representations? Recanati (2005) has characterized
this shift amongst philosophers of language as a movement from literalism to contextualism, with many intervening positions.
Within linguistics some pragmaticists, such as neo-Griceans and relevance theorists, without jettisoning the importance of truth conditions to meaning, have increasingly seen the notion as orthogonal to the distinction between pragmatics and
semantics. Relevance theorists in particular view the difference in terms of a computational distinction between linguistic decoding and inference. For others, like
Chomsky (1995), who reject the relation to states of the world as integral to the
semantic description of natural languages, the semantic representations are not of
course underspecified. However here the question of how linguistic rules and representations map to conceptual structures is part of semantics while a correspondingly large role is given to pragmatics in the derivation of contextually appropriate
interpretations. Finally, cognitive semanticists, as we have seen, see any clear distinction between pragmatics and semantics as an artificial construct deriving from
a mistaken distinction between expressing information and managing communication.
Notes
1. But cf. Collavin this volume.
2. Here we leave aside those writers, for example Montague (1974), who propose a formal
logical approach to meaning which treats natural languages like artificial logical languages and identifies truth relations between linguistic forms and extensions, with no intervening psychological entities.
3. See for example Stalnakers definition: Pragmatics is the study of linguistic acts and the
contexts in which they are performed (1999: 43).
4. See Stanley and Szab (2000) and Stanley (2002) for discussion of domain restriction.
5. For the following see also Huang this volume.
6. It is not certain though that this is a feature of Manner implicatures, where effects are produced by selecting longer or shorter equivalent descriptions.
7. Vincent Cable: The dance-crazy politician aiming to match Gordon Brown step for step,
Marie Woolf, Independent Newspaper, Monday, 8 December 2003.
485
References
Agirre, Enoko and Philip Edmonds (eds.)
2006
Word Sense Disambiguation: Algorithms and Applications. Dordrecht: Springer.
Almog, Joseph, John Perry, and Howard Wettstein (eds.)
1989
Themes from Kaplan. Oxford: Oxford University Press.
Anderson, Stephen R. and Edward L. Keenan
1985
Deixis. In: Timothy Shopen (ed.), Language Typology and Syntactic Description. Vol. III: Grammatical Categories and the Lexicon, 259308. Cambridge:
Atlas, Jay David
1979
How linguistics matters to philosophy: presupposition, truth, and meaning. In:
David Dinneen and Choon-Kyo Oh (eds.) Syntax and Semantics 11: Presupposition, 265281. New York: Academic Press.
Atlas, Jay David
2005
Logic, Meaning and Conversation: Semantical Underdeterminacy, Implicature, and their Interface. Oxford: Oxford University Press.
Austin, John Langshaw.
1975
How to Do Things with Words. 2nd edition. Oxford: Clarendon Press. (First
published 1962).
Bach, Kent
1994
Conversational impliciture. Mind and Language 9: 12462.
Barker, Chris
2002
The dynamics of vagueness. Linguistics and Philosophy 25.1: 136.
Blakemore, Diane
1992
Understanding Utterances: an introduction to pragmatics. Oxford: Blackwell.
Blakemore, Diane
2002
Linguistic Meaning and Relevance: The Semantics and Pragmatics of Discourse Markers. Cambridge: Cambridge University Press.
Borg, Emma
2004
Bring, Daniel
1997
The Meaning of Topic and Focus. London: Routledge.
2005
Carnap, Rudolf
1942
Introduction to Semantics. Cambridge, MA: Harvard University Press.
Carston, Robyn
1998
Informativeness, relevance and scalar implicature. In: Robyn Carston and
Sachico Uchida (eds.), Relevance Theory: Applications and Implications,
Carston, Robyn
1999
The semantics/pragmatics distinction. In: Ken Turner (ed.), The semantics/
pragmatics interface from different points of view, 85125. Oxford: Elsevier.
Carston, Robyn
2002
486
John Saeed
Carston, Robyn
2004
Truth-conditional content and conversational implicature. In: Claudia Bianchi
(ed.), The Semantics/Pragmatics Distinction, 65100. Stanford, CA: CSLI.
Carston, Robyn
2006
Linguistic communication and the semantics/pragmatics distinction. UCL
Working Papers in Linguistics 18: 3769.
Chomsky, Noam
1988
Language and Problems of Knowledge. The Managua Lectures. Cambridge,
MA: MIT Press.
Chomsky, Noam
1995
Language and nature. Mind 104.413: 161.
Davidson, Donald
1984
Inquiries into Truth and Interpretation. Oxford: Clarendon.
Davidson, Donald
1990
The structure and content of truth. Journal of Philosophy 87: 279328.
1992
Rethinking Context. Cambridge: Cambridge University Press.
Fauconnier, Gilles
1988
Quantification, roles and domains. In: Umberto Eco, Marco Santambrogio and
Patrizia Violi (eds.) Meaning and Mental Representations, 6180. Bloomington, IN: Indiana University Press.
Fauconnier, Gilles
1994
Mental Spaces: aspects of meaning construction in natural language. 2nd edition. Cambridge: Cambridge University Press.
Fauconnier, Gilles
1997
Mappings in Thought and Language. Cambridge: Cambridge University Press.
Fauconnier, Gilles and Mark Turner
2002
The Way We Think: Conceptual Blending and the Minds Hidden Complexities.
New York: Basic Books.
1982
Frame semantics. In: The Linguistic Society of Korea (ed.), Linguistics in the
Morning Calm, 111138. Seoul: Hanshin.
1985
Frames and the semantics of understanding. Quaderni di Semantica 6:
222254.
Frege, Gottlob
1980
Translations from the Philosophical Writings of Gottlob Frege. Edited by Peter
Geach and Max Black. Oxford: Blackwell.
Grice, Herbert Paul
1975
Logic and conversation. In: Peter Cole and Jerrold L. Morgan (eds.), Syntax
and Semantics. Vol. 3: Speech Acts, 4358. New York: Academic Press.
Grice, Herbert Paul
1978
Further notes on logic and conversation. In: Peter Cole (eds.) Syntax and
Semantics. Vol. 9: Pragmatics, 113128. New York: Academic Press.
Grice, Herbert Paul
1989
Studies in the Way of Words. Cambridge, MA: MIT Press.
487
Gumperz, John
1992
Contextualization and understanding. In: Duranti and Goodwin (eds.), 229
252.
Hanks, William F.
1992
The indexical ground of deictic reference. In: Duranti and Goodwin (eds.),
4376.
Horn, Laurence R.
1984
Toward a new taxonomy for pragmatic inference: Q-based and R-based implicatures. In: Deborah Schiffrin (ed.), Meaning, Form and Use in Context: Linguistics Applications, 1142. Washington, DC: Georgetown University Press.
Horn, Laurence R.
1989
A natural history of negation. Chicago: University of Chicago Press.
Horn, Laurence R.
1992
The said and the unsaid. In: Chris Barker and David Dowty (eds.), SALT II:
Proceedings of the Second Conference on Semantics and Linguistic Theory.
163192. (Ohio State Working Papers in Linguistics 40.) Columbus, OH: Ohio
State University Linguistics Department.
Horn, Laurence R.
2004
Implicature. In: Laurence R. Horn and Gregory Ward (eds.), The Handbook of
Huang, Yan
2007
Jackendoff, Ray
1990
Semantic Structures. Cambridge, MA: MIT Press.
Kaplan, David.
1989a
Demonstratives. In: Joseph Almog, John Perry, and Howard Wettstein (eds.),
Themes from Kaplan, 481563. Oxford: Oxford University Press.
Kaplan, David
1989b
Afterthoughts. In: Joseph Almog, John Perry, and Howard Wettstein (eds.),
Kennedy, Chris
2007
Vagueness and grammar: the semantics of relative and absolute gradable adjectives. Linguistics and Philosophy 30.1: 145.
King, Jeffrey C. and Jason Stanley
2005
Semantics, pragmatics and the role of semantic content. In: Szab (ed.),
111164.
Kyburg, Alice and Michael Morreau
2000
Fitting words: vague language in context. Linguistics and Philosophy 23:
577597.
Lakoff, George
1987
Women, Fire and Dangerous Things. Chicago: University of Chicago Press.
Lakoff, George
1988
Cognitive semantics. In: Umberto Eco, Marco Santambrogio and Patrizia Violi
(eds.), Meaning and Mental Representations, 119154. Bloomington, IN: Indiana University Press.
1987
Foundations of Cognitive Grammar. Vol. 1. Stanford, CA: Stanford University
Press.
488
John Saeed
1991
Foundations of Cognitive Grammar. Vol. 2: Descriptive Applications. Stanford, CA: Stanford University Press.
1999
Grammar and Conceptualization. Berlin: de Gruyter.
Lasersohn, Peter N.
1999
Pragmatic halos. Language 75.3: 522551.
Lepore, Ernest and Barry C. Smith (eds.)
2006
The Oxford Handbook of Philosophy of Language. Oxford: Oxford University
Press.
1983
1987a
Minimization and conversational inference. In: Jef Verschueren and Marcella
Bertuccellii Papi (eds.), The Pragmatic Perspective, 61129. Amsterdam:
Benjamins.
1987b
Pragmatics and the grammar of anaphora. Journal of Linguistics 23: 379434.
1991
Pragmatic reduction of the binding conditions revisited. Journal of Linguistics
27: 10761.
2000
Presumptive Meanings: the Theory of Generalized Conversational Implicature. Cambridge, MA: MIT Press.
Lewis, David K.
1973
Counterfactuals. Oxford: Blackwell.
Lyons, John
1977
Semantics. 2 volumes. Cambridge: Cambridge University Press.
Montague, Richard
1974
Formal Philosophy: selected papers of Richard Montague. Edited and with an
introduction by Richmond H. Thomason. New Haven: Yale University Press.
Murphy, Gregory
1997
Polysemy and the creation of novel word meanings. In: Thomas B. Ward,
Steven M. Smith and Jyotsna Vaid (eds.), Creative Thought: An Investigation
of Conceptual Structures and Processes, 235265. Washington, DC: American
Psychological Society.
Ortony, Andrew
1979
Metaphor and Thought. Cambridge: Cambridge University Press.
Peleg, Orna, Rachel Giora and Ofer Fein
2004
Contextual strength: the whens and hows of context effects. In: Ira A. Noveck
and Dan Sperber, (eds.), Experimental pragmatics, 172186. Basingstoke:
Palgrave.
Perry, John
2001
Reference and Reflexivity. Stanford, CA: CSLI Publications.
Pietroski, Paul M.
2005
Meaning before truth. In: Preyer and Peter (eds.), 255302.
489
Preyer, Gerhard and Georg Peter (eds.)

2005
Contextualism in Philosophy: Knowledge, Meaning, and Truth. Oxford: Clarendon Press.
2007
Context-Sensitivity and Semantic Minimalism. Oxford: Oxford University
Press.
Recanati, Franois
1989
The pragmatics of what is said. Mind and Language 4: 295329.
Recanati, Franois
1993a
Recanati, Franois
1993b
Domains of discourse. Linguistics and Philosophy 19: 445475.
Recanati, Franois
2000
Oratio Obliqua, Oratio Recta: an Essay on Metarepresentation. Cambridge,
MA: MIT Press.
Recanati, Franois
2004
Literal meaning. Cambridge: Cambridge University Press.
Recanati, Franois
2005
Literalism and contextualism: some varieties. In: Preyer and Peter (eds.),
171196.
Saba, Walid S. and Jean-Pierre Corriveau
2001
Plausible reasoning and the resolution of quantifier scope ambiguities. Studia
Logica 67.2: 27189.
Sadock, Jerrold M.
1984
Whither radical pragmatics? In: Deborah Schiffrin (ed.), Meaning, Form, and
Use in Context: Linguistic Applications. Georgetown University Roundtable,
139149. Washington, DC: Georgetown University Press.
Saeed, John Ibrahim
2009
Semantics. 3rd ed. Oxford and Malden, MA: Wiley-Blackwell.
Schank, Roger C. and Robert P. Abelson
1977
Scripts, Plans, Goals and Understanding. Hillsdale, N.J.: Lawrence Erlbaum.
Searle, John
1979
Metaphor. In: Ortony (ed.), 83111.
Simpson, Greg B.
1984
Lexical ambiguity and its role in models of word recognition. Psychological
Bulletin 96: 316340
Simpson, Greg B.
1994
Context and the processing of ambiguous words. In: Morton Ann Gernsbacher
(ed.), Handbook of psycholinguistics, 359374. San Diego: Academic Press.
1995
Relevance. 2nd ed. Oxford: Blackwell.
1997
The mapping between the mental and the public lexicon. UCL Working Papers
in Linguistics 9: 120. (Republished in Peter Carruthers and Jill Boucher
(eds.), 1998. Language and Thought, 184200. Cambridge: Cambridge University Press).
490
John Saeed
Stalnaker, Robert
1999
Stanley, Jason
2002
Nominal restriction In: Gerhard Preyer and Georg Peter (eds.), Logical Form
and Language, 365388. Oxford: Oxford University Press.
Stanley, Jason and Zoltn Gendler Szab
2000
On quantifier domain restriction. Mind and Language 15.2/3: 219261.
Strawson, Peter Fredrick.
1950
On referring. Mind 59. 235: 320344.
Szab, Zoltn Gendler
2001
Adjectives in context. In: Istvan Kenesi and Robert M. Harnish (eds.), Perspectives on Semantics, Pragmatics and Discourse, 199146. Amsterdam:
Benjamins.
Szab, Zoltn Gendler (ed.)
2005
Semantics versus Pragmatics. Oxford: Clarendon Press.
Szab, Zoltn Gendler
2006
The Distinction between semantics and pragmatics. In: Lepore and Smith
(eds.), 361389.
Talmy, Leonard
2000
Toward a Cognitive Semantics. 2 vols. Cambridge, MA: MIT Press.
Wilson, Deirdre and Robyn Carston
2007
A unitary approach to lexical pragmatics: relevance, inference and ad hoc concepts. In: Noel Burton-Roberts (ed.), Pragmatics, 23059. Basingstoke: Palgrave Macmillan.
Wilson, Deirdre and Dan Sperber
2004
Relevance theory. In: Lawrence Horn and Gregory Ward (eds.), The Handbook
of Pragmatics, 607632. Oxford: Blackwell.
17.

Elizabeth Couper-Kuhlen
This chapter is an attempt to take stock of advances in the field of prosody, in particular those which have had a direct impact on pragmatics and the study of language use in social contexts. It begins by looking backward, to where the field of
prosody began and at what has been achieved so far with respect to the pragmatic
functions of prosody, and ends by looking forward from the present standpoint to
the challenges that lie ahead. By prosody I refer primarily to those auditory but
non-segmental dimensions of speech which have a variable relation to the words
used (Crystal 1969, Couper-Kuhlen 1986).1
1.
Looking backward
Where and how did the field of linguistic prosody start? What have been the mileposts along the way to its present position? The history of prosody as a linguistic
enterprise dates from approximately the middle of the twentieth century, with
scholars such as J.R. Firth, David Abercrombie, David Crystal, Michael Halliday
in the UK and Kenneth Pike, George Trager, Dwight Bolinger in the US playing a
leading role in establishing its credentials as a discipline within the language
sciences.
Early students of linguistic prosody, particularly those working in the University College of London tradition, focused primarily on intonation and conceptualized its basic unit, the tone unit, as corresponding roughly to the sentence. Descriptive models of the tone unit were developed primarily through introspection and
the examination of read-aloud laboratory speech. The tone unit, it was thought,
could be mapped onto sentences according to syntactic category (cf. e.g. Armstrong and Ward 1926, OConnor and Arnold 1961, Halliday 1967). Five features
characterized this early view of intonation:
(i) metaphor of superposition
In the American tradition, pitch and stress were called suprasegmentals (Trager
and Smith 1957). These were conceptualized as being superimposed on the segmentally based units of language (syllables, morphemes, words) when the latter
were spoken or read aloud. In the British tradition, the stressed syllables were
treated as a frame on which the melodic line was suspended.
492
(ii) the structuralistic perspective

Pitch and stress, working in tandem, were thought to create the basic unit of spoken
language, the tone unit, which itself was divided into smaller structural parts: e.g.,
pre-head, head (or body), nucleus and tail (cf. e.g. Crystal 1969, OConnor and Arnold 1961). Speech was viewed as exhaustively segmentable into such discrete
structural units.
(iii) the context-free approach
Tunes were mapped onto sentences with different syntactic shapes in a contextfree manner. These sentences were consequently isolates, and when pitch and
stress were superimposed on them, the resulting melodies were also isolates. Intonation was acknowledged to be relative, but little thought was given to the question of relative to what? There was only one voice producing little snippets of
melody with no relation to prior tunes or to other voices. In short, there was no
provision for harmony.
(iv) syntactically based functions
The type of melody used was thought to be determined first and foremost by syntactic sentence type (Armstrong and Ward 1926). For instance, declaratives and
wh-interrogatives were said to take final falls in the neutral or default case, yes-no
interrogatives to take final rises. Departures from these neutral contours were said
to convey attitudinal meaning, declaratives with final rises to express doubt, yes-no
interrogatives with final falls to express seriousness or urgency (OConnor and Arnold 1961).
(v) phonological approach to meaning
Intonation was believed to create contrasts in closed systems of meaning, as could
be shown through minimal pairs (Halliday 1967, Schubiger 1958). For instance,
sentences such as She speaks English naturally, George has plans to write or Isnt
he sure of himself were said to be disambiguated by placement of unit boundaries,
position of tonic stress and type of main accent, respectively.
In the early days then, prosody was equivalent to intonation, and intonation (=
pitch and stress) was thought to be superimposed on the occasion of speaking on
particular syntactic sentence types, which themselves were conceptualized as
single utterances, considered in isolation from other utterances. Today we have
come to think quite differently about prosody.2 In part thanks to developments in
the field of sociology, including Berger and Luckmanns social construction of
reality (1967), Sacks, Schegloff and Jeffersons ethnomethodological conversation
analysis (1974), John Gumperz contextualization theory (1982), and the work of
many others whose names go unmentioned here, we have now come to think of
language as a tool for communication and interaction, and consequently of prosody
in terms of social action.
493
The social actions which prosody is implicated in are as varied as are the tasks
which participants must address in communicating with one another. For the conduct of mundane conversation, that primordial site of communication, the following are among those involved:
Turn construction
Action formation
Floor management
Multi-unit turn construction
Turn continuation
Sequence organization
Repair
Stance and alignment
In the following I attempt to summarize what is known at present about the contribution of prosody as a resource for each of these conversational tasks.
1.1.
Prosody as a resource for turn construction
Intonation is no longer conceptualized as something which is optionally overlaid

or superimposed on a piece of decontextualized language. Instead intonation and
prosody more generally is considered to be an integral part of language (Linell
2005: 63). Language products, we have now come to see, do not exist without prosody, i.e. without some kind of pitch, loudness and timing, even if these are realized
silently in reading and writing. Since the natural habitat of language is interaction and it is in turns-at-talk that bursts of language are found (Schegloff 1996),
we now think of intonation and prosody primordially as affording resources for the
construction of turns-at-talk. Given that there are no such things as turns-at-talk
without prosodic formatting, prosody is seen as one of the building blocks of turns.
Prosodic resources are used alongside lexical and syntactic resources as practices
for the construction of turns-at-talk.
1.2.
Prosody as a resource for action formation
Today it is no longer customary to think in terms of simple and direct equations between sentence type (e.g. interrogative vs. declarative) and final pitch movement
(rise vs. fall). Instead, grammar and prosody are thought of as being deployed together to build turns which implement particular social actions in interaction. Prosody then always has some kind of supporting role to play in action formation. A
number of studies have highlighted this function.3 For instance, in dialect survey
interviews specific prosodic features on word repetitions have been shown to cue
the difference between an understanding check of a vocabulary item and a mulling over to determine its familiarity (Kelly and Local 1989). In riddle guessing
494
sequences on call-in radio, another set of prosodic features accompanying the repetition of a guess by the compre cues the difference between an understanding
check of that guess and a negative, disparaging assessment of it (Couper-Kuhlen
1996). In child-adult picture labeling, the prosody with which care-takers repeat
the childs version of a label conveys the difference between an invitation to repair
that label and a confirmation that it is correct (Tarplee 1996). Interestingly, studies
like these testify to the fact that not only final pitch movement but also less wellknown prosodic dimensions such as phonation type (e.g. breathy vs. non-breathy
voice), pitch register (absolute vs. relative) and timing (e.g. delayed vs. welltimed) operate in tandem in specific sequential environments for action formation.
One demonstration of the power of prosodic and phonetic formatting for implementing particular actions will be found in recent work on assessments [or
evaluation] (actions which evaluate some personally experienced object or situation). Assessing turns which are produced in second position, i.e. in response to a
prior assessment, and with prosodic and phonetic downgrading, i.e. with narrower
pitch span, lower pitch accents and reduced volume vis--vis the prior assessment,
will be interpretable as projecting disagreement, rather than agreement, with the
first assessment, even though their assessing term may be the same or lexically upgraded (Ogden 2006).4 As Ogden points out, this kind of phenomenon is revealing
with respect to what prosodic skills a competent speaker of English is expected to
have: viz., they must be able to match their own prosodic and phonetic production
of a turn to that of another speaker and possibly manipulate it for the creation of interactional meaning.
Importantly, the relation between prosody and action formation is not contextfree but dependent on (a) co-occurring lexico-syntactic forms and (b) location in a
particular sequential position. For example, (a) topic proffers5 in English conversation will have final falling pitch if they are done with a wh-interrogative but final
rising pitch if they are yes-no interrogatives (Couper-Kuhlen, forthcoming, b). (b)
Turns which are repeated following other-initiation of repair are likely to be done
with prosodic and phonetic upgrading (i.e. expanded pitch range, longer word durations, greater loudness and articulatory resettings) if they were originally well
placed in the sequence but without this kind of upgrading if they were sequentially
misplaced (Curl 2004). Thus, just as there are no simple equations between sentence type and final pitch, so there are no simple relations between prosodic shape
and action type without a concomitant consideration of lexico-syntax and sequential position.
1.3.
Prosody as a resource for floor management
Acknowledging that language, and thus prosody as well, inhabits turns at talk, we
have also come to recognize its role in signaling transition readiness. Here the
studies which have been carried out so far indicate that prosody, in conjunction
495
with lexico-syntax, has an important role to play in turn-taking. Our judgment of

whether someone has finished speaking or not depends to no little extent on
whether a stretch of speech sounds prosodically complete or not. But the contribution of prosody to turn transition is sensitive to considerations of syntactic and
pragmatic completeness as well (Ford, Fox and Thompson 1996, Ford and Thompson 1996, Selting 1996a, 2000, 2005). That is, a possibly complete prosodic contour will not signal transition readiness unless it is accompanied by possibly complete syntax and unless the activity underway has also reached a point of pragmatic
completion.
The role of prosody in turn-taking is now known to involve more than simply
rising or falling intonation (Szczepek Reed 2004).6 We have learned, for instance,
that different kinds of pausing, e.g. whether a speaker trails off into silence or
produces silence by holding a glottal stop over time, can distinguish turn-yielding
from turn-holding (Local and Kelly 1986). Although we by no means fully understand how turn transition works, we do now know that English dialects can differ
significantly in terms of how syllable-based pitch, loudness and duration are configured at the ends of floor-yielding turns (Local, Kelly and Wells 1986, Wells and
Pepp 1996, Wells and Macfarlane 1998). And we know that speakers of some languages rely much more heavily on prosodic features other than pitch in managing
the floor in interaction. In Finnish, for instance, it appears to be non-modal voice
quality (including creaky phonation) which does much of the work (Ogden 2004,
Ogden and Routarinne 2005). Other languages, by contrast, appear to rely only minimally on prosody for managing the floor (Tanaka 2000, 2004).
Prosody is also known to have a crucial role to play in cueing how overlapping
talk is understood. Competitive incomings7, for instance, i.e. incomings which lay
an immediate claim to speakership well before a current speakers point of possible
completion, will be marked as such with high pitch and increased loudness (French
and Local 1983). The current floor-holder will typically respond either with a prosodic return of competition (increased loudness, decreased pace) or with a yield
to competition (decreased volume). By contrast, entries at non-completion points
which do not lay claim to the floor, e.g. quips and asides, are lower in pitch and volume than prior talk and the speakers norm. Here the current floor-holder does not
treat the incoming speech as turn-competitive but drops out immediately, only to
resume (with a small delay) once it is finished (French and Local 1983).
In sum, current research indicates that, in one way or another and to varying degrees, prosody is implicated in all aspects of turn-taking, including whether and
how speakership transfers from one participant to another in interaction. In this
sense it is central to the concerns of both interactants and analysts.
496
1.4.
Prosody as a resource for multi-unit turn construction
Today it is also customary to look beyond the single intonation unit or phrase to
consider how adjacent intonation phrases relate to one another. This means focussing on the prosodic seams between adjacent units in single-speaker turns-attalk. To do this, the beginnings of new intonation phrases must be related to the
endings of preceding ones, patterns of rhythmic timing across units identified and
attention paid to long-range articulatory and phonatory settings at unit boundaries.8
Registering this level of detail allows us to say that some neighboring intonation
phrases cohere more closely than others. It is the presence of prosodic cohesion between adjacent intonation phrases in a speakers turn which is related to various
practices for multi-unit turn construction, or building one turn out of several turn
units. For instance, at points of possible turn completion there is the practice of
abrupt joins: here what we find at the seam between two adjacent units is a local
speeding-up, a sudden diminuendo and closer than expected temporal proximity
between the two units. Interestingly, these features display that the actions involved are being treated as disjunct with one another (Local and Walker 2004).
Similarly, there is the practice of parenthetic inserts: here it is shifts to low density
of accentuation, faster tempo and lower overall pitch and amplitude compared to
prior talk which signal that the current bit of talk is being presented as parenthetic
(Uhmann 1992, Local 1992, Wichmann 2001, Mazeland 2007). Further, there is
the practice of reported speech, where speakers will be heard to use clear prosodic
breaks in pitch level, rhythm and loudness when switching from their own voice to
that of another (Couper-Kuhlen 1999, Anward 2002, Holt and Clift 2007). Such
prosodic shifts function like quotation marks in speech, although their placement
with respect to the quotation itself is more flexible (Klewitz and Couper-Kuhlen
1999).
Yet not only dis-junctive, also con-junctive relations between units within a
speakers turn are conveyed by clusters of prosodic features linking neighboring
intonation phrases. There is, for instance, the practice of rush-throughs. This involves speeding up the pace of talk towards the end of one turn-constructional unit,
withholding outbreaths and other marks of finality, and moving quickly into the
next turn-constructional unit, so as to present the two units in question as welded
together quasi-seamlessly (Schegloff 1987). Or the practice of projecting, at the
end of one unit, that another unit will follow. For instance, in conjunction with the
particle no when it is used to begin a turn following a topic proffer, mid-to-high
pitch range and increasing intensity project that the speaker will go on to produce
another unit of related talk (Ford, Fox and Hellermann 2004). Prosodic conjunction can also be created through the practice of emphatic scansions, where the
isochronous and isometric9 distribution of syllables across units in a turn-at-talk
produces a scanned multi-unit turn with special rhetorical effects (Auer, CouperKuhlen and Mller1999). Further, there is the practice of conversational listing,
497
where in conjunction with lexico-syntactic parallelism, the use of recurrent plateau contours and other forms of prosodic stylization creates an enumerating
multi-unit turn (Selting 2004, 2007). All of these practices involve particular prosodic features which, together with specific lexico-syntactic forms, link adjacent
units and signal that disjunctive or conjunctive relationships hold between them.
It is thus by comparing and contrasting prosodic formatting at the boundaries
between adjacent units within a single speakers turn that we discover distinct
combinations of pitch configuration, loudness and timing and specific patterns of
phonatory and articulatory setting which together, and along with lexico-syntax,
contribute to the construction of specific kinds of complex, multi-unit turns-at-talk.
1.5.
Prosody as a resource for turn continuation
Today we now also compare and contrast prosodic formatting across non-adjacent
same-speaker turn units and turns. Obviously, there are challenges here due to the
temporal discontinuities involved, but techniques have been developed to deal
with them. By comparing the prosody at the provisional end of a speakers turn
with that at the beginning of a subsequent continuation of that turn by the same
speaker, increments10 can be described as prosodically fitted to their hosts if they
are not significantly higher in pitch or louder in volume than the turn they are continuing and do not involve noticeable changes in articulatory setting or rate of articulation (Walker 2004). More generally, any type of same-speaker continuation
following a possible turn completion can be said to be either prosodically integrated if it is tacked on to the prior intonation contour, or prosodically exposed if it
has a significant pitch jump together with a marked change of tempo, loudness or
rhythm (Auer 1996). Further, when a speakers turn is interrupted or discontinued
and returned to again later, there is now a prosodically and phonetically grounded
way to talk about continuing that turn vs. re-starting it: resumption of a curtailed
turn is done by matching the pitch and loudness levels at drop-out, whereas recycled material is presented as a new contribution if its pitch and loudness levels
are re-set (Local 1992).
In addition to prosodic formatting across same-speaker turns, and just as importantly, prosodic formatting across turns by different speakers can now be compared and contrasted. This is no small feat, given the fact that the speakers involved
may have significantly different pitch spans and loudness ranges as well as different habitual speech rates and rhythms. Yet normalization procedures have been developed for comparing pitch and loudness, accent timing, tempo and speech rate in
cross-speaker talk (cf. e.g. Couper-Kuhlen 1996, Local 2005, Ogden 2006). As a
result, meaningful statements can now be made about the prosody of a speakers
unfinished (or finished) turn and the prosody of a collaborative completion (or continuation)11 of that turn by another speaker (Local 2005, Szczepek Reed 2006).
Collaborative completions, for instance, have been found to be typically no faster
498
or louder than the turn they are completing and to typically match the overall pitch
of the prior talk. They are thus displayed, and accountably so, not only as lexicosyntactic but also as prosodic continuations of what a prior speaker has begun but
not (yet) brought to completion (Local 2005).
In sum, when strict adjacency between stretches of talk is lacking, it is often
prosody which contributes to marking cohesive relations between them. Current
research is revealing just how sensitive participants are to the prosodic details of
continuing a turn-at-talk, be it ones own or that of another. Along with lexico-syntactic considerations, it is these details which help convey whether the talk is being
presented as part of a turn already begun or as something new.
1.6.
Prosody as a resource for sequence organization
We are now also beginning to appreciate the role of prosodic and phonetic resources in signaling how a new turn-at-talk relates to prior turns. For instance, in
radio phone-in calls following an exchange of greetings and possibly how-areyous, callers can signal that a turn in first-topic slot is the reason-for-the-call by
using extra high pitch on its first accented syllable (Couper-Kuhlen 2001).12 If it is
merely an aside, it will lack these features. Once a sequence has reached a point of
possible completion, a cluster of prosodic cues will conspire to convey whether the
next turn is beginning a new sequence or continuing the ongoing one: turns which
are new beginnings are typically louder and higher in pitch from the outset than
prior same-speaker talk and are often accompanied by rhythmic breaks (Goldberg
1978, Couper-Kuhlen 1993, 2004). Turns which continue the prior line of topical
or sequential development lack these features. New sequences which are begun
after a prior one has been closed down are also often marked as phonetically discontinuous by the use of clicks and loud inbreaths (Wright 2007).
Prosodic and phonetic features also contribute to signaling the possible end of a
sequence. Unit-final conjunctionals such as but, and and so have been considered ambivalent in terms of whether the speaker is projecting further talk or not
(Jefferson 1983), but a consideration of their prosodic and phonetic delivery resolves much of the ambivalence (Local and Kelly 1986). Stand-alone variants of
so, for instance, will be heard as potentially sequence-closing if they are relatively
low in pitch, have quiet volume and lack final glottal closure, but as projecting
more on-topic talk if they are relatively high in pitch, loud and have final glottal
closure (Local and Walker 2005).
As Schegloff (1990) has pointed out, adjacent turns-at-talk can either clump
together, i.e. form coherent courses of action, or not. Recent research shows that
such clumping is often underlined by specific prosodic and phonetic devices.
1.7.
499
Prosody as a resource for repair
In addition to signaling structural dimensions of talk, prosody, along with lexicosyntactic resources, is now known to be crucial with respect to what Goffman has
called the ritual requirements of talk. In self-repair, for instance, phonetic and glottal cut-offs will mark a potential trouble source in a speakers turn and initiate its
repair (Jasperson 2002); rhythmic breaks and extreme spikes in pitch and loudness
will accompany the upcoming self-repair itself (Cutler and Ladd 1983). As for
other-initiation of repair, interactants use prosodic resources when they produce a
noticeable silence instead of the conditionally relevant response and thereby implicitly convey that the prior turn may need some reworking. Noticeable silences
are of course notoriously problematic for analysts: how much elapsed time between turns counts as a delay? Yet for participants they are less so. Recent studies
suggest that participants orient to inter-turn silence in relation to the pacing of prior
talk (Couper-Kuhlen 1993, Auer, Couper-Kuhlen and Mller 1999). Prosodic and
phonetic resources are also deployed in other-initiated repair when a turn has been
targeted as problematic via a generalized repair initiator such as what? or huh? and
is repeated with prosodic and phonetic upgrading or downgrading (Curl 2004,
2005).
In short, from what we know so far, it is fair to say that if we did not pay attention to prosodic and phonetic practices in talk we would be unable to account for
much of repair-related behavior, including when repair is initiated, what is targeted
as a trouble source, whether a piece of talk is carrying out a repair or not and how
the responsibility for remedial action is being assigned.
1.8.
Prosody as a resource for displaying and negotiating stance and alignment
Finally, prosody is currently coming to be appreciated as a resource for the display

and negotiation of stance and alignment. The most persuasive research findings
here have situated the use of specific prosodic and phonetic practices within particular action sequences (Goodwin and Goodwin 2000). For example, in news deliveries: the valence with which a piece of news is delivered can be ascribed not
only lexically but also prosodically. Good news will often be signaled with high
pitch level, wide pitch range, animated contour and faster speech rate, while bad
news will have low pitch level, narrow pitch range, stretched vowels, slower pace
and often breathy or creaky voice quality (Freese and Maynard 1998). And if a
piece of news is receipted with a freestanding news-receipt token such as oh,
this turn will be heard as displaying surprise if the oh is lengthened and delivered
with an extra-high and pointed (rising-falling) contour13 but as displaying sympathy or disappointment if it has low pitch, compressed pitch range, soft volume
and weak and/or breathy voice (Reber 2008, Couper-Kuhlen 2009). Further, in
childrens game disputes, when cries of hey!, no! and out! are produced with sig-
500
nificant lengthening, extra high pitch and pronounced falling-rising contours, this
will mark not only disagreement but an affective dimension of strong opposition
(M. Goodwin, C. Goodwin and Yaeger-Droh 2002). As these and similar studies
show, it is by examining prosodic and phonetic resources for the implementation of
specific actions situated in particular sequences and activities that their distinctive
contribution to the display and interpretation of emotive involvement can be appreciated.
But surely one of the most fascinating tasks for which interactants deploy prosody is the negotiation of stance and alignment. In telephone openings, for instance,
callers attend to subtle indications of mood and enthusiasm as displayed by the
pitch level at which their co-participants deliver and respond to greetings and how
are yous. When pitch levels are adjusted to mutually accommodate, this leads to
the joint negotiation of the tenor at which the conversation is to be conducted
(Schegloff 1998). Similar online prosodic negotiation will be found with respect to
rhythm and pacing in conversation openings and closings on the telephone
(Couper-Kuhlen 1993, Auer, Couper-Kuhlen and Mller 1999). More generally
speaking, in virtually any kind of second position, speakers adjust the timing of
their incoming according to the pacing of the prior turn. The way they do this cues
meaningful inferences such as Interaction proceeding apace or I have a problem
with responding as expected (Auer, Couper-Kuhlen and Mller 1999). Recipients
also orient in other respects to their interlocutors prosody and thereby convey
alignment with the prior turn or speaker (Szczepek Reed 2006).14 This is the dimension of prosodic harmony which was missing in early treatments of intonation,
and it is perhaps one of the most exciting aspects of interactive prosodic practice
currently being investigated today.
1.9.
Summary
Looking back now at the work in prosody and interaction which has emerged over
the recent years, it becomes apparent that the nature of the enterprise has subtly
changed. For one, attention is no longer paid only to final, or nuclear pitch movement. Consideration is also given to onset level, pitch range and pitch register; to
the timing of accented and unaccented syllables; to amplitude shifts and levels of
loudness, to pausing and audible breathing; to phonation types and articulatory settings. Moreover, since the approach is an inductive one, further ways in which prosodic, phonetic and paralinguistic dimensions of speech are relevant to interaction
are likely to be discovered in the future.15 Some significant advances have already
been made. We have new ways of describing prosodic and phonetic relationships
with categories such as prosodic matching and prosodic orientation, phonetic
upgrading and downgrading, hypo- and hyper-articulation, etc. And we have
new tools for analyzing prosodic and phonetic patterns, e.g. the speech analysis program P RAAT , which is not only robust enough to handle non-laboratory data and
501
relatively large chunks of speech at a time but is also freely available and constantly
being updated.16 But we are still debating what kind of notation system is needed for
the representation of prosodic and phonetic features in talk-in-interaction.17
2.
Looking forward
What does the future hold in store? What challenges still remain for research into
prosody and interaction? Two recent developments are bringing about some reorientation among students of interactional prosody. They concern (i) the scope of
the field and (ii) its status as a resource for social action in interaction.
2.1.
Scope of prosody
What should we call prosodic? With the increasing awareness that many phonatory, articulatory and voice-quality parameters form clusters along with traditional prosodic parameters for structuring conversation and conveying interactional meaning, the original narrow understanding of prosody as restricted to pitch,
loudness and timing is gradually giving way. Just like the more traditional prosodic
dimensions, phonetic and paralinguistic features also convey what belongs together and what does not (the disjunctive-conjunctive dimension of talk) as well as
what matches and/or upgrades a prior turn-unit and what downgrades it (the affiliative-disaffiliative dimension of talk). For this reason, it is becoming increasingly
artificial to restrict ones attention to a mere subset of these phenomena. Local has
described the situation this way:
() in structuring their talk-in-interaction speakers systematically draw on bundles of
phonetic features which cut across the traditional classification of phonetic parameters
into prosodic on one hand versus articulatory (or segmental) on the other. This suggests that phonetic parameters are best treated as falling into functional bundles or
clusters, irrespective of their prosodic or segmental characteristics, on the basis of
how speakers deploy them to achieve particular interactional goals. (Local 2004:396)
So prosodists today are faced with a terminological dilemma: what to call the prosodic, alias paralinguistic, alias phonetic features which speakers routinely use for
the management of interaction in all its structural/organizational and its ritual/
interactional facets? Should everything be called phonetic? This runs the risk of
making the object of inquiry indistinguishable from that of linguistic phoneticians
whose concern is the phonetic-phonemic interface. Should everything be called
prosodic? This runs the risk of allowing articulatory dimensions to be forgotten
along the way. One is reminded of the debate over sexism in language: are women
made invisible through the use of linguistic forms such as everyone he? Just as
with sexist language, so too discriminatory scientific labels should be avoided. The
solution may thus lie in joint mention, with reference being made, where relevant,
502
to prosodic and phonetic dimensions of language use in interaction. On this

understanding, the present chapter might be more aptly titled: Prosody and phonetics as social action.
2.2.
Status of prosody
How central to social interaction is prosody? With recent advances in video technology there has been a concomitant increase in the attention paid to so-called
multi-modality, i.e. to those embodied and visually based semiotic systems implicated in the conduct of interaction. Prosody is thus no longer an only child; it
has acquired a number of siblings and must now take its place alongside other nonverbal practices for conducting interaction, including gaze, facial expression, gesture, body posture and kinesics. Ironically, among all the various types of embodiment, prosody for decades treated as lying at the edge of language (Bolinger
1972) now appears the most strongly linguistic. Despite the fact that many linguists continue to treat it as at best paralinguistic, in the larger picture of interactional conduct prosody is much more closely tied to language than is, say, gaze
or facial expression. This is in part because prosody is intrinsic to speech and
speech-like vocalization, and in part because prosodic resources are primarily
auditory and are thus organized linearly in ways in which visible resources are not.
So prosody, and research on prosody in interaction, must find a new place in the
multimodal world of interaction analysis. And unless language is to be prioritized,
that place may not be front stage.
Compared to the more dominant, visual semiotic systems deployed in interaction, prosodic and phonetic resources may appear somewhat limited, especially if
they are treated as auditory only. Actually there is often a visible component to the
prosodic and phonetic formatting of talk: so-called smile voice, for instance, is
quite naturally embodied in spread lips and is therefore clearly visible. Surges in
loudness and pitch are typically accompanied by noticeable changes in jaw and larynx position. The temporal onset and offset of speech is visible, and many articulations, not to mention articulatory settings (e.g. labialization), have clearly visible
dimensions. In short, it would behoove prosodists and phoneticians to pay more attention to the visible cues associated with prosodic and phonetic practices. In this
respect prosody and phonetics also have a visual contribution to make. But surely
their most important contribution to multimodal interaction is time or temporality.
In C. Goodwins words: time is an intrinsic component of the units used to
build events and action within situated human interaction. The location of relevant
units is a key task facing both analysts and participants (C. Goodwin 2002:
47). Even in the larger, multimodal picture, temporality remains the backbone of
interaction: it is in real time that interactive events emerge and position themselves
sequentially and simultaneously. And prosodists and phoneticians can help track
this temporal unfolding.
2.3.
503
The way forward
Despite the dominance of multimodality in the world of interaction, prosody and

phonetics in social action remain a central concern both for participants and for
analysts. Without prosodic and phonetic resources, interactants would lack a crucial means for implementing actions indirectly; they would have no way of determining when a particular action is due or overdue; they would be ill-equipped to
mark disjunctive relationships between adjacent pieces of talk or conjunctive relationships between non-adjacent ones; the beginnings of their courses of action
would be lacking in projectability and therefore indistinct from endings; their repairs would risk being taken for continuations; their stances towards actions and towards co-participants would remain opaque, etc. It has often been claimed that
meaning in interaction is dependent on composition (what sorts of parts) and on
position (located where) (e.g. Schegloff 2007: 81). Yet just as crucial is the question of manner (produced how): how high or low, how soft or loud, how fast or
slow is a piece of talk? Is it well-timed or delayed, glottalized or not, said with
spread lips (smile voice) or not? It is answers to such questions that help illuminate the omnipresent issue in interaction: Why that now? (Schegloff and Sacks
1973: 299). The way forward in the field of pragmatics and social interaction will
thus require continued attention to minute prosodic and phonetic detail. Concomitantly, it will require a flexible and, where needed, prosodically and phonetically
explicit transcription system. But it will also demand of its practitioners the willingness and ability to integrate prosodic and phonetic systems with other semiotic
modalities in order to fully appreciate the rich multimodal ecology of sign systems (Goodwin 2002) which underlies human social interaction.
Notes
1. As will become apparent in the course of this chapter, however, the actual delimitation of
the field of prosody is currently a subject of debate.
2. By we I mean those engaged in the empirical and in vivo investigation of prosody as a
pragmatic resource.
3. For reasons of space the following list is necessarily selective (the examples mentioned
all involve word repetition, for instance). Here and elsewhere in this chapter, barring indication to the contrary, mentions are restricted to work on English.
4. In Ogdens most telling case, a second speaker responds to the assessing turn isnt that
good at long last with thats thats (.) very good news but in a way which is prosodically
and phonetically downgraded. In subsequent talk the latter turn is treated as intimating
mild disagreement with the co-participants first assessment (2006: 1771).
5. These are interrogative turn-constructional units which propose a new topic of talk, either
after the opening section of the call or later in the body of the conversation after a prior
topic has been brought to possible closure (Couper-Kuhlen 2004, Schegloff 2007).
6. If only for this reason the negative results of de Ruiter, Mitterer and Enfields laboratory
experiments on final intonation in Dutch turn-taking (2006) are not surprising.
504
7. Competitive incoming refers to a would-be speaker trying to take over the floor immediately, although the current speaker is nowhere near completing their turn.
8. It is in part because articulatory and laryngeal settings are often co-implicated in this
kind of cohesive linking that some scholars have argued for speaking of phonetic instead of prosodic resources (Local and Walker 2005). See also section 2.1.
9. Isochronous refers to a type of rhythmization in which rhythmic intervals have
approximately the same temporal duration; isometric refers to another type of rhythmization in which rhythmic intervals have the same number of syllables (Auer, CouperKuhlen and Mller 1999).
10. An increment is a stretch of same-speaker talk which (a) is produced after a point of
possible completion in that speakers turn, and (b) is syntactically symbiotic with the
prior talk (cf. e.g. Schegloff 1996, Couper-Kuhlen and Ono 2007).
11. Collaborative completion is a term used to refer to a participant finishing a turn-constructional unit which their interlocutor has left (provisionally) incomplete. Collaborative continuation refers to a participant adding on to a turn-constructional unit which
their interlocutor has just (provisionally) completed.
12. This typically happens after the production of uh or uhm (Schegloff 2006).
13. This is especially so when the news being delivered has been designed to elicit surprise
(Wilkinson and Kitzinger 2006). Cf. also Local (1996), and Selting (1996b) for German.
14. Szecepek Reed (2006), for example, documents prosodic orientation involving intonation contour, pitch step-up, register, loudness, speech rate and voice quality.
15. One good example of a recent discovery comes from work on crying in interaction
(Hepburn 2004, Hepburn and Potter 2007), a follow-up to Jeffersons (1985) study of
laughter.
16. Developed by Paul Boersma and David Weenink, Institute of Phonetic Sciences, University of Amsterdam; available at www.praat.org.
17. Among the options are the autosegmental, laboratory-based ToBI system (Beckman and
Ayers 1994), the discourse-based Discourse Transcription system (Du Bois et al 1992)
and the CA-based Gesprchsanalytisches Transkriptionssystem: GAT (Selting et al
1998, 2009).
References
Anward, Jan
2002
Other voices, other sources. In: Per Linell and Karin Aronsson (eds.), Jagen
och rsterna: Goffman, Viveka och samtalet / Selves and Voices: Goffman,
Viveka, and Dialogue, 128147. Linkping: Tema K, Linkpings Universitet.
Armstrong, Lilias E. and Ida C. Ward
1926
A Handbook of English Intonation. Cambridge: W. Heffer and Sons.
Auer, Peter
1996
On the prosody and syntax of turn-continuations. In: Elisabeth Couper-Kuhlen
and Margret Selting (eds.), Prosody in Conversation, 57100. Cambridge:
Auer, Peter, Elizabeth Couper-Kuhlen and Frank Mller
1999
Language in Time: The rhythm and tempo of spoken interaction. New York:
505
Beckman, Mary E. and Gayle. M. Ayers

1994
Guidelines for ToBI labeling. Version 2.0. Columbus, OH, Ohio State University, Department of Linguistics.
Berger, Peter and Thomas Luckmann
1967
The Social Construction of Reality. Harmondsworth: Penguin.
Bolinger, Dwight
1972
Intonation. Selected readings. Penguin: Harmondsworth, Middlesex.
Couper-Kuhlen, Elizabeth
1986
An Introduction to English Prosody. Tbingen: Max Niemeyer and London:
Edward Arnold.
1993
English Speech Rhythm: Form and function in everyday verbal interaction.
1996
The prosody of repetition: On quoting and mimicry. In: Elizabeth CouperKuhlen and Margret Selting (eds.), Prosody in Conversation, 366405. Cambridge: Cambridge University Press.
1999
Coherent voicing: On prosody in conversational reported speech. In: Wolfram
Bublitz and Uta Lenk (eds.), Coherence in Spoken and Written Discourse:
How to create it and how to describe it, 1132. Amsterdam: Benjamins.
2001
Interactional prosody: High onsets in reason-for-the-call turns. Language in
Society 30: 2953.
2004
Prosody and sequence organization: The case of new beginnings. In: Elizabeth
Couper-Kuhlen and Cecilia E. Ford (eds.), Sound Patterns in Interaction,
2009
A sequential approach to affect: The case of disappointment. In: Markku
Haakana, Minna Laakso and Jan Lindstrm (eds.), Talk in Interaction comparative dimension. Helsinki, Finnish Literature Society.
Forthc. Some truths and untruths about final intonation in conversational questions. In:
Jan P. de Ruiter (ed.), Questions, 94123. Cambridge: Cambridge University
Press.
2007
Incrementing in conversation. A comparison of practices in English, German and Japanese. Pragmatics 17.4: 513552.
Crystal, David
1969
Prosodic Systems and Intonation in English. Cambridge, Cambridge University Press.
Curl, Traci S.
2004
Repetition repairs: The relationship of phonetic structure and sequence organization. In: Elizabeth Couper-Kuhlen and Cecilia E. Ford (eds.), Sound
Patterns in Interaction, 273298. Amsterdam, Benjamins.
506
Curl, Traci S.
2005
Practices in other-initiated repair resolution: the phonetic differentiation of
repetitions. Discourse Processes 39.1: 144.
Cutler, Anne and D. Robert Ladd (eds.)
1983
Prosody: Models and measurements. Berlin: Springer.
Du Bois, John W., Stephan Schuetze-Coburn et al.
1992
Discourse Transcription. Santa Barbara Papers in Linguistics 4: 1225.
Ford, Cecilia E. and Sandra A.Thompson
1996
Interactional units in conversation. Syntactic; intonational and pragmatic resources for the management of turns. In: Elinor Ochs, Emanuel A. Schegloff,
and Sandra A. Thompson (eds.), Interaction and Grammar, 134184. Cambridge: Cambridge University Press.
Ford, Cecilia E., Barbara A. Fox and Sandra A.Thompson
1996
Practices in the construction of turns: The TCU revisited. Pragmatics 6:
427454.
Ford, Cecilia E, Barbara A. Fox and John Hellermann
2004
Getting past no: Sequence, action and sound production in the projection of
no-initiated turns. In: Elizabeth Couper-Kuhlen and Cecilia E. Ford (eds.),
Sound Patterns in Interaction, 233272. Amsterdam: Benjamins.
Freese, Jeremy and Douglas W. Maynard
1998
Prosodic features of bad news and good news in conversation. Language in
Society 27:195219.
French, Peter and John K. Local
1983
Turn-competitive incomings. Journal of Pragmatics 7: 1738.
Goldberg, JoAnn
1978
Amplitude shift: A mechanism for the affiliation of utterances in conversational interaction. In: Jim Schenkein (ed.), Studies in the Organization of Conversational Interaction, 199218. New York, New York Academic Press.
Goodwin, Charles
2002
Time in action. Current Anthropology 43: 1935.
Goodwin, Marjorie and Goodwin, Charles
2000
Emotion within situated activity. In: Alessandro Duranti (ed.), Linguistic Anthropology: A Reader, 239257. Oxford: Blackwell.
Goodwin, Marjorie H., Charles Goodwin and Malcah Yaeger-Droh
2002
Multi-modality in girls game disputes. Journal of Pragmatics 34: 16211649.
Gumperz, John
1982
Halliday, Michael A. K.
1967
Intonation and Grammar in British English. The Hague: Mouton.
Hepburn, Alexa
2004
Crying: Notes on description, transcription and interaction. Research on Language and Social Interaction 37: 251290.
Hepburn, Alexa and Jonathan Potter
2007
Crying receipts: Time, empathy, and institutional practice. Research on Language and Social Interaction 40.1: 89116.
Holt, Elizabeth and Rebecca Clift (eds.)
2007
Reporting Talk: Reported speech in interaction. Cambridge: Cambridge University Press.
507
Jasperson, Robert
2002
Some linguistic aspects of closure cut-off. In: Cecilia E. Ford, Barbara A. Fox
and Sandra A. Thompson (eds.), The Language of Turn and Sequence,
Jefferson, Gail
1983
On a failed hypothesis: Conjunctionals as overlap-vulnerable. Tilberg Papers
in Language and Literature 28: 133.
Jefferson, Gail
1985
An exercise in the transcription and analysis of laughter. In: Teun van Dijk
(ed.), Handbook of Discourse Analysis. Vol. 3, 2534. London, Academic.
Kelly, John and John Local
1989
Doing Phonology. Manchester: Manchester University Press.
Klewitz, Gabriele and Elizabeth Couper-Kuhlen
1999
Quote-unquote. The role of prosody in the contextualization of reported speech
sequences. Pragmatics. 9.4: 459485.
Linell, Per
2005
The Written Language Bias in Linguistics. Its nature, origins and transformations. London: Routledge.
Local, John
1992
Continuing and restarting. In: Peter Auer and Aldo di Luzio (eds.), The Contextualization of Language, 273296. Amsterdam, Benjamins.
Local, John
1996
Conversational phonetics: Some aspects of news receipts in everyday talk. In:
Elizabeth Couper-Kuhlen and Margret Selting (eds.), Prosody in Conversation: Interactional studies, 177230. Cambridge: Cambridge University
Press.
Local, John
2004
Getting back to prior talk: and-uh(m) as a back-connecting device in British
and American English. In: Elizabeth Couper-Kuhlen and Cecilia E. Ford
(eds.), Sound Patterns in Interaction, 377400. Amsterdam: Benjamins.
Local, John
2005
On the interactional and phonetic design of collaborative completions. In:
W. Hardcastle and J. Beck (eds.), A Figure of Speech: a Festschrift for John
Laver, 263282. Mahwah, NJ: Lawrence Erlbaum.
Local, John and John Kelly
1986
Projection and silences: Notes on phonetic and conversational structure.
Human Studies 9: 185204.
Local, John, John Kelly and William H.G. Wells
1986
Towards a phonology of conversation: Turn-taking in Tyneside English. Journal of Linguistics 22: 411437.
Local, John and Gareth Walker
2004
Abrupt-joins as a resource for the production of multi-unit, multi-action turns.
2005
Methodological imperatives for investigating the phonetic organization and
phonological structures of spontaneous speech. Phonetica 62: 120130.
508
Mazeland, Harrie
2007
Parenthetical sequences. Journal of Pragmatics 39: 18161869.
OConnor, J.D. and Arnold, Gordon F.
1961
Intonation of Colloquial English. London: Longmans.
Ogden, Richard
2004
Non-modal voice quality and turn-taking in Finnish. In: Elizabeth CouperKuhlen and Cecilia E. Ford (eds.), Sound Patterns in Interaction, 2962.
Ogden, Richard
2006
Phonetics and social action in agreements and disagreements. Journal of Pragmatics 38: 17521775.
Ogden, Richard and Sara Routarinne
2005
The communicative function of final rises in Finnish intonation. Phonetica 62:
160175.
Reber, Elisabeth
2008
Affectivity in Talk-in-interaction: Sound objects in English. PhD diss., University of Potsdam.
de Ruiter, Jan Peter, Holger Mitterer and Nick J. Enfield
2006
Predicting the end of a speakers turn; a cognitive cornerstone of conversation.
Language 82.3: 515535.
1974
1987
Analyzing Single Episodes of Interaction: An Exercise in Conversation Analysis. Social Psychology Quarterly 50.2: 101114.
1990
On the organization of sequences as a source of coherence in talk-in-interaction. In: B. Dorval (ed.), Conversational Organization and its Development,
5177. Norwood, NJ, Ablex.
1996
Ochs, Emanuel A. Schegloff and Sandra A. Thompson (eds.), Interaction and
Grammar, 52133. Cambridge: Cambridge University Press.
1998
Reflections on studying prosody in talk-in-interaction. Language and Speech
41: 235263.
2006
On uh and uhm and some of the things they are used to do. brandial, Potsdam, September 1113, 2006.
2007
Sequence Organization in Interaction. A primer in conversation analysis I.
1973
Schubiger, Maria
1958
English Intonation: Its form and function. Tbingen, Niemeyer.
509
Selting, Margret
1996a
On the interplay of syntax and prosody in the constitution of turn-constructional units and turns in conversation. Pragmatics 6: 371388.
Selting, Margret
1996b
Prosody as an activity-type distinctive cue in conversation: the case of socalled astonished questions in repair. In: Elizabeth Couper-Kuhlen and Margret Selting (eds.), Prosody in Conversation, 231270. Cambridge: Cambridge
University Press.
Selting, Margret
2000
The construction of units in conversational talk. In: Language in Society 29:
477517.
Selting, Margret
2004
The upward staircase intonation contour in the Berlin vernacular: An
example of the analysis of regionalized intonation as an interactional resource.
In: Elizabeth Couper-Kuhlen and Cecilia E. Ford (eds.), Sound Patterns in Interaction, 201232. Amsterdam: Benjamins.
Selting, Margret
2005
Syntax and prosody as methods for the construction and identification of turnconstructional units in conversation. In: Auli Hakulinen and Margret Selting
(eds.), Syntax and Lexis in Conversation, 1744. Amsterdam: Benjamins.
Selting, Margret
2007
Lists as embedded structures and the prosody of list construction as an interactional resource. Journal of Pragmatics 39: 483526.
Selting, Margret, Peter Auer, Birgit Barden, Jrg Bergmann, Elizabeth Couper-Kuhlen, Susanne Gnthner, Christoph Meier, Uta Quasthoff, Peter Schlobinski, and Susanne Uhmann
1998
Gesprchsanalytisches Transkriptionssystem (GAT). Linguistische Berichte
173: 91122.
Selting, Margret, Peter Auer, Dagmar Barth-Weingarten, Jrg Bergmann, Pia Bergmann,
Karin Birkner, Elizabeth Couper-Kuhlen, Arnulf Deppermann, Peter Gilles,
Susanne Gnthner, Martin Hartung, Friederike Kern, Christine Mertzlufft,
Christian Meyer, Miriam Morek, Frank Oberzaucher, Jrg Peters, Uta Quasthoff, Reinhold Schmitt, Wilfried Schtte, Anja Stukenbrock, and Susanne Uhmann
2009
Gesprchsanalytisches Transkriptionssystem 2 (GAT 2). Gesprchsforschung
Online (http://www.gespraechsforschung-ozs.de/).
Szczepek Reed, Beatrice
2004
Turn-final intonation in English. In: Elizabeth Couper-Kuhlen and Cecilia E.
Ford (eds.), Sound Patterns in Interaction, 97118. Amsterdam, Benjamins.
2006
Prosodic Orientation in English Conversation. Basingstoke: Palgrave Macmillan.
Tanaka, Hiroko
2000
Turn-projection in Japanese talk-in-interaction. Research on Language and
Social Interaction 33.1: 138.
Tanaka, Hiroko
2004
Prosody for marking transition-relevance places in Japanese conversation:
The case of turns unmarked by utterance-final objects. In: Elizabeth Couper-
510
Kuhlen and Cecilia E. Ford (eds.), Sound Patterns in Interaction. 6396.

Amsterdam, Benjamins.
Tarplee, Clare
1996
Working on young childrens utterances: Prosodic aspects of repetition during
picture labelling. In: Elizabeth Couper-Kuhlen and Margret Selting (eds.),
Prosody in Conversation, 406435. Cambridge: Cambridge University Press.
Trager, George. L. and Henry L. J. Smith
1957
An outline of English structure. Studies in Linguistics: Occasional Papers 3:
4052.
Uhmann, Susanne
1992
Contextualizing relevance: On some forms and functions of speech rate
changes in everyday conversation. In: Peter Auer and Aldo di Luzio (eds.), The
Contextualization of Language, 297336. Amsterdam: Benjamins.
Walker, Gareth
2004
On some interactional and phonetic properties of increments to turns in talk in
interaction. In: Elizabeth Couper-Kuhlen and Cecilia E. Ford (eds.), Sound
Wells, Bill and Sue Pepp
1996
Ending up in Ulster. Prosody and turn-taking in English dialects. In: Elizabeth
Couper-Kuhlen and Margret Selting (eds.), Prosody in Conversation, 101
Wells, Bill and Sarah Macfarlane
1998
Prosody as an interactional resource: Turn-projection and overlap. Language
and Speech 41: 26594.
Wichmann, Anne
2001
Spoken parentheticals. In: Karin Aijmer (ed.), A Wealth of English, 177193.
Gteborg: Gothenburg University Press.
Wilkinson, Sue and Celia Kitzinger
2006
Surprise as an interactional achievement: Reaction tokens in conversation.
Social Psychology Quarterly 69.2: 150182.
Wright, Melissa
2007
Clicks as markers of new sequences in English conversation. In: Proceedings
of the 16 th International Congress of Phonetic Sciences, Saarbrcken, 1069
1072.
18.

Jacob L. Mey
1.
Introduction
Literary pragmatics is about how language is used in the production and consumption of literary texts. Since much of literature is about telling stories (including the
stories that we value as lyrical expressions of a poets feelings), the notion of narrativity looms large here. To guide the author and reader in their collaborative work
of creating and re-creating the fictional universe of the story, the concept of voice
is of paramount importance. Voices represent characters, as they are created by the
author and perceived, re-created, by the reader; it is the characters that in the end
determine the storys importance rather than the simple plot, or the action, as it is
nowadays called. Important to notice here is that even when the storyteller narrates
own or others real experiences (as in autobiography, or when expressing a reflection on a factual event), the universe that is created by this narration is still entirely
fictional, and so are the characters, real or imagined, that populate the story and the
voices that express their respective points of view. The reader collaborates with
the author in identifying (re-creating) the characters, as they are entering on the
literary scene and voicing their parts. But the voices do not always sound in perfect
unison or dramatic counter-point; when voices become mixed up and clash, the
reader becomes confused and may abandon the common literary enterprise altogether.
Thus, voice clashes (along with the related phenomena of transition between,
and confusion of, voices) offer an interesting indirect perspective on the roles of
authors and readers in the literary enterprise. These roles will be seen as intimately
interrelated, indeed dialectically presupposing each other: the literary enterprise is
a collaborative effort.
2.
Storytelling and stories
Contrary to what many people believe, storytelling is never a matter of getting the
facts, all the facts, and nothing but the facts. What makes a story interesting is the
way we consider those facts as they are presented to us by a narrator; this is what
makes a story different from a simple retelling of factual happenings, as in the familiar kind of grade school composition that starts each sentence with And then .
Stories are part of our lives: they extend into our lives and branch out from
there; but also, stories can only be understood if they are localized, fleshed out in
512
Jacob L. Mey
terms of our daily reality. A story that has no relation to my world is probably not
going to interest me very much. Stories of earlier centuries (like the few surviving
novels of classical antiquity or the interminable chevaleresque poems of the
Middle Ages) appear often boring to a modern reader by their monotonic, endless
repetition of the same themes: but these very qualities may have been what endeared this literature to the contemporary readership and contributed to its popularity. Similarly, the 18th century British author Samuel Richardsons Sir Charles
Grandison (an endless epistolary novel in seven ( ! ) volumes, called Series of
Letters) must have received some appreciation among its contemporaries; but already a century later both Jane Austen (in Northanger Abbey) and Aleksander S.
Pushkin in Evgenij Onegin saw fit to quote the hapless hero as the paragon of reading-induced boredom; compare Pushkins sneering reference to
Grandison, the precious bore
Whose gallant lines provoke our snore
(bezpodobnyj Grandison
kotoryj nam navodit son )
Evgenij Onegin (ch. III, verse ix; my translation)
Stories like Grandison and its predecessors Clarissa and Pamela seem end- and
pointless, precisely because we cannot relate to the reality from which they derive
and to which they point back.
On the other hand, we also find stories in older literature, which today seem as
fresh as when they first were written down. I already mentioned Austen and Pushkin: both furnish outstanding examples of stories that are as fascinating and pageturning today as they were when they were written, nearly two centuries ago. For
who among us has not chuckled over the pointed ironies of Austens characters, or
swooned with Tatiana in her romantic, unrequited love for Onegin? Or (to take an
even earlier case): Boccaccios Decameron is, after three quarters of a millennium,
still good for a thrill and a laugh, on a great number of exciting levels.
While it is difficult to pinpoint what makes a story succeed, it is perhaps easier
to spot at least a few of the causes that make a story fail, or let it fall into oblivion
after the story producers and the consumer generation that supported them have
passed away. In his Evgenij Onegin, Pushkin satirizes the ennui of the Russian genteel country nobility by pointing to the fabulous array of has-been authors who
constituted the daily spiritual bread of the country manor, and whose fame faded as
their pages yellowed (the incomparable Grandison, whose bore is matched by our
snore, is only one among many).
In this and similar cases, the interested readership is confined to students of literature or history; the regular consumers will find it difficult to relate to the characters and their voices, as these represent an entirely different social and societal
environment. The question thus is what makes for a story, and in particular, for a
513
good story. Literary scholars and the common readership have pondered about this
question for centuries, both theoretically, in studies of narrative, and practically, in
university courses of the Creative Writing 101 type and in popular do-it-yourself manuals.
In the following, I will first consider what we mean by narrativity: briefly, the
question is about the essential features of story-telling. Next, the voices of the
narrative will be brought into play, and I will show how they interact, and sometimes conflict. After that, I will identify some, but not all of the linguistic and pragmatic devices that authors use when they tell their stories and that readers rely on
for a successful interpretation and consumption of the narrative. And finally, I want
to go back to the key figures in the process of literary production and consumption:
the authors and the readers, as seen at work in their collaborative efforts, and as
perceived through their respective voices.
3.
Narrativity and voices
Narration is about storytelling. But a story has to be told by someone; let us call
him or her the narrator. Already, it is clear that the narrator is different from the
author: in every story, the author creates a world of fiction, the narrative, in which
the narrator plays an eminent role, even though not always discernible on the surface. Narrativity thus deals with the techniques and devices that a narrator has at
his or her command when telling a story; among these, the notion of character
and the accompanying voice are among the most important.
To see how character and voice connect and interrelate in sometimes rather oblique and inscrutable ways, consider the case of the homodiegetic novel, or
stories told in the first person. Clearly, we cannot assume that the speaking I is
identical to the author, who likewise cannot be held responsible for the narrators
claim as being his own. When the 19th century Russian Mikhail Yu. Lermontov
starts his classic story A hero of our times (1841) with the famous line I traveled
by stagecoach from Tbilisi , every reader will understand that this does not
mean that Lermontov himself did the traveling: the words are spoken by the narrator (referring to his conversation while sitting on top of the coach with a returning NCO, the ranking shtabskapitan serving in Chechenya).
So, the author is not the narrator, not even if the story is told in the first person
singular, as in the Lermontov case. The author creates the narrator as the embodiment of the storys authoritative competence; it is really not important whether this
authority assumes the I-role or pretends to be an outside voice: Trust the tale,
not the teller, as the old adage has it. What characterizes the story is its narrativity, its quality of being narrated and received as a story. The latter part is important, but it is often overlooked; Robert Scholes has pointed out that the very idea of
narrativity is dependent on the reader actively entering the story world and partici-
514
Jacob L. Mey
pating in the narration: a process by which a perceiver [author or reader, JM] actively constructs a story from the fictional data provided by any narrative medium
(1982: 60). Thus, on the part of the reader, there corresponds to the act of narration
an act of active collaboration, an active narrativity.
However we define narrativity, there is a need to distinguish between author
and narrator. A further necessary distinction is that between narrator and actor, and
this is where the concept of narrative level comes in. Following Bal (1985), I distinguish between a primary level, where the action unfolds, and a secondary (and
possibly further) level(s), where the actors come in and their voices are heard. On
the primary level, we have the narrators text; on the secondary (embedded)
level, we are confronted with the actors text. The latter text does not necessarily
contain only what the actors say or think in the context of the primary level; they
may start a story of their own, and this embedded narrative may then again have
several levels of its own, where the secondary story is related to the primary one in
various ways.
A good example is provided by the classic novella by the German author Theodor Storm, Der Schimmelreiter (The man on the white steed, 1888; Engl. transl.
The Dyke Master, 1996), where the narrator on the primary level is the authoritative retired school teacher, who frames the story by telling it to his friends in the
local pub (in German, this kind of narrative is called a Rahmenerzhlung, lit. a
story within a (narrative) frame). But the narrative is not about the teacher and his
audience; rather, it concerns the legendary local official in charge of dams and levees (in German: der Deichgraf, the dyke master), whose expertise and innovative efforts form the secondary narrative, which in fact is the story that we, the
readers, co-construct and remember. The teachers primary voice, as perceived in
his narration and buttressed by his memory, lends credence and local coloring to
the embedded story and its secondary voices.
Similarly, in one of the most famous collections of stories ever told, the 1001
Nights, the main character, Sheherazade, invents and orchestrates the stories,
which she then attributes to secondary narrators on the embedded level. These secondary narrators remain implicit, that is, they are heard, but not seen; in contrast,
another famous collection of stories, Boccaccios Decameron, has the actors on the
primary level narrate their own stories as narrators of secondary level-stories.
In what I have said so far, I have implicitly assumed that actors speak in narratives; in other words, I have attributed them a voice. But how does this happen?
More precisely: How do we know whose voices we are hearing in a narrative? How
do we decide if what we are hearing (and this includes of course reading) is
spoken by the narrator, by an actor (primary or secondary), or by some other
(third) party in a story? To better understand, and answer this question, let me introduce another notion: that of focus, or focalizing, as it is often called (Bal 1985).
Focalizing is what makes an uninteresting series of events into a potentially
fascinating story. To illustrate my point, I will recount a personal happening in
515
which some unrelated events somehow suddenly coalesced into something resembling a story.1
At one time during the winter semester of 1994, I was sitting in the sauna at the
Crown Aquatic Center of Northwestern University, Evanston, Ill. I was enjoying
myself, thinking about nothing in particular, paying not much attention to what
was going on either inside or outside of the sauna room. People came in and went
out again; some took a shower, some stayed longer, some shorter, some were reading their newspapers, others were talking, some were throwing water on the stove
to produce steam and get the temperature up.
All this had been going on for quite a while without my paying much attention.
But after a while, I became aware of a balding, middle-aged man sitting on a chair
outside the sauna and looking in through the big glass door. I remembered him
vaguely: he had been inside and gone out, then had come back in again; now, having taken another shower, he was sitting outside, watching. I am directly in his line
of vision, and suddenly my perspective is no longer that of one watching a scene
with some persons and objects in the background, but of me watching some person
watching me. In other words, I start focalizing.
My immediate, conscious reaction is, of course, that I am making things up.
The man may be there for a number of reasons, and why should he be watching
me? Still, something interesting happens the moment I realize I am no longer observing an inconsequential series of events, or looking at some irrelevant piece of
humanity, but at somebody who is somebody, even though I dont know him. When
I began to focalize, what happened is that I set in motion a narrative process: in
other words, I started making up a story in my head.
Of course, most of this story is just daydreaming, and it happens somewhere
below the explicit level of consciousness. For example, depending on my previous
experiences (or, if you want, my private hang-ups), I attribute a character to the
man, trying to figure out if this is somebody I know, or ought to remember. In a
more interesting scenario, I might consider him to be an elderly homosexual, waiting for me to get out of the sauna so he can make a pass at me.
Now, even though my story can be said to be pure, self-centered fantasy (and
even though I may know it to be exactly that), in terms of narrativity it is still the
case that I have given this person a role as a character in my story, an actor in my
little play; and the next thing I know is that the character starts speaking, starts having a voice. In my head, I even hear snatches of dialogue: the character is asking me
questions (Do I come here often; where do I live; what kind of music do I like; do I
have a phone number, and so on), and I listen to myself providing non-committal,
evasive answers (No, not really; somewhere in South Evanston; no particular
preferences; no, I dont have a phone, etc.). Thus, my focalization makes an indifferent piece of my environment come alive in my self-constructed narrative; from
being a silent prop, the person outside the sauna becomes a talking head. Focalizing implies vocalizing, that is, providing a voice to a character in a narrative.
516
Jacob L. Mey
No matter that the above ctional dialogue never took place, and never will, and
discounting any possible psychological explanations of my sauna daydreaming (earlier experiences, repressed feelings, anxieties, or what have you), what this little anecdote is supposed to show is that we cannot separate focus (and in general, the narrative perspective) and voice. Moreover, focalization presupposes a point of view,
inasmuch as it belongs to a character; hence, every focalization happens according to
a characters perspective. Similarly, every characters voice presupposes the existence of a focalizing perspective in which that character is vocalized. In the next section, I will look at how the characters voices interact, and how they sometimes clash.
4.
Characters and voices
In a story, the characters come to life through their voices. The characters are kept
apart, and their voices separated, by means of simple typographic means (such as
quotation marks or line shifts) or in other, more sophisticated ways; the point of
these techniques is to permit the reader or listener to keep track of what is going on,
to place and/or reposition the focus of his or her attention.
Consider now the following extract:
He [Jack Kemp] joined the Buffalo Bills after an injury.
I hit a helmet with my passing hand and dislocated a finger so severely that I
had to literally decide what shape I wanted it in. So I put my hand on a football
and they put a cast over it.
In those days football was pretty Darwinian. Wed do anything to survive.
I glanced at the finger. Yep.
(From an interview with the late former Republican U.S. Congressman Jack
Kemp, who had been a professional football player in the major leagues NFL/
AFL; Washington, D.C., October 1994; Parks 1994: 22)
In the above story, we have first the voice of the narrator, the journalist who is telling us the story of his interview with Jack Kemp and provides both a historical
frame (He joined the [minor league team] Buffalo Bills after an injury) and a
running comment (In those days football was pretty Darwinian).
Then there is Kemps own, quoted voice, telling us how he hit a helmet with
[his] passing hand and dislocated a finger. According to the conventions of journalism, we are expected to believe that this was how Kemp spoke while being interviewed; it also provides the background for the continued story (again, in
Kemp-speak) of what happened to the speakers hand (I had to literally decide
what shape I wanted it in. So I put my hand on a football and they put a cast over it).
But third, we encounter a voice that has not been heard so far, and a character
that unmistakably is neither Kemp nor the journalist-narrator of the interview. The
517
focus of the narration has changed: what we are hearing in the short utterance
Yep is the voice of the journalist himself, as an active participant in the narrated
scene who suddenly makes his presence in the interview explicit (as opposed to
earlier, when he was only the implicit narrating character telling us, in a running
comment, that He [Jack Kemp] joined the Buffalo Bills after an injury).
This Yep does not belong to the interview, as it is not part of any dialogue; it
represents the journalist Parkss comment to himself on what Kemp has just told
him about the consequences of his injury. This mental comment figures as a kind of
reality check: Parks looks at the finger in question (and Kemps football-curved
hand, as he calls it later on), and affirms the correctness of the story by inspecting
the evidence.
But notice that this affirmation is not part of the interview as narrated. It is the
authors internal, unspoken comment, uttered to himself: I looked, and there it
was all right: a football-shaped finger. Even though the tone of the commentators
voice belongs characteristically in spoken language (being recognizable as such by
its typically conversational nature: yep, unlike its formal equivalent yes, belongs in direct, spoken discourse). The comment is not spoken, and the person uttering it does not belong to the interview scene. One could perhaps say that Parks,
the journalist, intrudes upon the scene by offering his two-bits of commentary
aside, as in an Ibsenian stage direction. His voice is not heard directly, but rather
as a kind of ventriloquizing, as the Russian semiotician and linguist Mikhail M.
Bakhtin has called it.2 It is (self-)reported by the character as an unspoken confirmation of what just has been stated by the interviewee in the reported, uttered dialogue.
Summing up: In the extract above, we have first the (implicit) narrator, the
journalist: his (unheard) voice records the interview for the magazine publishing it,
The Smithsonian (the official publication of the Smithsonian Institution in Washington, D.C.). Then, we have the reported voice of the ex-football player-cum-congressman, Jack Kemp: he is quoted by his very words, in direct discourse. The
third voice is that of the journalist, now coming out so to speak: he sheds his protective gear and changes his status from implicit to explicit narrator: I glanced
at the finger; and he follows up by quoting himself in an internal, likewise unheard but clearly identifiable sotto voce: Yep.
From this little exercise in voicing, we see that in narrative, characters tend to
maintain their respective voices (and their corresponding stances, or points of
view) throughout the narration. They do not speak out of order or out of turn; their
vocalizations help us focalize, such that we do not lose the thread of the story, or
are uncertain of who is speaking or what any of the characters is saying.
In the next sections, we will see that voices not only collaborate in unison but
that they sometimes engage in intra-narrative (and sometimes even extra-narrative) conflicts: the case of the clashing voices.
518
Jacob L. Mey
5.
When voices clash
5.1.
Free Indirect Discourse (FID)
In the present section, I want to examine the phenomenon of vocalization under the
general angle of what I above called focalization (often referred to, as I have done
above, by the term point of view). Since the point from which a voice is heard is
directly dependent on the place of the voicer in the story, it will be useful to
examine this place (and its surrounding discourse) in some detail. I will illustrate
my findings with appropriate examples culled from literary and other discourse; in
particular, many of my examples will be taken from what is often referred to as
free indirect discourse, FID, or represented speech and thought, as it is also
called.
The difference between direct and indirect discourse on the one hand, and FID
on the other, is that in the former, we have explicit devices indicating the point
from which the stream of narration emanates. This point is either localized in the
narrative instance (as in most of traditional novel writing), or in one or the other of
the actors (as it happens in dialogue); or finally (and exceptionally), in the author
him- or herself. Whereas in both direct and indirect discourse, such origins are
marked with conventional means (e.g. quotation marks, or some form of a verbum
dicendi et sentiendi, or an inquit),3 in the case of FID we have no such formal
marks.
FID is especially favored among writers who exploit the technique known as
stream of consciousness, a technique often prototypically associated with modern writers (such as Joyce or Hemingway), but by no means a recent invention or
innovation. Here is an example from Jane Austen:
when [Catherine] thought of Henry, so near, yet so unconscious, her grief and agitation were excessive. The day which she had spent at that place had been one of the
happiest of her life. It was there, it was on that day, that the General had made use of
such expressions with regard to Henry and herself, had so spoken and so looked as to
give her the most positive conviction of his actually wishing their marriage. Yes, only
ten days ago had he elated her by his pointed regard had he even confused her by his
too significant reference! And now what had she done or what had she omitted to do,
to merit such a change?
Jane Austen, Northanger Abbey (Chawton edition, pp. 190191).
In this fragment, even though Catherine is not mentioned as saying or thinking,

there is no doubt that the discourse is Catherines. The personal pronoun she in
the second sentence and onwards refers univocally to the novels heroine. But
since there is no formal marking of this indirect discourse, we call it free, as opposed to regular indirect discourse, which is preceded by a verbum dicendi or sentiendi (such as Catherine said to herself, or Catherine thought).
5.2.
519
Voices clashing in discourse
The interesting thing about FID is that it seems so natural, so without effort, both
on the part of the reader and of the author; yet, this naturalness is an illusion. It does
not take much effort to provoke a clash of voices, wreaking havoc in the discourse, such that the true origin of a particular voice, the characters point of view
is obscured, or even made to disappear.
The French semiotician Roland Barthes has suggested using what one could
call a literary litmus test in order to determine the true provenience of a particular
voice in the discourse. The test consists in exchanging the pronominal reference
for a narrated third person with a (narrating) first person, and then observing how
the various voices get repositioned with regard to their origin in the discourse. The
technique is applied by Robert Scholes to one of Ernest Hemingways short stories;
below, I will repeat Scholes experiment and discuss its implications.
According to Scholes (1982: 116117), in Hemingways A very short story
(1965/1925), there are two third person characters of apparently equal consequence (1982: 116). Scholes goes on to show how the substitution of the first person for each of the two third persons in the narrative gives an absurd result in the
case where the substituted third person in reality was not the original voicer of
the FID. That is to say, the point of view represented by this character is not that of
the I character of the story, which actually represents a covert, first-person narration, as Scholes remarks; but the same thing could, of course, be said of much of
Hemingways work, making up its weakness as well as its strength.4
In the original story, the hero, an American soldier who was wounded in the
Great War (he), has to undergo an operation. While the hero is being readied for
the operation table, he wants to make sure he behaves himself well in the eyes of
the Italian nurse, Luz, with whom he has fallen in love; afterwards, during his convalescence, he takes pains to please her by performing some of her chores.
He went under the ansthetic holding tight on to himself so he would not blab
about anything during the silly, talky time After he got on crutches he used
to take the temperatures so Luz would not have to get up from the bed As he
walked back along the halls he thought of Luz in his bed.
(Hemingway 1965: 141142),
Here is Scholes first transposition, with I replacing the occurrences of the he
third person:
I went under the ansthetic holding tight on to myself so I would not blab
about anything during the silly, talky time After I got on crutches I used to
take the temperatures so Luz would not have to get up from the bed As I
walked back along the halls I thought of Luz in my bed.
520
Jacob L. Mey
And here is the same text with I substituted for the Luz third person:
to take the temperatures so I would not have to get up from the bed As he
walked back along the halls he thought of me in his bed.
Clearly, the first transposition makes sense, and in fact has the same overall reader
effect as does the original story. The occurrences of the third person he keep their
true ring, even when transposed to the first person; they continue to be attributed to
the narrators point of view. The omniscient narrator is telling us what happens
with the hero; likewise, the hero is able to tell us what is happening directly, the
discourse emanating from himself rather than from the narrator. In the same vein,
his thoughts are accessible to the omniscient narrator; there is no clash between the
voice directly representing he, his, and him, and the voice of the narrator telling us the same things in narrative mode.
The second transposition, though, is very odd and voice clashes abound. The
overall point of view here is that of the I person, the woman called Luz but she
is not the omniscient narrator, so how could she possibly know that the male hero
thinks of her in his bed (which is in reality hers)? Similarly for the heros reservations about the blabbing: unless he has told her this (and unless we are informed
that he has done so!), she has no legal narrative access to this knowledge.
Clashes like these, where the voices of the characters and that of the narrative
attributer or voice provider apparently contradict one another, have their origins in the different points of view that the voices represent: clashing vocalizations
are due to incompatible focalizations. While in the first rewriting, the original FID
stands uncontradicted as the male protagonists discourse, a voice clash happens
when we introduce other points of view, and still want to keep alive the pretense of
an FID (as it is done in the second rewriting). In the first rewriting, the focalizer
and the voice are one and the same, just like in the original; in the second rewriting,
the voices are at war.
In our example, the clashes between the different voices were made visible by
the use of a syntactic device: a pronoun which is being switched around with respect to person reference. Clashes of this kind are, however, not bound to a particular syntactic device such as a referential pronoun; among others, the techniques of
changing tenses and moods are frequently used as indicating a change in point of
view.
A clash between different points of view occurs whenever the changes are not
made in accordance with the readers normal expectations. In this case, a shifted
point of view is usually rendered in a different tense and mood than what had been
used in the original discourse. This phenomenon is often called back-shifting: in
FID, a present tense in direct discourse is shifted back into the past, whereas an indicative mood becomes a subjunctive.
521
Compare the following (constructed) examples (due to Bal 1985: 141):

Direct Discourse:
[Elizabeth said] I refuse to go on living like this. (quotation marks; present
tense, indicative)
Free Indirect Discourse:
Elizabeth would not go on living like this. (no quotation marks; past tense, subjunctive)
In the latter case, the focalizing point is not Elizabeths (who is quoted in the first
example as speaking the words herself, using her own voice); rather, the voice is
that of the narrator (the narrative instance), crawling inside Elizabeths head, so
to say, and rendering her monologue as one who was privy to the characters
speech and thought. The words, they are the words of Elizabeth; but the voice is
that of the implicit narrator: the narrator acts as a Biblical impostor, taking on the
identity of the character.
As to the term back-shifting itself, this is somewhat misleading. Actually, the
point of view that is responsible for the FID may shift not only backwards in time
and mood, but also in a forward direction. Compare the following stretch of narrative (also due to Jane Austen):
[Catherine] travelled on for about eleven hours without accident or alarm,
and between six and seven oclock in the evening found herself entering Fullerton.
A heroine returning, at the close of her career, to her native village, in all the
triumph of recovered reputation, and all the dignity of a countess, is an event
on which the pen of the contriver may well delight to dwell; it gives credit to
every conclusion, and the author must share in the glory she so liberally bestows. But my affair is widely different: I bring back my heroine to her home in
solitude and disgrace; and no sweet elation of spirits can lead me into minuteness. A heroine in a hack post-chaise is such a blow upon sentiment as no attempt at grandeur or pathos can withstand. Swiftly, therefore, shall her postboy drive through the village, amid the gaze of Sunday groups; and speedy
shall be her descent from it.
Jane Austen, Northanger Abbey, Chawton edition, pp. 190191.
In this extract, the shifting points of view are clearly marked by the shifts in tense:
first, we have a shift from the narrative past tense ([Catherine] travelled on) to
some gnostic authorial comments offered in present tenses (A heroine returning is an event, and so on). Subsequently, the narrative is extended, in future
tense, to embody the authors own voice, when she is describing to her readers
what she is going to do to her heroine: swiftly shall her post-boy drive speedy
shall be her descent .
522
Jacob L. Mey
Admitting that such direct author interventions in the text are the exception in
modern literary prose (John Fowles being one, fairly recent exception; earlier authors such as Anthony Trollope literally reveled in it), we have to acknowledge
Austens skill in avoiding a clash of voices by making her presence on the scene
clear, among other things, by using a different tense. Readers usually have no beef
with this: if omniscience is the narrators privilege, then omnipresence (albeit in
differing tense garb) may well be the authors. But in order to be properly perceived, the textual roles and their voices have to remain clearly separated, not least
with the help of the linguistic devices that are at the authors and readers disposal.
The next section will discuss how meaning enters into the picture when it
comes to properly distinguishing between voices, and avoiding their clashes.
5.3.
Voicing and reality
Voice clashes often have a semantic background, as when an actor pretends to command a knowledge that is not her or his at the time of speaking: we witness the
author sneaking onto the scene, as it were, without any forewarning or contextual
restrictions. In cases like these, the rules of discourse organization and voice attribution are not observed: what we get is rampant omniscience and an improbable
authorial omnipresence. Often, the author omits to cloak her- or himself in the narrator role when describing events and sentiments that the actors cannot be familiar
with; since the readers depend on the narrator for their being properly introduced to
the specific parts of the literary scene where the narratives events take place, respectively the characters thoughts and sentiments are voiced, their entry into what
is called the fictional space becomes blocked or hindered. Alternatively, the narrator loses control of the narration, and misconstrues the state of the world-atlarge, as it is reflected inside the narrative; anachronisms and erroneous place attributions (along with other malapropisms) are then apt to occur, all resulting in potential voice clashes.
When this happens, the person to blame is ultimately the author who does not
respect the limitations that the fictional space imposes on the narrator and the addressees. Here is an example from a modern novel; the narrative is set in the year
1943, and the female protagonist is enrolled in flight training at a Home Guard base
somewhere in Texas.
As the summer deepened, the heat thickened. The only time she was ever
comfortable was aloft, where the wind and altitude provided natural air-conditioning.
Marge Piercy, Gone to soldiers. (New York: Ballantine Books, 1987, p. 302)
The voice heard here is that of the main character, a young airwoman volunteer;
however, there is another voice, clashing with hers, that of the author who (without
announcing herself or properly introducing her voice) comments on the heat in
523
terms that belong to a much later period. Air conditioners, as we know them today,
were not part of peoples natural reference points in the forties; the most one could
imagine in the line of space cooling technology were devices based on evaporation,
often in combination with a fan or stuck outside of a car window, blowing humid
air into the overheated space and lowering the temperatures by letting the excess
heat be absorbed by the evaporating water. Thus, air conditioners were not natural
comparison points for people in the early forties, and it is for this reason that the
protagonists voice in this passage becomes entangled in a reality clash (usually referred to as anachronism).5
Or, to take another example from the same book (the speaker is an American
soldier commenting on various places where he has been assigned):
They say Switzerland is the same scene on a cleaner scale. But you have to
rank higher than I do and have more top honchos for your friends to pull that
service.
(Piercy 1987: 182)
Here, the voice clash is blatant: the soldier, who is speaking in 1943, uses a word
that didnt come into use until well after the war was over: honcho.6 The characters voice reveals itself as containing a foreign element; the author (speaking in
her own voice-that-knows) intrudes on the narrators voice by putting inappropriate words into the mouth of the character (letting him speak with a voice-that-cannot-know). Its a bit like watching a thirties movie in which you suddenly spot a
vehicle having flashing turn indicators instead of the traditional, semaphore-type
directionals of the earlier period; or spotting a wrong period animal, such as a New
World monkey hanging by its tail, displayed in a presumed painting by Hubert van
Eyck, when all van Eyck could have possibly known in his days were Old World,
tailless monkeys.7
Other instances of this type of reality clash occur characteristically when authors impute knowledge, feelings, or wants to their characters that either are impossible to verify or that could not be imagined in the actual lives of those authors. To
take an example: in Susan Sontags novel, The Volcano Lover (1992: 182), a scene
depicts the deposed Queen of France, Marie-Antoinette, on the scaffold, waiting to
be executed, and feeling the guillotines yoke chafing around her neck. The narrative cannot be attributed to the queen herself, who is soon to die, and whose voice
is only heard internally and has left no record. The voice we do hear through the
implicit narrator is that of the queens sister, Charlotte, Queen of Naples, who
(across the waters) conjures up the horrid scene with the cart, the mob, the steps up
to the scaffold, even the executioners muddy boots and the creaking of the gruesome machinery as the axe starts to fall. Here, the narrator vicariously invests the
story with the feelings, observations, emotions, and horrors of the French queens
sister, as well as providing the reader with a self-styled internal monologue in
Marie-Antoinettes voice.
524
Jacob L. Mey
In this case, too, we are confronted with a shift in point of view that only can be
attributed to the twentieth-century authors invisible appearance, in the guise of
her implicit narrator, on the late eighteenth-century scene. Where earlier authors,
such as Henry Fielding, would think nothing of entering the narrative scene in person, to openly deliver their narrative homilies to the reader (a tradition that still was
alive in the days of Anthony Trollope, and even has been occasionally used by
modern authors), in the scene depicted above the author remains unseen, while
speaking through a variety of inconsistent mouthpieces. Apparently, however,
neither readers nor critics seem object to this kind of vicarious FID being
smuggled into the narrative.8
Contrary to the free indirect discourse that we find in texts like the Jane Austen
passages quoted above a discourse which is duly attributed to one of the characters (usually the hero/ine of the story), in the case of the quotes above, we have a
reality clash (either an involuntary one, as in the Sontag case, or one exploited for
a particular purpose, as it is done by Robertson Davies). In either case, the semantics of the story are not borne out by the reality of the speaking voices. In fact, we
are getting close here to the pragmatic aspect of voice clashing; I will deal with this
in the next section.
6.
The pragmatics of voicing
6.1.
Positioning the voices
Pragmatic voice clashes are bound up with the notion of narrative level; by this, I
understand the relative spatial positioning of the voices within the text. A voice
may be speaking from within the text, as in the case of the narrators or an actors
voice; or it can be heard as speaking from the outside, as it were, on an extra-textual metalevel. In the latter case, we may be listening to what the author or another
text-external authority has to tell us: their voices may emanate from an authoritative source, as it is often done in newspaper reporting, or they may be attributed
to a scholarly authority (propped up by the proper bibliographic references), and
so on.
Voice positioning is important for a proper understanding of the text; but also,
the proper attribution of a voice, no less than its proper use, is part and parcel of the
pragmatics of voicing. Referring to an event is not just a matter of naming place
and date; it requires an understanding of why this event is important for the interlocutors, and how a mention of the event should be voiced. As a clear case of the
need for proper voicing, consider the way ones condolences are expressed when
we meet with a friend or acquaintance who has gone through the painful personal
experience of losing a loved one. Here, correct voicing is so important that most
people will not take any chances, but prefer to use some pre-set formula, like
525
Sorry to hear about your loss, or My prayers and thoughts are with you (the
latter kind is usually restricted to written use).
In the following sections, I will discuss the pragmatics of voicing from a
number of angles, including that of territory of information and empathy. I will
conclude by offering some thoughts on the linguistic realization of voice positioning, and on some of the clashes that may occur when proper positioning is not observed.
6.2.
Territory of information and empathy
The late Japanese linguist Akio Kamio has, in a number of articles and books (e.g.
1994, 1995, 1997), offered a new perspective on the problem of where to properly
position a texts voice. His concept of territory of information implies that we all
have access to some information in ways that others normally do not; our territory
is so to speak sacrosanct in the sense that it cannot be intruded into or operated
upon unless the intruder, the interlocutor, has been properly ratified with regard
to his or her rights and duties.
In certain cases, the territory is defined so personally that we cannot even refer
to phenomena occurring within it without committing a linguistic or social gaffe.
In Japanese, for instance, bodily phenomena, such as being hungry or thirsty, belong to the territory defined by the speaking I. Only the speaker can pronounce
I about him- or herself, as Karl Bhler (1982: 103) remarked a long time ago. But
also, only the speaking I has access to his or her own feelings and sensory experiences; thus, saying You are cold to a person in Japanese is not proper voicing,
since the position of the speaker is outside of the territory where such information
is found. The interlocutor can only say something like You look cold to me, or
You seem to be cold; and similarly for other expressions having to do with the
ego whose territory we are dealing with (like being hungry, having a problem,
and so on). Here is another example, borrowed from Kamio (1995):
In Japanese, it is usually allowable (and hence pragmatically correct) to be
asked, and answer a question, about ones wifes age.9 But while it is OK for a husband to utter
Kanai wa 46 desu My wife is 46
(a husband is supposed to be familiar with the number of years that have
elapsed since his wifes entry into this world), the same husband cannot properly put himself outside this territory of information by adding a quotative
maker (in Japanese -tte or soo) to the utterance. Hence both
Kanai-wa 46 desu-tte I hear my wife is 46
and
Kanai-wa 46 da-soo desu My wife supposedly is 46
526
Jacob L. Mey
are pragmatically ill-formed, as they do not reflect the proper positioning of the
husbands voice, viz., inside his and his wifes shared territory of information.
Clearly, the fact that there are limits to ones territory of information is not particular to Japanese or any other language; but the specific ways in which the territorys boundaries are expressed linguistically are often strictly bound to the respective territorialities of cultures and languages. Something similar is valid in the
cases discussed by Heritage under the general labels of accessibility and the
right to empathy (Heritage 2007). Congratulations, condolences, and other expressions of empathy should not be offered out of place: a faulty positioning of the
speaking voice may result in a fatal clash. A student cannot properly congratulate
his or her professor on the latters brilliant lecturing (see Mey 2008 for a case in
point). Similarly, as Heritage and Sefi remark, advice giving carries problematic
implications about the knowledge or competence of the intended recipient (1992:
389); anybody who has witnessed or experienced cases of spontaneously alter-offered, well-intentioned but ill-received advice knows how toe-curlingly irritating
such information-territorial transgressions can be. To enter or not to enter another
persons territory is not just a matter of fact; it has eminently to do with the pragmatics of voicing, and in particular with the proper positioning of voices within the
proper territory.
The linguistic mechanisms that are at our disposal to effect correct voice positioning and avoid voice clashes that result from undue changes of position or unwarranted infraction of territories are numerous. But more often than not they are
insufficiently understood and clumsily managed by authors and readers; as a result,
a clash of voices may occur. In the following section, I will consider one of the
most widely employed mechanisms that are instrumental for a correct pragmatic
positioning, to wit: the discourse markers often referred to as sentence adverbials, and discuss how they function pragmatically.
6.3.
Sentence adverbials
Among the most frequent mechanisms used for positioning the discoursal voices
are the so-called sentence adverbs (often called sentence adverbials, since they
not strictly have to be what are usually called adverbs, but may be made up of
longer stretches, phrases, and even entire sentences functioning adverbially, that
is to say, specifying a verb or verbal expression). Such adverbial expressions have
as their scope not just the (immediately) preceding or following verb (as in I admire her greatly), but may extend over a whole phrase or sentence. Their name,
sentence adverbials, expresses this fact: they act as ad-verbs, added not just to a
single adjective or verb, but potentially to whole phrases or sentences. In this way,
they are often equivalent to full-blown adverbial clauses; for instance, in an utterance such as To be honest with you, your article is not good enough for our journal, I could replace the entire adverbial clause To be honest with you by a single
527
adverb or adverbial phrase: Honestly/In all honesty, your article is not good
enough for our journal.10
Sentence adverbials have the property of coloring or characterizing the entire
sentence they scope over. Thus, they may be used to express the utterers or
writers viewpoint on what the sentence says; they serve to position the speakers
or writers voice.
Consider the adverb happily in the following sentence (taken from a letter addressed by a prospective author to a journal editor): The paper is formatted in MS
Word; I will happily send you a diskette if necessary. In the context, this tells the
editor that the sender of the message is willing to provide a diskette with her paper,
and moreover that she will be happy to do so.
Alternatively, a sentence adverbial may be used to tell the addressee where the
message in question is to be placed in the ongoing discourse; again, the senders
voice is positioned properly thanks to the adverbial. An adverb such as finally is
often used to tell us that what follows is the last element in a series of arguments,
examples, and so on. In this way, finally is equivalent to a meta-statement by the
author about his or her text: Im telling you now that the next argument or example
is the last one. And consequently, the sentence adverbial may transmit the additional message that there is no more to talk about, from the speakers point of view.
Of course, such meta-statements may remain implicit in the text, or be expressed in other ways, as is often the case in highly ritualized text productions,
such as sermons, political speeches, or newscasts. A particular inflection of the
voice may accompany a summing-up statement (For National Public Radio, this is
Melissa Brown); in such cases we do not need a finalizing adverbial phrase of
the type at the end of the day, or when all is said and done. Alternatively, I may
make use of the multi-purpose adverb anyway to conclude a discussion by indicating my position as the (purportedly) final contributor to the debate, or to voice a
need to return to the original argument the same argument that was expressed in
full sentential array by the judge in the medieval farce of Matre Pathelin, who
coined the famous expression Revenons nos moutons (literally; lets go back
to our sheep), meaning: Anyway, enough of these digressions, lets get back to
the subject.
Sentence adverbials are extremely useful devices, not only for positioning the
voices of the discourse, but also for assuming or denying speakers responsibility
for an utterance. Thus, a sentence adverbial like realistically speaking may indicate a speakers attitude to what is being said; hopefully characterizes the speakers wishful expectations as to the success of the action announced in an utterance;
and so on and so forth. In the language of the press and the courts, the adverbial allegedly is universally used whenever there is talk of a person suspected of being
involved in a criminal act, but without the speaker wanting to make a formal accusation. The sentence adverb allegedly tells us that what we are going to hear does
not necessarily reflect the speakers beliefs or intentions, but something he or she
528
Jacob L. Mey
has heard somebody else say: the speaking voice is located at a safe distance from
the (recording or reflecting) alleging voice, which is only properly heard through
the mouthpiece of the speaker. Thus, on the level of the informative speech act of
assertion, nothing is said that could incriminate, or create a legal liability for, the
speaker person: the catch-all sentence adverbial allegedly acts as a disclaimer for
the whole utterance.
Consider the following constructed example, in which a newspaper reporter
quotes what transpired from the courtroom proceedings in the case of an armed
robbery:
Allegedly, the defendant entered the 7&11 convenience store on South Congress at around 10 pm, demanded to be handed over the money, and when the
store owner declined, proceeded to hit him over the head with an 11 in. pipe
wrench.
Here, the speaker implicitly states that he is not responsible for the truth of the allegations: allegedly is spoken with a borrowed voice; that is to say, somebody
(for whom the speaker is not responsible) is making he statement at his or her own
risk. The responsibility for the truth of the utterance found within the scope of the
sentence adverbial is lifted off the shoulders of the speaker; the expression allegedly marks the whole reported phrase (the defendant entered the 7&11 convenience store and so on) as something that is merely alleged, and therefore not
necessarily true as reported.
In other words, the sentence adverbial in question functions as a disclaimer.
Whatever is within its scope, is said to be outside the verbal jurisdiction of the
speaker, outside his or her territory of information; consequently, the speaker cannot be legally held responsible for the effects of his or her words.
Non-adverbial expressions may serve a similar distancing function. As
examples, consider the words Watch my lips, famously uttered by the first US
President Bush (George H. W.), when questioned about his promise not to raise
taxes, or the fully-fledged disclaimers of the type I didnt mean to say that. This
latter type can often result in voice clashes commonly known as paradoxes: a celebrated case is that of the so-called Moores paradox, by which one denies what
one just has uttered (classic example: The cat is on the mat but I dont believe it,
attributed to the philosopher G.E. Moore; Mey 1993: 156).
Since sentence adverbials are so unobtrusive and impersonal, as contrasted
with explicit, full sentence disclaimers, their voices are positioned in neutral territory, so to speak, while their anonymity prescribes the correct interpretation of the
words uttered even in the absence of a concrete utterer: a speaker or writer. We are
in a borderland of voices here: on the one hand, we have sentence adverbials that
straightforwardly represent the attitude of the speaker like in Strictly speaking,
your request is out of order, where the voice of the speaker is heard as strictly applying certain rules and regulations. On the other, we have cases where the speaker
529
distances him- or herself from what s/he is going to say, as in Strictly off the record, I think youre indulging in a frivolous interpretation of the law. Here, the
speaker wishes to be quoted off the record, that is, the words spoken are not to be
recorded, that is to say, made part of some official account. But the wish to not say
something is not equivalent to not saying that very thing; when I say I dont mean
to insult you, but I think youre not behaving like a gentleman, I cannot countermand the insult, as the words have been spoken by what is rightly assumed to be
my voice, even though that same voice now tries to distance itself from what it just
has said, or is going to say. The speaker dons a speaking mask, so to speak, a persona in the classical sense of the term (like the masks worn on the Greek and
Roman dramatic scenes): speaking on the authority of this voice (and possibly employing what is currently on the verge of becoming a standard disclaimer, viz.,
speaking on condition of anonymity), the voice of the speaker assigns the voice
of the alleged or quoted utterance its correct positioning. We are confronted here
with a change of voice that we may experience as normal and legitimate, but also
on occasion as representing a clash.
The question may be raised whose voices we really are hearing in cases like the
above. Most hearers and readers will not usually be in doubt in the normal course
of events: it does not take too much reflection on the part of the addressee to correctly position the voices he or she is hearing and attribute them to the proper characters in the narrative or discourse. But there is of course some rhetorical trickery
involved in this process. By forcing the addressee to reflect on the positioning of
the narrative voices and on the characters mutual relationships, the speaker may
gain time and improve his or her own position: while the addressees defenses are
down, the speaker is able to get another thrust in edgewise. The clashing voices of
the narrative recall the clanging swords of the classic duel, and it often takes an uncommon skill on the part of the attacked (s)wordsman to position himself correctly
in the face of the attack.
In less martial contexts, the clashing voices may limit themselves to spreading
confusion, as in the following excerpt from a contemporary mystery story, where
Viera, the protagonist, suddenly realizes that the character she had trusted, in reality is the main villain. The development of the narrative, as told by the narrator (the
story is in the homodiegetic, that is, the I mode, as opposed to the heterodiegetic, or he/she mode) leads us to Vieras discovery, and thereby opens up a new
vision on reality for the storys protagonist, causing her to reshuffle her voice
cards.
But what about Vieras own voice: where does it position itself? In particular,
what does the sentence adverbial subconsciously in the following fragment tell
us about this? (Viera is conducting an internal monologue):
Subconsciously, I realized
(Elizabeth Powers, All that glitters. New York: Avon Books, 1981, p. 43).
530
Jacob L. Mey
Vieras internalizing the changed state of affairs, the new reality, may aptly be described by the verb to realize, even when uttered by the narrative voice (recall
what I said earlier about the omniscient narrators position in the story); but how
are we going to deal with the sentence adverbial subconsciously? If the voice we
are hearing is the voice of the character Viera, then the utterance subconsciously
in her monologue is not appropriately attributed to this voice. A person cannot realize something without being conscious of it; hence Viera could not properly utter
the sentence fragment in question without provoking a clash of voices and an ensuing confusion in the readers mind.
This is a pragmatic clash, and we can only solve the dilemma by appealing to a
super-voice that is not heard directly in the dialogue, but is supposed to be implicitly present: this the hidden voice of the omniscient author, inserting itself into
the discourse of the character, where it subsumes and suppresses the clashing
voices of the monologue. We lend credence to this maneuver because we trust the
author; but as readers, we experience a certain uneasiness. This uneasiness may be
formulated in the following simple question: How could Viera realize what she
subconsciously was not even thinking, let alone expressing in words? It is as if the
author were telling the character what its subconscious thoughts would sound like,
when brought to the surface (where they of course would provoke a clash). By
keeping it under wraps, so to speak, the author may avoid the clash, but not eliminate the readers uneasiness, caused by the narrators attempt to pull the literary
wool over our eyes, as she tells us what to believe about the character Vieras unexpressed state of consciousness rather a tall order, even for an omniscient narrative instance.
7.
Conclusion: narration and conversation
From a pragmatic point of view, both the literary text and the text generated in conversation are manifestations of human language use. It may be useful to briefly
compare the two uses in this concluding section, to see if and how they are different, and if any observed differences can shed light on the use of the literary voices
that we have been dealing with so far.,
If one asks how language use differs from narration to conversation, one notices that the ways we use language in the two cases are rather diverse. Narration
may include conversation, and conversely, conversations usually include stretches
of narration; yet, their respective ways of doing things with words are vastly different. Compare the meticulous wording that is the hallmark of literary prose, with
the spontaneous utterances that characterize everyday conversation: in the first
case, we have authors judiciously assigning the various voices of the text to their
respective characters (where clashing voices should not normally, but do occasionally, occur); in the second, we are confronted with what sounds like a cacophony of
531
voices, some speaking in unison, some interrupting, some dropping out, some
dominating, some subordinated, and so on.
The conversation analysts have shown us that (despite apparent irregularity)
the business of conversation is conducted according to strict rules of selection,
turn-taking, and (non-)overlapping. Beginning with the ground-breaking work of
people like Sacks, Schegloff and Jefferson (1974), a new discipline, Conversation
Analysis, has seen the light of day and is enjoying great popularity among a crosssection of the linguistics and sociology fields.
At first blush, the question Whose voice are we hearing? may seem a bit inappropriate in the conversational context. In conversation, we will always be able
to identify the current speaker (the one who has the floor, as it is called). The current speaker is the one who can be singled out by his or her use of the first singular
(or sometimes plural) personal pronoun (the speaking I, as the German psychologist Karl Bhler used to call it; 1982: 103). By the same token, the passages of a
novel in which the characters conduct conversation should be the ones that are the
least problematic with respect to the positioning of those voices.
But on closer inspection, the question of where the voices are coming from
turns out to be more complicated. As we have seen in the preceding, voices are not
always what they pretend to be, or sound like; an implicit voice may represent an
extraneous instance, speaking with the authority of a superaddressee (as Bakhtin
calls this instance), namely: God, absolute truth, the court of dispassionate human
conscience, the people, the court of history, science, and so forth (1994: 126).
Such a voice may not be as easy to detect as, say, the voice of a person using a local
or social dialect; but it may manifest itself in numerous indirect ways, e.g. by the
authority it invokes, the appeals it makes to the listeners, the value-laden judgments it proffers or reproduces, and so on.
When Jane Austen rebels against the common cant which undervalue[s] the
labour of the novelist (1947: 23), we hear the authors voice deploring the poor
appreciation allotted to novelists on the part of those whose voices belong in the
specific discourse, common to a particular type of readers: those who will pick up
an Austen novel, but hide it in shame when caught in the act of reading. Similarly,
when Virginia Woolf lets Clarissa Dalloway muse about the important meetings
her husband Richard has to attend (Armenians, he said; or perhaps it was Albanians; Woolf 1990: 119), we hear, through the speaking voice of the protagonist, the implicit background voices of the societal discourse of the time, where the
Armenian question played such an important role but not important enough for
Mrs. Dalloway to be able to keep Armenians apart from Albanians.
Similarly in everyday conversation, voices may appear on the interlocutory
scene that are not explicitly heard or observed, but stay hidden in the conversational backdrop. For instance, a person may ask an interlocutor a question that is
not appropriate in the context, as it presupposes an authority on the part of the
questioner that he or she may not have. The implicit voices we are hearing when
532
Jacob L. Mey
such questions are denied an answer are those of societys powerful instances (such
as the police or the judiciary), that pervade the entire fabric of our society. While a
policeman or -woman can stop me in the streets of an affluent neighborhood and
ask me for my identification if he or she perceives something untoward, a regular
neighborhood dweller is not allowed to do the same, except maybe in an act of
neighborhood watching (or in the US context) of making a citizens arrest. In
the latter cases, such a person would be speaking with the (implicit or impersonated) voice of the police official, ventriloquating (to use Bakhtins colorful expression) the sounds emanating from the official instance, the police superaddressee. In conversation as in narration, the proper positioning of the speakers and
their voices is every bit as important as the produced text (narrative or conversational) itself.
In the pragmatic analysis of a text (narrative or conversational), correctly placing the voices in the discourse is as crucial for the readers as it is for the interlocutors. Both have to know, at any given moment of the (literary or conversational) interaction, where they stand in relation to the other participants, and where the
protagonists of the literary production: viz., the characters (including the author
and the narrator) of the narrative, respectively the interlocutors in the conversational situation, are located in the discourse.
Pragmatics, when applied to narrative as well as conversational texts, raises a
double question: From which point of view are we being addressed as hearers and
readers; and: Whose language is being spoken, whose voices are being heard? In
the preceding pages, I have tried to make it clear that, in order to answer these (not
unrelated) questions, we have to rely on the linguistic and pragmatic concepts that I
have referred to in the preceding pages, first of all, the central notion of voice and
its (harmonious or clashing) manifestations in literary and conversational discourse.
Notes
1. The story has been told earlier in my book When Voices Clash (Mey 2000: 148150).
2. The author speaks as it were through language, a language that has somehow become
objectified, that he merely ventriloquates [sic] it (Bakhtin 1992:299).
3. These terms are taken from traditional Latin (school) grammar. Verba dicendi are, e.g.,
to say, state, affirm, exclaim, ; verba sentiendi express the thoughts or feelings of the
subject (e.g. to think, mean, believe, ); inquit is the 3rd person singular of the defective Latin verb inqueo to say, speak.
4. Of course, as Scholes also remarks, [t]he third-person narration of the original text is a
disguise, a mask of pseudo-objectivity worn by the text for its own rhetorical purposes.
(1982:117)
5. According to the Encyclopaedia Americana, air conditioners for home use began to be installed in the U.S. in the late 40s-early 50s.
533
6. The words first official mention (as far as I have been able to ascertain) is in the
Oxford American Dictionary (1980). It is said to be derived from the Japanese hancho, leader of a band; it was first introduced into English, according to Wikipedia, in
1955.
7. And in this way, the clash reveals the true date of the forged painting, The Harrowing
of Hell, as the incident is described by the Canadian novelist Robertson Davies in
the second volume of his Cornish trilogy, Whats bred in the bone. (Penguin 1985,
pp. 350351)
8. Compare the British critic Natasha Walters comments on this episode, in her review of
Sontags work: there are scenes to remember in the book: Marie-Antoinette on the
scaffold, feeling the yoke choking her (The Independent, October 2, 1992, p. 27)
9. Or ones own, for that matter. Other cultures may have other, special socio-cultural conventions of propriety; compare: Women dont have ages, as Talleyrand supposedly remarked to Metternich at the Vienna Congress.
10. Sentence adverbials are among the lesser studied pragmatic discourse operators. The
fullest treatment (by now somewhat dated) is by the Canadian-Czech linguist Eva Koktov (1986).
References
Austen, Jane
1947 [1815] Northanger Abbey (Chawton edition). London: Allan Wingate.
Bhler, Karl
1982 [1934] Sprachtheorie: Die Darstellungsfunktion der Sprache. Stuttgart and New
York: Gustav Fischer.
Bakhtin, Mikhail M.
1992
Discourse in the novel. In: Michael Holquist (ed.), 259422. [First Russian
publication 1975].
Bakhtin, Mikhail M.
1994
The problem of the text in linguistics, philology and the human sciences. In:
Caryl Emerson and Michael Holquist (eds.), 103131. [First Russian publication 1976].
Bal, Mieke
1985
Narratology: Introduction to the theory of narrative. Toronto and London:
Emerson, Caryl and Michael Holquist (eds.)
1994
Mikhail M. Bakhtin: Speech Genres and other late essays. Trans. by Vern W.
McGee. Austin, TX: University of Texas Press. [First Russian publication
1986].
Hemingway, Ernest
1965 [1925] A very short story. New York: Scribner.
Heritage, John
2007
Territories of knowledge, territories of experience: (Not so) empathic moments
in interaction. Keynote speech at the XVth Symposium About Language and
Society, Austin (SALSA), Austin, TX, April 14, 2007.
534
Jacob L. Mey
Heritage, John and Sue Sefi

1992
Dilemmas of advice. In: Paul Drew and John Heritage, Talk at work: Interaction in institutional settings, 359417. Cambridge: Cambridge University
Press.
Holquist, Michael (ed.)
1992
The dialogic imagination: Four essays by M.M. Bakhtin. Austin, TX: The University of Texas Press.
Kamio, Akio
1994
The theory of territory of information: The case of Japanese. Journal of Pragmatics 21(1): 67100.
Kamio, Akio
1995
Territory of information in English and Japanese and psychological utterances.
Journal of Pragmatics 24(3): 235264.
Kamio, Akio
1997
Theory of territory of information. Amsterdam: Benjamins.
Koktov, Eva
1986
Sentence adverbials in a functional description. Amsterdam: Benjamins.
Mey, Jacob L.
1993
Pragmatics: An introduction. Oxford: Blackwell.
Mey, Jacob L.
2000
When voices clash: A study in literary pragmatics. Berlin: Mouton de Gruyter.
Mey, Jacob L.
2008
Impeach or exorcise? Or, whats in the (common) ground? In: Istvan Kecskes
and Jacob Mey (eds.), Intention, common ground and the egocentric speakerhearer, 254275. Berlin: Mouton de Gruyter.
Parks, Edward
1994
Around the Mall and Beyond. The Smithsonian 25(8): 2025.
Piercy, Marge
1987
Gone to soldiers. New York: Ballantine Books.
Powers, Elizabeth
1981
All that glitters. New York: Avon Books.
1974
Language 50(4): 696735.
Sontag, Susan
1992
The volcano lover: A romance. New York: Farrar Straus Giroux.
Scholes, Robert E.
1982
Semiotics and interpretation. New Haven, Conn.: Yale University Press.
Storm, Theodor
1996 [1888] The Dyke Master. London, Angel Books. (German original: Der Schimmelreiter. Transl. by Dennis Jackson,).
Woolf, Virginia
1990 [1925] Mrs. Dalloway (Harvest edition). San Diego, New York and London: Harcourt Brace and Company.
Part V
Methods and tools
19.

Monika Bednarek
1.
Introduction: approaching the data of pragmatics
As the editors of this series, Wolfram Bublitz, Andreas Jucker and Klaus Schneider
argue pragmatics is a transdisciplinary subject which has a substantial and
wide-ranging significance for nearly all areas of language description. From the
broad conceptualization of pragmatics that they envisage, pragmatics includes approaches that are traditional and contemporary, linguistic and philosophical, social
and cultural, text- and context-based, as well as diachronic and synchronic. In this
view, pragmatics shares interests with other related and not always easily differentiable linguistic sub-disciplines, such as discourse studies, semiotics, psychoand sociolinguistics, conversation analysis, or ethnography of speaking (see
Wodak 2008: 34 for a list of seven dimension these fields have in common), sharing for instance the focus on language use or linguistic (inter)action. There is thus a
common commitment of these linguistic sub-disciplines to language in use the
pragmatics of language.
It seems reasonable to propose that any discipline interested in language use
must have some kind of approach to its object of investigation the data of linguistic (inter)action, or simply, the data of pragmatics. This includes a methodological standpoint from which to approach this data, some kind of research
methodology. However,
the commitment to particular methods or procedures often fails to deal with a range of
important questions: where or from whom should the data be collected and how? (For
example, shall we use texts that we obtain in interviews or published self-descriptions?
Do we include the news on the notice board and the graffiti in the toilets? Are texts sufficient for our purposes or should we also carry out observations?) Equally undecided is
the matter of how the data collection is to be organized. (For example, do we generate
our texts in group interviews or individual interviews? In what order shall we do the interviews?) The question of how the material is to be stored also remains open. (For
example, will audio recordings be sufficient or do we need video recordings? What rules
of transcription shall we apply?) (Titscher et al 2000: 1314, see e.g. Kreuz and Riordan
in this volume on transcription systems).
Thus, while the focus of this chapter is on the use of different data in pragmatics
and related disciplines (compare first paragraph above) rather than a general discussion about research methodology, epistemology, or evaluation criteria such as
ermk 2001; Silverman 2004;
reliability, validity, replicability (see Taylor 2001; C
McNeill and Chapman 2005; Drnyei 2007), reference will be made to the insights
gained from different approaches and methodologies to data and the different per-
538
Monika Bednarek
spectives they provide on the linguistic data. The starting point of this chapter is
thus the differentiation between different kinds of data, rather than between different kinds of methodologies (e.g. quantitative vs. qualitative).
2.
Issues of data collection and analysis
Issues that have to be addressed by all researchers using data are whether all data
should be systematically collected before the analysis or whether further data can
and should be added after some data has already been analyzed. This question relates to the design of data collections or corpora (see e.g. Wynne 2005, Voormann
and Gut 2008). Another issue is the difference between inductive (data-driven) and
deductive (theory-driven) research. In other words, do we analyze linguistic data
according to some pragmatic theory such as politeness (Leech 1983, Brown and
Levinson 1987) or do we deduce some pragmatic theory from the analysis of the
linguistic data itself? This kind of distinction cuts across the use of different data,
and different methodologies, such as quantitative vs. qualitative research. For instance, speaking of large-scale corpus linguistics, Tognini-Bonelli (2001) makes a
distinction between corpus-driven and corpus-based linguistics. The term corpusbased is employed to refer to a methodology that avails itself of the corpus mainly
to expound, test or exemplify theories and descriptions that were formulated before
large corpora became available to inform language study (Tognini-Bonelli 2001:
65), and the term corpus-driven refers to an approach where the commitment of
the linguist is to the integrity of the data as a whole, and descriptions aim to be
comprehensive with respect to corpus evidence (2001: 84). Similarly, Bednarek
(2006a) speaks of a text-driven methodology, where texts are analyzed trying to
make as few a priori theoretical assumptions as possible before analyzing the text
data, and a text-based methodology, where texts are analyzed using a previously
established theoretical framework. In contrast to corpus-based/corpus-driven research, text-based/text-driven research is based on the manual (and more qualitative) analysis of small-scale text corpora rather than on automated quantitative
large-scale corpus analyses (Bednarek 2006a). However, what Halliday notes with
respect to corpus-based and corpus-driven research is also true for text-based and
text-driven analysis: the two approaches represent a continuum rather than a strict
dichotomy, and much research is situated somewhere in between (Halliday 2004:
24). This may also be the case more generally with respect to inductive vs. deductive approaches to data analysis in pragmatics.
Other issues that need to be addressed by all researchers include legal, ethical
and technical considerations, with many research proposals and projects needing
the approval of ethics committees at universities, and much data collection and
publication requiring sometimes costly copyright consent of respective copyright
owners. Most spoken data also require consent of speakers to use and publish these
539
data. Relevant technical issues concern questions such as: what technology is
needed to collect the data (e.g. audio and video recordings, keying in of data, conversion of internet data into other formats), to store the data (e.g. large databases,
large electronic storage facilities) and to analyze the data (e.g. computer software)?
Such issues can place real constraints on the analysis of linguistic data and should
in general be considered before data collection.
3.
Types of data
3.1.
Non-attested data
I use the term non-attested data to refer to any kind of data that has not been observed or recorded as having been produced by a particular speaker or writer. This
covers predominantly the use of introspection thinking and reasoning about language use based on ones own knowledge and experience of language. The term
non-attested does not, however, refer to recourse to anecdotal evidence data
that has been observed by the researcher at a particular time without necessarily
having information on when it was observed, by whom it was produced, what the
context of use was etc. This kind of attested but anecdotal data can sometimes be
useful, even if systematic data collection is to be preferred. For example, whereas
Bublitz (1998) uses corpus evidence to argue that totally has a negative semantic
prosody (Louw 1993), Bednarek (2008c) uses informal observations of the use of
totally on broadcasts on MTV to argue that, in AE young adult language, totally
does not exhibit such a prosody.
Much classic research in pragmatics seems to rely on introspection, particularly that related to logics and language philosophy (Austin 1975, Searle 1969,
Grice 1989, Kompa and Meggle in this volume). Adolphs (2008) goes as far as
stating that most research in pragmatics has not used attested examples of language-in-use (Adolphs 2008: 7). One example is speech act theory, although more
recent approaches in pragmatics study speech acts with respect to elicited (e.g. discourse completion tasks; see below) or corpus data. For instance, Adolphs (2008)
focuses on describing and exploring lexico-grammatical patterns that are associated with speech acts. Thus, why dont you can be related to the act of suggesting
and could you just can be related to the act of requesting. She also explores the distinction between direct and indirect speech acts with the help of corpus data.
In general, introspection can be useful for forming hypotheses, but these need
to be applied to attested language data for confirmation, rejection or modification.
Indeed, many corpus linguists would argue that what we think about language use
does not always correspond to language use. The focus of introspection also seems
to be more on what is possible in language use, rather than on what is actual, typical, salient or frequent.
540
3.2.
Monika Bednarek
Attested data: elicited and non-elicited
Attested data means language data which was actually produced by speakers/
writers on particular occasions. Such data can be elicited (by the researcher) or can
have occurred more naturally, that is, without being produced in a research context.
3.2.1.
Elicited data
Under the heading of elicited data we can include interviews (involving researchers), verbal reports, elicited narratives, discourse completion tasks, and experimental data.
3.2.1.1. Interviews
Interviews are a standard way of collecting data across the social sciences (see
e.g. Miller and Glassner 2004), including linguistics. There are various kinds of
interview data. Structured interviews use prepared questions, which are often
standardized, and thereby pre-define topics, whereas unstructured interviews
leave room for topics to be explored, and are more like an informal conversation (McNeill and Chatman 2005: 56). Semi-structured interviews feature a
combination of the two. While unstructured interviews thus may be more authentic, they are also more subjective, less representative, and tend to involve a
smaller number of participants. Depending on mode, interviews can be in written
or spoken form (with written interviews shading into questionnaires; see below);
depending on participants, they can involve only one participant or more than
one. Sometimes focus groups are set up featuring discussions in which one or
more researchers moderate the discussion and interaction between members of a
group.
There are at least two ways in which interview data are used in research. Either
what people say in interviews is seen as a reflection or description of what they
actually do/think or what people say in interviews is itself analyzed as a particular kind of linguistic construal. Benwell and Stokoe (2006) write:
Unstructured interviews and focus groups as data collection methods have been the
focus of much debate and discussion across the social sciences, largely due to the way
they are often treated as a transparent description of opinion or unmediated report of
practice. [] Some commentators argue that interview data must be treated as a researcher-driven, occasioned and situated account of identity and practices. (Benwell
and Stokoe 2006: 194)
There is a point in arguing that interviews may tell us what people think but not
necessarily what they do or even that interview data are too much influenced by the
interview situation, relations between interviewer and interviewee, social values
541
etc to tell us anything about the topic of the interview. Wetherell et al (2001) ask:
when a researcher interviews someone, how far can their talk be treated as a reflection of something else, such as established memories and ready-formed
opinions? Should it instead be analysed as the place and process in which memories and opinions are constituted? (Wetherell et al 2001: 7). Group discussions in
focus groups can raise new problems of about-ness: perhaps the interaction is
now about group relations rather than the apparent topic of discussion? (Wetherell
et al 2001: 28).
It may thus be more useful to use interview data to explore not what participants think about or how they handle the ostensible topic of the interview, but to
focus the analysis on other issues of the linguistic data. For example, Galasinski
(2004) makes use of interviews with 26 male and 29 female participants, where the
interview topic was a variety of issues related to middle age [], the experience
of its onset, and its relation to family and working relationships (Galasinski 2004:
24). He then explores the data with respect to emotion and gender, looking in particular at how emotion narratives (Galasinski 2004: 24) work in the construal of
masculinity. In particular, participants can choose to distance themselves from
emotions using a variety of distancing strategies, such as not directly describing
themselves as experiencing a particular emotion. This is the case in this interview
extract where a situation is described as being traumatic/dramatic rather than as directly and explicitly referencing the speakers emotion (as in the alternative I really
hated it or I was really sad):
(1)
I: and can you describe what it was like when your daughter moved out?
IF: well (.) permanently (.) it sort of crept up on us. I mean the night we left her
at university (.) she didnt go in halls, she went into a students house. And that
night was pretty traumatic, the drive back down the M6 was pretty dramatic.
[]
(Galasinski 2004: 53, my bold)
Such distancing strategies can be used by male speakers to maintain the pretence
of meeting the requirements of the dominant model of masculinity (Galasinski
2004: 73). Such use of data thus includes considerations of context and dominant
ideologies (e.g. prevailing models of gender), and seems a more interesting way of
looking into emotion narratives than interviews where participants are asked about
what they think about emotion and gender or where men and women are asked
about their emotional experience.
In general, interviews have a number of methodological advantages, as they are
argued to be
a natural and socially acceptable way of collecting information that most people feel
comfortable with and which can be used in a variety of situations and focusing on di-
542
Monika Bednarek
verse topics to yield in-depth data. The interviewers presence allows for flexible approaches, probing into any emerging new issue, while the interview guide helps to
maintain a systematic coverage of the domain. (Drnyei 2007: 143)
However, interview data are clearly not naturally occurring; further, to set up and
run interviews and to analyze the resulting data can be challenging and time-consuming; and there are also questions concerning the nature of the data with respect
to the interview situation (as discussed above). As already noted, it is problematic
to consider interview data as reflecting reality or producing reliable stories of experience; rather, interviews concern the repetition of familiar cultural tales (Miller and Glassner 2004: 125) and the interview is obviously and exclusively an interaction between the interviewer and interview subject in which both participants
create and construct narrative versions of the social world (Miller and Glassner
2004: 125).
3.2.1.2. Verbal reports
Verbal reports are self-reflections on participants language use, and are primarily
used in second language studies, interlanguage studies and cross-cultural research
(Woodfield 2007: 2). Verbal reports are employed by researchers to reconstruct
cognitive processes in language use, and to find out about speakers pragmatic
knowledge and their choice and formulation of speech act strategies (Woodfield
2007: 2). Verbal reports can be concurrent (on-line) e.g. as think aloud, or they
can be retrospective (subsequent). However, it is necessary to point out that when
interviewing participants after interactions, participants might not themselves be
able to recollect their own interpretations at a later stage (Adolphs 2008: 41), and
that trying to access cognitive processes is problematic because such processes
may be unconscious and thus not amenable to self-reporting. As with interviews, it
is worth problematizing whether or not verbal reports can be a reflection of actual behavior, cognitive processes and knowledge structures. Further, the influence
of the researcher and fellow informants on the resulting verbal report cannot be neglected (Woodfield 2007, citing Haastrup 1987).
3.2.1.3. Discourse completion tasks
Much cross-linguistic research in pragmatics (e.g. the Cross-Cultural Speech Act
Realization Project) or research in Second Language Acquisition into speech acts
uses oral or written discourse completion tasks (DCTs). DCTs were first introduced
by Levenston (1975) for testing communicative competence and then employed by
Blum-Kulka (1982) for speech acts (Rose and Ono 1995: 192). Written DCTs are
questionnaires that involve a description of different situations, potentially some
dialogue, and empty slots for subjects to produce a particular speech act or to fill in
appropriate responses (e.g. to a compliment). For example, to elicit English re-
543
quests, the following description may be used: You have been asked by your lecturer to give a seminar paper but you need more time. You go to see her to ask for
extra time to prepare your paper. Ask for an extension (from Woodfield 2007, who
also gives an overview of different request situations used in research). Participants in DCT resist such tasks to some extent, commenting on the non-interactive
and inauthentic nature of DCTs and their lack of contextual data (Woodfield 2007).
Although only limited efforts have been made to evaluate the validity of this
method (Yuan 2001: 272), it has been suggested that both written and oral DCTs
have a number of limitations in that they do not capture dynamic discourse features
such as turn-taking, conversational structure, speaker-listener interaction, and
pragmatic features (Kasper 2000, cited in Woodfield 2007; see also Yuan 2001).
Thus, responses given in DCTs are shorter in length, simpler in wording, less
face-attentive and negotiatory in the negotiation process, and less emotionally involved than naturally-occurring speech (Yuan 2001: 272). They may capture
pragmatic knowledge rather than usage. In as far as utterances in DCTs are more
decotextualized, more decontextualized and less multimodal than natural discourse, they can be argued to neglect the complexity of language usage. Adolphs
(2008) notes that the fact that data generated through DCTs is based around single
utterances distorts the overall picture of a speech act which is often negotiated and
developed over a number of turns in a dynamic discourse event. It is therefore unclear whether DCT or role play data display the same patterns and attributes as naturally occurring data (Adolphs 2008: 22).1
A further disadvantage of DCTs is that they are difficult to construct, and there
is a need for rigor in questionnaire construction and for extensive pilot-testing of
DCTs to control for various possible effects for prompt, rejoinder, and so forth
(Rose and Ono 1995: 198). As with other questionnaires, general issues that need
to be considered concern the sampling of the population of the participants (on
sampling see e.g. Drnyei 2007: 95100),2 its design (e.g. formulation of situation
in terms of appropriate detail, multimodality) and administering the DCT.
On the other hand, DCTs, like other questionnaires, allow for control and quantification; they are replicable; they are not time-consuming; and they are versatile.
For a more detailed discussion of DCTs in pragmatics see Woodfield (2007), who
proposes that DCTs represent highly constrained instruments of data collection in
terms of the degree to which the data is predetermined by the elicitation instrument (Woodfield 2007: 5). With respect to the distinction between oral and
written DCTs, the former share more features of natural speech than the latter
(Yuan 2001), and thus may be preferable. Yuan concludes that an oral DCT
generates data that are, to a certain extent, representative of natural speech; it allows researcher control; it is efficient in terms of the amount of data it can elicit: and it is ethical. Nevertheless, [] the oral DCT suffers similar drawbacks as do the written DCT
[sic] in that it cannot elicit elaborated negotiations and indirect compliment exchanges
seen in everyday conversations. (Yuan 2001: 289)
544
Monika Bednarek
3.2.1.4. Experimental data

Experimental data in pragmatics and related areas are used in educational research
(see Drnyei 2007), with respect to child language acquisition as well as in cognitive linguistics. For example, experimental research on understanding and recall
asks subjects to reproduce sentences with the help of specific cues. Such research
has suggested that understanding may proceed with the help of mental models
(frames, schemas etc, see Bednarek 2005). With respect to the sentence The fish attacked the swimmer, the word shark worked better as a recall cue than fish, suggesting that readers constructed a mental model for the sentence (Brown and Yule
1983: 251, reporting on research by Anderson et al 1976). Experimental pragmatics bridges pragmatics and psycholinguistics. The 4th Experimental Pragmatics conference in April 2009 included research on topics such as semantic approaches, language development, impaired communication, non-verbal and
non-human communication, non literal language, negation, presupposition, reference, conditionals, sentence processing. (http://experimentalpragmatics.org/, last
accessed 17 October 2008).
With respect to experimental research it is questionable whether it is possible to
find out about how people behave in real life rather than in the artificial and simplified lab situation. Other problematic issues are related to how many subjects are
used and how representative they are, and to what degree they are influenced in
their language use by the experimental situation and the presence of the experimenter. Some of these factors may be less problematic in field experiments that do
not take place in a lab but rather in a real-world environment such as a classroom. With the help of a carefully constructed experiment and control groups, it is
argued that the advantage of experimental data is that they allow for the testing of
hypotheses, the control of variables and the analysis of cause-effect relationships.
(For a more detailed discussion of experimentation-based pragmatics see contributions to Noveck and Sperber 2004, Breheny in this volume.)
3.2.1.5. Elicited narratives
An important strand of research on language use analyzes elicited narratives, employing various techniques (including the use of non-linguistic data) to elicit narratives from subjects. For example, participants are shown a story represented in a
sequence of pictures or a film narrative and are then asked to retell the story. Thus,
Bamberg (1991) shows subjects a 24-picture story called Frog, where are you?
and subsequently asks them to tell the story depicted in this sequence of pictures.
The data are then analyzed in terms of narrative structure and narrative activity.
Similarly, Labovs classic analysis of the structure of narratives (abstract, orientation, complicating action, evaluation, result/resolution, coda) is based on elicited
narratives. Tannens (1993) data stem from a project in which small groups of
545
young Greek and American women watched a film (the pear film),3 and afterwards talked about its contents to another woman, who they were told had not seen
it (Tannen 1993: 21). Tannen analyzes the womens accounts of the film and categorizes sixteen types of linguistic evidence such as negatives and repetitions
which reveal their expectations or, in cognitive linguistic terms, their schemas (see
above). Thus, elicited narratives allow a variety of linguistic enquiries about topics
including narrative structure, the construal of narratives, and cognitive phenomena
to name but a few. However, they share some disadvantages with interviews, for
instance that they are not naturally occurring, and that it can be very time-consuming to elicit the data in the first place.
With respect to all elicited data discussed in the above sections (interviews,
verbal reports, discourse completion tasks, experimental data, and elicited narratives) as well as other kinds of elicited data not discussed here (such as role plays
and different types of questionnaire,4 e.g. open or multiple choice questionnaires)
the disadvantage is that they are elicited rather than naturally occurring. In
contrast, the following sections look at different types of non-elicited data as used
in pragmatics and related research into language use.
3.2.2.
Non-elicited data
Under the heading of non-elicited data we can include observational ethnographic

data and different kinds of text/discourse data, including corpora.
3.2.2.1. Ethnographic data
Ethnography has its roots in anthropology, in particular research by Bronislaw
Malinowski and Franz Boas as well as in the linguistic works of Edward Sapir (Titscher et al 2000: 90). It is self-reflexive, culturally founded and based on participant observation (Titscher et al 2000: 90). Ethnographic methods have emerging
research foci, based on what evolves from the researchers immersion in a particular context for a period of time (at least 612 months). Thus, participant observation is a crucial method in ethnography, which may be complemented by other
techniques such as interviews or document analysis (McNeill and Chapman 2005:
89). However, not all participant observation is ethnographic: A study becomes
ethnographic when the fieldworker is careful to connect the facts that s/he observes
with the specific features of the backdrop against which these facts occur, which
are linked to historical and cultural contingencies (Baszanger and Dodier 2004:
12). In other words, texts are interpreted in terms of cultural structures or used to
reconstrue such structures (Titscher et al 2000: 91).
The participant in participant observation can be actively involved in the social
group/activity which is the focus of research (covertly or overtly) but need not be.
His/her observations can be recorded in observational schemas, reports or field-
546
Monika Bednarek
notes. Data can be recorded and then transcribed or jotted down in notebooks as it
occurs. Yuan (2001) points out that the latter type of linguistic data is not very accurate in capturing the details of language use. When participant observation is
overt, the observers paradox (Labov 1972, cited in Wardhaugh 1986: 18) may play
a role. This paradox is that the aim of linguistic research is to find out how people
talk when they are not being systematically observed, but the data are available
only through systematic observation (Wardhaugh 1986: 1819). Ethnographic
data can be argued to show the complexity and dynamics of authentic behavior;
ethnographic methods are flexible and open to the observation of new phenomena.
However, ethnographic research is time-consuming, not replicable or quantifiable,
and not necessarily representative. It may, however, be ideal for generating initial
hypotheses about something totally unknown (Drnyei 2007: 133). While different types of ethnographic research exist (Baszanger and Dodier 2004: 19) in linguistics it is usually associated with John Gumperz and Dell Hymes and their ethnography of speaking or communication (e.g. Gumperz and Hymes 1972). For
instance, Hymes is interested in speech communities, speech events, speech styles,
and ways of speaking as units of analysis (Hymes 1972: 5358). Within speech
acts he proposes to look at various components such as setting, scene, participants,
outcomes and goals, key (tone/manner), genres, norms of interaction and interpretation. Key questions in such research are: Who speaks, with whom, when, where
and in what code, about what? (Titscher et al 2000: 95).
3.2.2.2. Texts/discourse data
With respect to the use of spoken and written text/discourse data, we can identify a
continuum in terms of size, with analyses ranging from individual case studies of
one or just a few texts, to the use of small-scale corpora consisting of a range of
texts to the recourse to large-scale corpora of hundreds of thousands or millions of
words.
Continuum of text/discourse data
Individual text(s):
case study/ies
small-scale
corpus
large-scale
corpus
Figure 1
Discourse data also vary in terms of their content and context of production such
that analyses can focus on political documents, speeches, media texts, conversations, radio interviews, reality TV talk, books, academic articles, advertisements, songs, poems, television series, university prospectuses, chat room interaction or websites to name but a few. These types of data offer different advantages.
For example, written language is usually easier to collect and to study than casual
547
conversation. Public spoken language (e.g. radio or internet broadcasts) is easier to

get hold of than conversations among friends or partners. Printed text
is infinitely easier to collect than audiovisual data and more permanent than most Web
material. It is already out there, ready to be gathered, and does not require time-consuming transcription before analysis. Also, the observers paradox is not an issue. Unlike, say, informants elicited responses or structured interviews, printed material is not
in danger of changing just because it is being observed. (Mautner 2008: 32)
In contrast,
[t]he variation in the types of contexts recorded as part of different spoken corpora, as
well as in the parameters used for collecting data, makes it difficult to accumulate a
large enough sample of spoken data that follows the same design principles. As a result,
research based on spoken data is often limited in terms of the number of examples of a
particular linguistic feature in the corpus, especially once contextual filters have been
applied. (Adolphs 2008: 33)
With respect to the recording of spoken data, the observers paradox may again be
relevant, since people may talk differently when they are being recorded, although
there are techniques for minimizing this problem, such as recording people for
long periods of time. Spoken data also needs to be transcribed according to specific
transcription systems (see Kreuz and Riordan in this volume). Since the emergence
of the internet and other technologies, data from the new media (such as emails,
chats, social forums, message boards etc) are also increasingly studied in pragmatics (Bublitz 2008; Gruber 2008).5 The majority of the new media incorporate
semiotic systems other than language, and multimodal research becomes important. Wodak (2008) even speaks of the visual turn (Wodak 2008: 2) in linguistics,
necessitating new toolkits and theories. Multimodality is a very vibrant field in linguistics nowadays, with many different approaches and notation/transcription systems (e.g. Hbler 2007; Jewitt 2009). In general, multimodal analysis, whether of
traditional or of new media, can become highly complex and time-consuming
(see Pollak 2008 on film analysis).
These and other types of spoken and written discourse data may be used in case
studies, small-scale corpus studies and discourse analyses, and large-scale corpus
linguistics. (Detailed descriptions of different kinds of corpora are given in most
introductions to corpus linguistics.) The way in which these data are analyzed
varies widely (see below and Andersen as well as Ruehlemann in this volume), but
all methodologies have in common the focus on the actual (rather than the possible) and all value authentic language-in-use. There are a variety of approaches that
usually analyze individual texts or small-scale collections of texts (including
specially designed corpora) such as ethnomethodological conversation analysis
and membership categorization analysis (Schegloff 2006; see also Holly and
Domke in this volume), interactional sociolinguistics, politeness theory, discursive
psychology, critical discursive psychology, content analysis, grounded theory, eth-
548
Monika Bednarek
nographic methods (discussed above), narrative semiotics, SYMLOG (system for

the multiple level observation of groups), critical discourse analysis, (German)
functional pragmatics, distinction theory text analysis, objective hermeneutics,
and systemic functional discourse analysis. See Taylor (2001), Stubbe et al (2003),
and Titscher et al (2000) on evaluations of and comparisons between these approaches to discourse data.
The advantages of looking at single texts or a limited number of texts or smallscale corpora are that the texts can be analyzed in depth through close-reading and
attention to manifold aspects of language use. Such an analysis allows researchers
to focus on the localized construal of social phenomena such as identity in particular contexts, often resulting in a complex, rich, interpretive, dynamic and flexible
analysis of micro-contexts and capturing the dynamic and negotiatory nature of
much language use. If video data are used (Heath 2004), the multimodal context can
be taken into account and attention can be paid not only to intonation and prosody
but also to gesture, posture, body movement and facial expression. However, while
individual case studies allow a focus on significant or important texts and events,
they lose out on generalizability and representativeness. They raise the question
of how far statements made about one particular text are transferable to other contexts. Case studies can also be time-consuming, relatively subjective, and are not
easily replicable. Analyses that make use of small-scale corpora or collections of
texts are more representative but can be just as subjective depending on the features
analyzed and the method of analysis. In addition to the advantages of case studies,
they offer observation of recurring regularities, practices or patterns, at least to
some extent. Thus, conversation analyses have uncovered regularities of turn-taking and sequencing using collections of spoken data. Small-scale corpus studies
have, for instance, been undertaken by Semino and Short (2004) with respect to reported speech and thought, by Bednarek (2006b) focusing on evaluative meaning in
a 70,000 word corpus of hard news stories, by Martin and White (2005: 165) on a
systemic functional linguistic appraisal analysis in 85 items from journalistic discourse and by Bednarek (2008a) focusing on emotion talk in an 85,000 word corpus
comprising conversation, news reportage, fiction and academic discourse. The corpora used in this research are small enough for manual, context-sensitive analysis
but large enough to show at least some patterns and allow some generalizability.
Coding can be supported through computer assisted manual annotation (CAMA),
where a human annotates the text in terms of patterns that generally computers
cannot recognize (ODonnell 2007). The latter makes use of smaller corpora because manual coding takes up time but it allows the researcher to explore semantic
and pragmatic patterning. ODonnell has designed computer software that facilitates such analyses: the UAM CorpusTool (http://www.wagsoft.com/CorpusTool/),
but other coding/annotation software also exists for use across the social sciences.
Finally, a few words on large-scale corpus data. Corpus linguistic applications
include discourse analysis, lexicography, stylistics, forensic linguistics, language
549
variation studies and language teaching (Baker 2006: 23). Although corpora have
in the past been neglected in pragmatics, corpora are increasingly used in studies of
language use, such as e.g. Adolphs (2008) and contributions to Skaffari et al
(2005). Compare also the three sub-issues on corpus linguistics in the Journal of
Pragmatics (2004: 36/2, 36/6 and 36/9).
A crucial advantage of corpus linguistic investigations lies in the representativeness of the data and the use of empirical, systematic, and sophisticated data. The
analyses are often reliable and replicable, especially when focusing on formally defined items. Further, corpus linguistics can uncover features of language that are inaccessible to intuition or a close reading type analysis of individual texts. This concerns patterning, typicality of usage, and quantification (e.g. type-token ratios,
frequency lists, lexical density) and particular kinds of semantic-pragmatic meanings such as semantic prosody (Louw 1993, Bednarek 2008c). Corpus data also
give researchers access to shared connotations (Coffin and OHalloran 2005) and
the experience of language users (Thompson and Hunston 2000: 15). They allow us
to study typicality adequately through access to large amounts of actually occurring
discourse, and enable easy application of tests of statistical significance through
user-friendly software. Thus, hypotheses can be verified, falsified or modified and
new language features can be uncovered with the help of large scale corpora.
However, there are also some disadvantages6 in that researchers are only able
to search the corpus for formally defined items, unless the corpus data have previously been annotated manually or semi-automatically (in which case there is a
limitation on the size of the corpus because this process can be quite time-consuming). Social categories such as gender or age are often treated as stable variables
rather than as discursively construed (Butler 1999), and many (though not all) corpus analyses fail to take into account reception, the dynamic nature of discourse, its
context or textual structure. As Stubbs (2008) notes with respect to keyword analysis, since the texts have been ripped apart into lists of individual words and/or
n-grams, the patterns ignore text segmentation. They are a feature of global textual
cohesion, but not textual structure (Stubbs 2008: 5). It has also not yet been discussed sufficiently to what extent statistical tests are adequate for data of authentic
language-in-use, to what extent linguists can evaluate and understand the information provided by statistics correctly, and to what extent they are able to
choose the most appropriate statistical tests for their analyses.
The majority of corpora are mono-modal, and the incorporation of semiotic
systems other than language is not yet wide-spread. With respect to multimodality,
Bauldry and Thibault (2006) stress the importance of using multimodal corpora in
semiotic analysis while noting that their design is very much in its infancy
(Bauldry and Thibault 2006: 181). Johnston (2008) has forcefully argued for the
need for sign language corpora.
Corpus analyses are also usually interested in the repeated, the typical, the frequent, rather than the individual and outstanding (but see e.g. Hoey 2005 on cre-
550
Monika Bednarek
ativity). For example, some texts are highly valued in the community or []
[have] special significance in some domain such as history or politics. Here we
treat texts as artefacts objects of study in their own right (Matthiessen 2006:
108). An emphasis on repeated patterns of semiotic behavior neglects the importance of outstanding singular texts that might have an impact on phylogenesis and
ontogenesis that goes beyond the impact of ordinary singular texts that achieve
their impact through repetition. An example from environmental discourse would
be the documentary An Inconvenient Truth produced and presented by Al Gore,
which has been widely credited with responsibility for a seismic shift in the attention of politicians, governments and the media towards environmental matters and
in policy changes around issues of environmental sustainability. In fact, its impact
was recognized in the awarding of the Nobel Peace Prize to Gore in 2007 for efforts to build up and disseminate greater knowledge about man-made climate
change, and to lay the foundations for the measures that are needed to counteract
such change (http://nobelprize.org/nobel_prizes/peace/laureates/2007/). The impact of such extraordinary texts, which are more powerful than normal texts produced countless times by speakers over time, presumably depends on diverse factors such as the profile of the author/speaker, the context of reception, and in the
case of An Inconvenient Truth, the sophisticated use of semiotic resources (Bednarek, Caple and Hood 2008). Studying such unique texts, then, also provides important insights into the construal of categories, and is an argument for qualitative
analyses of individual texts.
Further, some publications are more read than others and thus might have
greater impact on phylo- and ontogenesis:7 According to claims, the most likely
document that an ordinary English citizen will cast his or her eyes over is The Sun
newspaper (Sinclair 2005: 7). While it is not clear what exactly the impact of such
(valued and/or wide-spread) texts is, this has at least two implications for largescale corpus data. On the one hand, it needs to be acknowledged that special texts
become lost in the corpus where they have only as much weight as any other text:
[F]requent patterns of language do not always necessarily imply underlying hegemonic
discourses. Or rather, the power of individual texts or speakers may not be evenly distributed. A corpus which contains a single (unrepresentative) speech by the leader of a
country or religious group, newspaper editor or CEO may carry more weight discursively than hundreds of similar texts which were produced by ordinary people. (Baker
2006: 19)
On the other hand, this question clearly needs to be discussed in relation to corpus
design. For example, with respect to the quote above, Sinclair asks if we should include more texts from The Sun than from other newspapers in a corpus of British
English on account of its importance (Sinclair 2005: 7). In fact, many texts that are
included in corpora such as the British National Corpus differ in circulation status
(size of readership/circulation level) (Lee 2001: 68). Sinclair notes that issues of
551
representativeness and balance in corpus design are far from resolved at present
(Sinclair 2005: 79) and recommends that in the meantime corpus design should
be documented fully with information about the contents and arguments in justification of the decisions taken (Sinclair 2005: 8). With respect to corpus design
(Wynne 2005) other problematic issues include the thorny notion of representativeness (Mahlberg 2004), procedures of sampling (see Mautner 2008, Voormann
and Gut 2008), and practicalities of coding and inputting as well as technological
exigencies. It may also be difficult to find a corpus that has enough natural occurrences of individual pragmatic phenomena that the researcher is interested in (such
as speech acts, see Yuan 2001: 275) and not all pragmatic phenomena can be formally defined necessitating the use of small-scale corpora at least in a pilot study
or as complementary data.
4.
Conclusion: triangulating the data of pragmatics
In view of the above observations, it seems that all data of pragmatics and associated methodologies provide certain insights and perspectives, without necessarily
being able to capture the complexity of language in use in its entirety. The use of
different data and methodologies contributes variability to the field of pragmatics.
With respect to this, Kasper and Dahl (1991: 215) speak of a double layer of variability: variability that reflects the social properties of the speech event, and the
strategic, actional, and linguistic choices by which interlocutors attempt to reach
their communicative goals and the variability induced by different instruments
of data collection (Kasper and Dahl 1991: 215, cited in Woodfield 2007).
It thus seems reasonable to argue for a mixed-methods approach (Drnyei
2007) in pragmatics, combining quantitative and qualitative research. Triangulation, or the use of a variety of methods, allows researchers to cross-check and
verify the reliability of a particular research tool and the validity of data collected
(McNeill and Chatman 2005: 23). Researchers arguing for a combination of
methods include Holmes (1997), Taylor (2001), Holmes and Schnurr (2005),
Baker (2006) and Wodak and Krzyzanowski (2008).
If there is a team of researchers with various backgrounds, triangulation may be
less difficult than if there is one individual researcher only. Large research projects
with research teams are thus at an advantage, as they can combine expertise in different research methodologies and distribute the analyses among researchers. An
individual researcher may not be able to gain sufficient expertise in various research methods, and there are limitations to the extent to which s/he can undertake
different types of analysis for a single research project. One possibility of triangulation for discourse analysis is the three-pronged approach suggested by Bednarek (2008a, b, 2009). In essence, such an approach extends previous linguistic research that involves both corpus and discourse analysis, for instance corpus-based
552
Monika Bednarek
discourse analyses (Baker 2006), recent sociolinguistic research (e.g. Holmes and
Schnurr 2005), studies in Critical Discourse Analysis (Fairclough 2000; Barker
and Galasinski 2001: 26; Mautner 2008), and Matthiessens (2006) two-pronged
approach (Matthiessen 2006: 110). The three-pronged approach combines a
large-scale quantitative approach with a small-scale corpus analysis to be complemented by qualitative case studies. The large-scale corpus studies use a large (~
more than 500,000 or 1 million words) corpus and computer software to investigate the frequency and distribution of lexico-grammatical items. If the corpus design is reasonable, it can provide a relatively representative sample of language
that allows us to formulate hypotheses about general functions of the respective
lexico-grammatical items. On account of the large size of the corpus, however, the
description is less detailed than would otherwise be possible. In contrast, a smallscale corpus analysis uses a corpus of a size that is amenable to manual or semiautomated analysis, e.g. of semantic or pragmatic meanings, and that still shows us
intertextual patterning of such meanings. On account of its smaller size, such a corpus will be less representative, but its analysis will be more detailed and more interpretive. Finally, undertaking one or several case studies allows for qualitative
discourse analysis, which can take into account textual development and the sociocultural context (e.g. of production and reception) of the discourses at hand. It will
be least representative, in that researchers will be least able to generalize from findings, but its analysis will be most detailed and most interpretive. Thus, applying
these three methodologies can provide us with various perspectives on the discourse(s) at hand. However, even with this three-pronged discourse analytical approach, a time span of at least two to three years is necessary for the development
of theory, methodology, corpus design, software expertise, and analysis. In terms
of its focus on discourse, the analysis is also limited with respect to the various
ways in which this discourse can be taken up by its readership or audience. Generally speaking, the development of other mixed-method approaches in discourse
analysis and pragmatics is called for, and there is also a need for further evaluation
of how to approach the data of pragmatics, as there are only few studies that evaluate data-gathering and analysis in pragmatics (cf. Yuan 2001: 273 on the dearth of
such studies within empirical pragmatics). As Rose and Ono (1995) put it: We are
not claiming that one source of data is better than any other; we know too little to
make such claims and, frankly, we are sceptical of their value. It would be more
productive to ask better for what purpose? (Rose and Ono 1995: 207).
Notes
1. Research that compares DCT and other data includes Bardovi-Harlig and Hartford
(1993), Golato (2003), and Yuan (2001). Rose and Ono (1995) give a useful overview of
research on the methodology of DCTs in speech act studies.
553
2. Sampling procedures are relevant to many approaches to the data of pragmatics: different
types include ad hoc sampling, cluster sampling, multi-stage sampling, quota sampling,
random sampling, and stratified sampling (Titscher et al 2000).
3. About the history of the pear film:
In the mid-1970s Professor Wallace Chafe, a specialist in Native American languages,
decided to test how much a simple story will vary from language to language. With his
research team, he designed a very simple film to elicit stories from speakers around the
world, including uneducated speakers in rural areas. []
The Pear Stories film was designed to tap into universal experiences, by showing a
man harvesting pears, which are stolen by a boy on a bike. The boy has some other adventures with other children, before the farmer discovers that his pears are missing. The
film is six minutes long, in colour, with sound effects but no words. It was filmed in
northern California, near the University of California, Berkeley. The man who plays the
farmer is a Cuban.
The story line is deliberately loose and bland, to avoid imposing a strong U.S. cultural
bias. But the film makers deliberately inserted a short scene of a man leading a goat to
test descriptions of a background event with no later significance. The scene of falling
off the bike and spilling the pears can measure language for cause and effect. And the
unusual ping-pong toy tests how people describe an unfamiliar object. The final scene,
when the farmer discovers his fruit has been stolen, re-introduces a character who had
been off-screen for most of the film, and stimulates speakers to describe emotions and
state a moral. (http://www.pearstories.org/docu/ThePearStories.htm, last accessed
14 October 2008)
4. Questionnaires differ with respect to their object of analysis (e.g. factual/behavioral/attitudinal), their use or non-use of scales (e.g. the well-known Likert scale), whether they
are open-ended or closed-ended etc (Drnyei 2007).
5. Computer mediated communication (CMC) can be studied or can itself be made use of as
a research tool, with certain implications. On web-based studies compare e.g. Drnyei
(2007: 121122).
6. Baker (2006: 1721) discusses some limitations of corpus data and analysis, and Paltridge (2006: 174176) provides a useful, if brief, summary of and response to criticisms
of corpus studies.
7. Additionally, there may be differences in impact depending on who says what to whom
how in what context, and mental factors such as cognitive biases (Baker 2006: 11) are
also involved in discourse production and reception.
References
Adolphs, Svenja
2008
Corpus and Context. Investigating Pragmatic Functions in Spoken Discourse.
Anderson, Richard C., James W. Pichert, Ernest T. Goetz, Diane L. Schallert, Kathleen V.
Stevens and Stanley R. Trollip
1976
Instantiation of general terms. Journal of Verbal Learning and Verbal Behaviour 15: 667679.
554
Monika Bednarek
Austin, John L.
1975
How to Do Things with Words. 2nd ed. London: Oxford University Press.
Baker, Paul
2006
Using Corpora in Discourse Analysis. London/New York: Continuum.
Bamberg, Michael
1991
Narrative activity as perspective taking: the role of emotionals, negations and
voice in the construction of the story realm. Journal of Cognitive Psychotherapy 2: 275290.
Bardovi-Harlig, Kathleen and Beverly S. Hartford
1993
Redefining the DCT: Comparing open questionnaires and dialogue completion
tasks. Pragmatics and Language Learning 4: 143165.
Barker, Chris and Dariusz Galasinksi
2001
Cultural Studies and Discourse Analysis: A Dialogue on Language and Identity. Thousand Oaks: Sage.
Baszanger, Isabelle and Nicholas Dodier
2004
Ethnography: relating the part to the whole. In: David Silverman (ed.), Qualitative Research. Theory, Method and Practice, 934. London: Sage.
Baldry, Anthony and Paul Thibault
2006
Multimodal corpus linguistics. In: Geoff Thompson and Susan Hunston (eds.),
System and Corpus. Exploring Connections, 164183. London: Equinox.
Bednarek, Monika
2005
Frames revisited: the coherence-inducing function of frames. Journal of Pragmatics 37: 685705.
Bednarek, Monika
2006a
Epistemological positioning and evidentiality in English news discourse a
text-driven approach. Text and Talk 26/6: 635660.
Bednarek, Monika
2006b
Evaluation in Media Discourse. London/New York: Continuum.
Bednarek, Monika
2008a
Emotion Talk across Corpora. Basingstoke/New York: Palgrave Macmillan.
Bednarek, Monika
2008b
What the hell is wrong with you? A corpus perspective on evaluation and
emotion in contemporary American pop culture. In: Ahmar Mahboob and
Naomi Knight (eds.), Questioning Linguistics. 95126. Newcastle: Cambridge
Scholars Press.
Bednarek, Monika
2008c
Semantic preference and semantic prosody re-examined. Corpus Linguistics
and Linguistic Theory 4/2: 119139.
Bednarek, Monika
2009
Corpora and discourse: a three-pronged approach to analyzing linguistic data.
In: Michael Haugh, Kate Burridge, Jean Mulder and Pam Peters (eds.), Selected Proceedings of the 2008 HCSNet Workshop on Designing the Australian
National Corpus: Mustering Languages, Sommerville, MA: Cascadilla Proceedings Project.
Bednarek, Monika, Helen Caple and Sue Hood
2008
Communicating a sustainable Australia: discourses of persuasion. Unpublished manuscript, Faculty of Arts and Social Sciences, University of Technology, Sydney.
555
Benwell, Bethan and Elizabeth Stokoe

2006
Discourse and Identity. Edinburgh: Edinburgh University Press.
1982
Learning to say what you mean in a second language. Applied Linguistics 3:
2959.
1987
1983
Bublitz, Wolfram
1998
I entirely dot dot dot: copying semantic features in collocations with up-scaling intensifiers. In: Rainer Schulze (ed.), Making Meaningful Choices in English, 1132. Tbingen: Narr.
Bublitz, Wolfram
2008
Sailing the islands or watching from the dock: the treacherous simplicity of a
metaphor. How we handle new (electronic) hypertext versus old (printed)
text. In: Andrea Gerbig and Oliver Mason (eds.), Language, People, Numbers
Corpus Linguistics and Society, 249273. Amsterdam: Rodopi.
Butler, Judith
1999
Gender Trouble: Feminism and the Subversion of Identity. New York: Routledge.
Cermk, Frantisek
2001
Research Methods in Linguistics. Essential Principles, Based on a General
Theory of Science. Prague: Karolinum.
Coffin, Caroline and Kieran OHalloran
2005
Finding the global groove: theorising and analysing dynamic reader positioning using APPRAISAL , corpus, and a concordancer. Critical Discourse Studies
2/2: 143163.
Drnyei, Zoltan
2007
Research Methods in Applied Linguistics. Oxford: Oxford University Press.
Fairclough, Norman
2000
New Labour, New Language? London: Routledge.
Galasinski, Dariusz
2004
Men and the Language of Emotions. Basingstoke/New York: Palgrave Macmillan.
Golato, Andrea
2003
Studying compliment responses: a comparison of DCTs and recordings of naturally occurring talk. Applied Linguistics 24/1: 90121.
Grice, Paul
1989
Gruber, Helmut
2008
Analyzing communication in the new media. In Ruth Wodak and Michal Krzyz anowski (eds.), Qualitative Discourse Analysis for the Social Sciences,
5476. Basingstoke/New York: Palgrave Macmillan.
1972
Directions in Sociolinguistics. The Ethnography of Communication. New
York: Holt, Rinehart and Winston.
556
Monika Bednarek
Haastrup, Kirsten
1987
Using thinking aloud and retrospection to uncover learners lexical inferencing procedures. In: Claus Faerch and Gabriele Kasper (eds.), Introspection in
Second Language Research. 197212. Clevedon: Multilingual Matters.
2004
The spoken language corpus: A foundation for grammatical theory. In: Karin
Aijmer and Bengt Altenberg (eds.), Advances in Corpus Linguistics. Papers
from the 23rd International Conference on English Language Research on
Computerized Corpora. 1139. Amsterdam/New York: Rodopi.
Heath, Christian
2004
Analysing face-to-face interaction: video, the visual and material. In: David
Silverman (ed.), Qualitative Research. Theory, Method and Practice. 266
282. London: Sage.
Hoey, Michael
2005
Lexical Priming. A New Theory of Words and Language. London: Routledge.
Holmes, Janet
1997
Women, language and identity. Journal of Sociolinguistics 1/2: 195223.
Holmes, Janet and Stephanie Schnurr
2005
Politeness, humor and gender in the workplace: Negotiating norms and identifying contestation. Journal of Politeness Research 1: 121149.
Hbler, Axel
2007
The Nonverbal Shift in Early Modern English Conversation. Amsterdam: Benjamins.
Hymes, Dell
1972
Models of the interaction of language and social life. In John J. Gumperz and
Dell Hymes (eds.), Directions in Sociolinguistics. The Ethnography of Communication, 3571. New York: Holt, Rinehart and Winston.
Jewitt, Carey
2009
The Routledge Handbook of Multimodal Analysis. London: Routledge.
Johnston, Trevor
2008
Signed languages and corpus linguistics. Paper presented at the 2nd International Free Linguistics Conference, University of Sydney, 1112 October
2008.
Kasper, Gabriele
2000
Data collection in pragmatics research. In: Helen Spencer-Oatey (ed.), Culturally Speaking. Managing Rapport through Talk Across Cultures, 316341.
London/New York: Continuum.
Kasper, Gabriele and Merete Dahl
1991
Research methods in interlanguage pragmatics. Studies in Second Language
Acquisition 13/2: 215247.
Labov, William
1972
Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press.
Lee, David
2001
Genres, registers, text types, domains, and styles: clarifying the concepts and
navigating a path through the BNC jungle. Language Learning & Technology
5: 3772. Retrieved from http://llt.msu.edu/vol5num3/lee/ on 20 October
2005.
557
Leech, Geoffrey
1983
Levenston, Edward
1975
Aspects of testing the oral proficiency of adult immigrants to Canada. In: Leslie Palmer and Bernard Spolsky (eds.), Papers on Language Testing 1967
1974, 6774. Washington, DC: TESOL.
Louw, Bill
1993
Irony in the text or insincerity in the writer? In: Mona Baker, Gill Francis and
Elena Tognini-Bonelli (eds.), Text and Technology. In Honour of John Sinclair,
Mahlberg, Michaela
2004
Corpus design and the words in a dictionary. Lexicographica 20: 11429.
Martin, James R. and Peter R.R. White
2005
The Language of Evaluation: Appraisal in English. Basingstoke/New York:
Palgrave Macmillan.
Matthiessen, Christian M.I.M.
2006
Frequency profiles of some basic grammatical systems: an interim report. In:
Susan Hunston and Geoff Thompson (eds.), System and Corpus: Exploring
Connections, 103142. London: Equinox.
Mautner, Gerlinde
2008
Analyzing newspapers, magazines and other print media. In Ruth Wodak and
Michal Krzyzanowski (eds.), Qualitative Discourse Analysis for the Social
Sciences, 3053. Basingstoke/New York: Palgrave Macmillan.
McNeill, Patrick and Steve Chapman
2005
Research Methods. 3rd ed. London: Routledge.
Miller, Jody and Glassner, Barry
2004
The inside and the outside: finding realities in interviews. In: David Silverman (ed.), Qualitative Research. Theory, Method and Practice, 125139.
London: Sage.
Noveck, Ira A. and Dan Sperber (eds.)
2004
Experimental Pragmatics. Basingstoke/New York: Palgrave Macmillan.
ODonnell, Mick
2007
The UAM Corpus Tool. Paper presented at ICAME 28, Stratford-upon-Avon,
UK, 2327 May 2007.
Paltridge, Brian
2006
Discourse Analysis. London/New York: Continuum.
Pollak, Alexander
2008
Analyzing TV documentaries. In: Ruth Wodak and Michal Krzyzanowski
(eds.), Qualitative Discourse Analysis for the Social Sciences, 7795. Basingstoke/New York: Palgrave Macmillan.
Rose, Kenneth and Reiko Ono
1995
Eliciting speech act data in Japanese: the effect of questionnaire type. Language Learning 45/2: 191223.
Schegloff, Emanuel
2006
A tutorial on membership categorization. Journal of Pragmatics 39: 462482.
Searle, John
1969
558
Monika Bednarek
Semino, Elena and Mick Short

2004
Corpus Stylistics: A Corpus-based Study of Speech, Thought and Writing in a
Corpus of English Writing. London: Routledge.
Skaffari, Janne, Matti Peikola, Ruth Carroll, Risto Hiltunen and Brita Wrvik (eds.)
2005
Opening Windows on Texts and Discourses of the Past. Philadelphia: Benjamins.
Silverman, David (ed.)
2004
Qualitative Research. Theory, Method and Practice. 2nd ed. London: Sage.
Sinclair, John
2005
Corpus and text basic principles. In: Martin Wynne (ed.), Developing Linguistic Corpora. A Guide to Good Practice, 116. Oxford: Oxbow.
Stubbe, Maria, Lane, Chris, Hilder, Jo, Vine, Elaine, Vine, Bernadette, Marra, Meredith,
Holmes, Janet and Ann Weatherall
2003
Multiple discourse analyses of a workplace interaction. Discourse Studies 5/3:
351388.
Stubbs, Michael
2008
Three concepts of keywords. Revised version of paper presented to the conference on Keyness in Text, Certosa di Pontignano, University of Siena, June
2007. Retrieved from http://www.uni-trier.de/fileadmin/fb2/ANG/Linguistik/
Stubbs/stubbs-2008-keywords.pdf on 7 August 2008.
Tannen, Deborah
1993
Whats in a frame?: surface evidence for underlying expectations. In: Deborah
Tannen (ed.), Framing in Discourse, 1456. Oxford: Oxford University Press.
Taylor, Stephanie
2001
Evaluating and applying discourse analytic research. In: Margaret Wetherell,
Stephanie Taylor and Simeon Yates (eds.), Discourse as Data: A Guide for
Analysis, 311330. London: Sage in association with The Open University.
Thompson, Geoff and Susan Hunston
2000
Evaluation: an introduction. In: Susan Hunston and Geoff Thompson (eds.),
Evaluation in Text. Authorial Stance and the Construction of Discourse, 127.
Titscher, Stefan, Michael Meyer, Ruth Wodak and Eva Vetter
2000
Methods of Text and Discourse Analysis. Translated by Bryan Jenner. London:
Sage.
Tognini-Bonelli, Elena
2001
Corpus Linguistics at Work. Amsterdam: Benjamins.
Voormann, Holger and Ulrike Gut
2008
Agile corpus creation. Corpus Linguistics and Linguistic Theory 4/2: 235251.
Wardhaugh, Ronald
1986
An Introduction to Sociolinguistics. Oxford: Basil Blackwell.
Wetherell, Margaret, Stephanie Taylor and Simeon J. Yates (eds.)
2001
Discourse as Data: A Guide for Analysis. London: Sage in associated with The
Open University.
Wodak, Ruth
2008
Introduction: Discourse studies important concepts and terms. In: Ruth
Wodak and Michal Krzyzanowski (eds.), Qualitative Discourse Analysis for
the Social Sciences, 129. Basingstoke/New York: Palgrave Macmillan.
559
Wodak, Ruth and Michal Krzyzanowski (eds.)

2008
Qualitative Discourse Analysis for the Social Sciences Basingstoke/New York:
Palgrave Macmillan.
Woodfield, Helen P.
2007
Problematizing discourse completion tasks: voices from verbal report. Evaluation and Research in Education 20/2. Retrieved from
http://hdl.handle.net/1983/943 on 14 October 2008.
Wynne, Martin (ed.)
2005
Developing Linguistic Corpora. A Guide to Good Practice. Oxford: Oxbow.
Yuan, Yi
2001
An inquiry into empirical pragmatics data-gathering methods: Written DCTs,
oral DCTs, field notes, and natural conversations. Journal of Pragmatics 33/2:
271292.
20.
Experimental pragmatics*
Richard Breheny
1.
Introduction
Pragmatics is a mature discipline, focussed largely on the use of language or communication in context. Experimental psychology is a mature discipline, focussed in
part on language and communication. It may come as a surprise then to find that
Experimental Pragmatics, that is, the development of pragmatics as a theoretical
discipline via experimental methods, is more of a fledgling field. While there had
been a small but influential tradition in psychological research informed by pragmatic theory (see for instance, Clark 1996), until recently there was comparatively
little theoretical pragmatics research founded on the experimental methods of the
psychological sciences. Likewise, while there is a strong tradition of psycholinguistic research on the interpretation of discourse in context (see Graesser et al
2003 for a recent handbook on the topic), until recently little of this research had
been much interested in the subject matter of pragmatics. Why this state of affairs
persisted for so long and how the situation is changing are questions that will be
addressed in this entry. In addition, the current state of experimental pragmatics
will be reviewed and future prospects surveyed.
Recent experimental pragmatics has focussed on Gricean pragmatics, which
considers phenomena such as implicature, presupposition, figurative language and,
more generally, speaker meaning. Thus the focus of this entry will reflect this bias.
2.
The foundations of pragmatics and experimental methods
Modern experimental psychology aims to develop scientific theories of individual

mental, cognitive or psychological capacities; normally of how certain functions,
such as vision, planning or memory, are realised. Contemporary pragmatics, particularly as it interfaces with formal linguistics, stems largely from the tradition
of analytic philosophy through the work of Austin, Grice, Carnap, Lewis among
others. It is not straightforward to apply the methods of experimental psychology
to this field of inquiry since thinkers like Grice and Austin were not in the business
of proposing accounts of individual pragmatic competencies. For analytic philosophy, questions about how we cognise language in context are separate from critical philosophical questions about meaning in context. If we see pragmatics as a
continuation of the analytic tradition, there would seem to be little scope for the application of experimental psychological methods in pragmatics. Things are not that
562
Richard Breheny
different if we consider more socially oriented pragmatics, which takes inspiration

from social sciences such as anthropology, sociology and economics. While these
disciplines are more empirically grounded, their methods rarely overlap with those
of the experimental psychologist.
Given the philosophical and social scientific foundations of pragmatics, perhaps we should not be surprised to find that Experimental Pragmatics has been a
minority pursuit, at least until relatively recently. However, pragmatics has also
flourished not so much as a branch of social sciences like anthropology or of analytic philosophy but as a branch of contemporary linguistics. Given the nature
of the more dominant formal end of this discipline, it is more surprising to find
the lack of experimental pragmatics. After all, most of contemporary formal linguistics takes itself to be part of the cognitive scientific enterprise, where theory is
ultimately aimed at explaining the cognitive psychological underpinnings of linguistic and pragmatic phenomena.
3.
Introspection and its limits in formal pragmatics
Since its inception, the main method for collecting empirical data in formal linguistics has been to introspect on intuitions (and check with a small number of informants if judgments are not sharp). While this methodology has served the discipline well in syntax and semantics, it might be questioned in the case of
pragmatics. For pragmatics interfaces with formal linguistics mostly in the domain
of generalised pragmatic inference. Probably the best known example of generalised phenomena are the scalar implicatures formalised (to a greater or lesser degree) in the neo-Gricean works of Gazdar (1979), Horn (1989), and Levinson
(Atlas and Levinson 1981, Levinson 2000).1 In these works, Grices informal account of implicatures is recast so that generalised implicatures become default.
I.e. the implicatures can be derived solely on the basis of the structure of the sentence used and only become unavailable (cancelled) under specific contextual
conditions. While neo-Gricean theory provides a computationally more tractable
description of scalar implicature than Grices original informal accounts, and
hence better meets the desiderata of formal linguistics, the empirical basis for this
description is the intuition that these implicatures are default. To take a concrete
example, the assumption is that an implicature as in (1b) is available from the use
of a sentence like (1a) unless context overrides this (for example by the speaker
saying, In fact all of them did):
1. a. Some of the consultants had a meeting with the director.
b. Not all of the consultants had a meeting with the director.
Note however that any introspective evidence of the defaultness of (1b) depends on
implicit assumptions to the effect that implicatures are available independently of
563
an utterance situation (real or imagined). To the contrary, one could assume that no
implicature becomes available to introspection without a real or imagined utterance situation. In that case, one could observe that it would be dubious to try to ascertain introspectively whether an implicature is available when ones interest is in
whether the implicature in question is available. Although neo-Gricean theory has
been around for thirty years or more, no argument has been put forward to defend
the implicit assumption behind introspective evidence for the defaultness of certain implicatures. In a recent important set of studies, Geurts and Pouscoulous
(2009) have established that introspection on the availability of an implicature introduces a substantial bias in favour of judgements of availability. In their design,
Geurts and Pouscoulous had participants perform two tasks. Their inference task
mirrors the task undertaken by the theorist when she introspects: participants were
asked about whether a putative scalar implicature, like that in (1b), followed from a
simple sentence containing the implicature trigger, like (1a). Using separate items
in a verification task, participants had to verify the truth of a sentence like (1a)
when in fact all consultants had a meeting with the director (contrary to the implicature in (1b)). In a number of different kinds of construction, Geurts and Pouscoulous found that participants were significantly more likely to base their response on
an interpretation that included the implicature in the inference task than the verification task. Their conclusion is that, no matter how consistent and robust an introspected inference may be, that doesnt say anything about the frequency with
which it is drawn in practice (Geurts and Pouscoulous 2009 p.4.18). These results
reinforce the view that it is very important to establish the defaultness of a given
pragmatic effect (like scalar implicature) by methods other than introspection. As
we will see in the next section, much recent work in experimental pragmatics has
focussed on scalar implicatures and is informative on this issue.
4.
Testing for defaultness and preferred readings
One clear statement of a cognitively oriented neo-Gricean theory can be found in

Levinson (2000). Levinson argues that there exist utterance-type meanings that include default implicatures, and that these implicatures are actually generated upon
encountering the appropriate linguistic stimuli. This thesis has been directly and
indirectly tested in experimental work. One set of studies builds on a sentence
verification paradigm established in Noveck (2001). In those studies, participants
are asked to verify statements like Some elephants have trunks. Noveck (2001)
found that a substantial proportion of participants responded negatively, suggesting that those participants had based their judgement on the meaning of the sentence plus the implicature, not all elephants have trunks; while a smaller but still
substantial number responded yes, apparently basing their response on the plain,
logical meaning of the sentence. In subsequent studies, Noveck and Poseda
564
Richard Breheny
(2003) and Bott and Noveck (2004), measured response times to these questions
and found that it was significantly greater when participants responses were based
on the enriched meaning the one that included the implicature. In a follow up
study, Bott and Noveck (2004) also found that a group that was put under time
pressure to respond as quickly as possible gave significantly more logical responses than the group that was not. These results are relevant to the question of
whether generalised implicatures are generated by default. According to Levinson
(2000), where default implicatures are unavailable this is the result of cancellation a process akin to revision when a sentence processor has been garden
pathed. As such, responses based on an interpretation of a stimulus without an implicature ought, other things equal, to manifest a greater cost than those based on
an interpretation with an implicature. As just discussed, in the speeded verification
tasks, Noveck and colleagues found the opposite pattern.
In a more direct measure of on-line access of implicatures, Breheny et al (2006)
tested Levinsons theory employing a reading time methodology. In one experiment, participants read short discourses which provided either a context which
heavily biased a scalar implicature (as in (2)) or one in which the implicature is
generally felt to be unavailable due to lack of relevance (see (3)). The latter kind of
contexts are referred to as implicitly cancelling contexts:
2. Mary asked John where he would accommodate guests for the weekend of the
wedding. John replied that /some of his relatives/ would stay with him. /The
rest/ would book into hotels.
3. Mary asked John why he was cleaning his apartment. John replied that /some of
his relatives/ would stay with him. /The rest/.would book into hotels.
Reading times on the segment the rest confirmed that the and not all implicature
was highly salient in the biasing contexts but not so in the context where the implicature was irrelevant. In line with results in Bott and Noveck (2004), Breheny et
al (2006) found that reading times for the segments around the quantifier some
were significantly longer in the biasing context, like (2). These results again run
contrary to the predictions of Levinsons theory which holds that in the implicit
cancellation context, extra work is required to suppress an activated implicature.
By contrast, all of these results are in line with a more Gricean view whereby implicatures are only available and drawn where warranted by specific features of the
context (as per Sperber and Wilson 1986).
Breheny et al (2006) present a further study to directly compare the predictions
of Levinsons strong interpretation of neo-Gricean theory and a more fundamentally Gricean view. In this study, there was a simple manipulation of the position in
the sentence in which an implicature trigger occurs. For example, participants read
one of the discourses in (4a,b):
565
4. a. Some of the consultants had a meeting with the director. The rest did not
manage to attend.
b. The director had a meeting with some of the consultants. The rest did not
manage to attend.
Citation examples presented out of the blue in theoretical discourse or indeed in
experimental settings are sometimes referred to as neutral contexts (cf. Levinson
2000, Chierchia 2004). There is no agreed definition for neutral in this kind of
discourse but the suggestion seems to be that the examples are interpreted without
any presumptions about the utterance situations in which the items are produced. A
more Gricean view would argue that in as far as the examples are given an interpretation by the reader or experimental participant, there ought to be some imagined utterance situation in which these are produced. That is, they are treated as
utterances (albeit in written form), produced by a speaker with a purpose. Breheny
et al reasoned that if experimental participants treated these items as occurring in
imaginary utterance situations, then other things equal, they would accommodate
context incrementally. Thus it is more likely with (4a) than (4b) that the utterance
is addressing some concern about the consultants, and so in (4a) and not (4b) it is
more likely that participants would derive the implicature. So, by the Gricean account, we would expect to see this effect manifested in response times to the anaphor the rest, which is easier to interpret in contexts where the not all inference
has been drawn. Stronger versions of neo-Gricean theory, including Levinsons,
predict that in these neutral contexts the implicatures are equally likely to be generated. Thus there should be no difference. In line with the Gricean account, Breheny et al found a difference, suggesting that sentence position has an effect on implicature.
In a further test of the strong neo-Gricean view, Grodner and Sedivy (forthcoming) adapted a visual-world design (see Cooper 1974, Tanenhaus et al 1995) from
Sedivy et al (1999) to be discussed below which involves participants hearing
instructions such as in (5):
5. Pick up the tall glass.
Modified referential phrases like the tall glass often presuppose that there is more
than one object that satisfies the nominal description glass. In a series of papers,
Sedivy and colleagues have established that this extra presupposition is effectively
a Gricean implicature. Grodner and Sedivy (forthcoming) tested the claim in Levinson (2000) that this inference only requires the appropriate linguistic context to
be generated. One group of participants in their study were given information that
the speaker producing instructions like (5) was unreliable in certain relevant ways.
Grodner and Sedivy reasoned that if the inference is triggered solely by the linguistic form, even though there was reason to think that the speaker may not have
intended the implicature, in early eye-movements to the visual array, one should
566
Richard Breheny
see evidence that the implicature was accessed before it could be cancelled. By
contrast, if the information about the speaker is allowed to modulate interpretation,
then in contrast to the control group, one should see no effect of the implicature in
early eye-movements. In line with the off-line and reading time studies reviewed
above, Grodner and Sedivy again found that the contextual information about the
speaker does modulate the on-line access of implicatures.
The findings of these studies point toward the conclusion that even apparently
common implicatures are not generated solely on the basis of linguistic context in
on-line comprehension. While these findings tend to disconfirm Levinsons particularly strong version of neo-Gricean theory, other versions of neo-Gricean pragmatics could be adapted as the basis of a cognitive theory of pragmatics, for instance, those found in Horn (1984, 1989), Gazdar (1979) or Sauerland (2004).
These accounts presume the default availability of generalised implicatures, rather
than actual default generation. The difference is important for predictions in empirical studies that seek to measure participants responses based on the inclusion of
an implicature in the interpretation of an utterance. For, just as certain logically derivable entailments of a discourse may not necessarily be generated or accessed in
a given circumstance, so it may be that default implicatures may not necessarily be
derived or generated in comprehension. The parallels with deductive inferences
are important for another reason since it is quite reasonable to view pragmatic description as operating at a normative, rather than cognitive level. The relation between the normative level and the level of individual psychological competence
may be very indirect (see Saul 2002). The more indirect relation between neoGricean pragmatics and cognition is suggested in Horn (2006) and simply manifests a more specific version of the position held by Grice (see Warner 2001, Petrus
2010) and others in the analytic tradition.2 Adopting this stance results in a disengagement of pragmatics from empirical, experimental research of the kind undertaken in contemporary psychology. While it remains a very interesting question
whether it is viable to take either the normative stance or the more neutral a priori
stance towards pragmatics, it is a question which is rarely addressed. For the most
part, as mentioned above, those interested in Gricean pragmatics are not philosophers but linguists who work on the assumption that linguistics is in the business
of explaining individual behaviour via theories of the cognitive processes that
underlie that behaviour. From this perspective, pragmatic theory should be seen as
a theory of individual competence, and pragmatic description ought ultimately to
be specified in such a way that it becomes open to empirical (dis-) confirmation.
Thus an account of pragmatic competence based on neo-Gricean theory according to which implicatures merely become available by default cannot be tested
unless we make some additional assumptions about when available pragmatic inferences are accessed. For instance, neo-Gricean theory could be implemented in a
constraint-based framework (MacDonald and Seidenberg 2006) so that generalised implicatures would be accessed in the presence of certain cues. However, it
567
then falls to the proponent of default implicatures to specify in some more detail
under what condition implicatures are accessed.
It has become clear through the studies in Geurts and Pouscoulous (2009) and
the various studies testing Levinsons strong version of neo-Gricean theory that
there is very little evidence that even the most proto-typical scalar implicatures are
default. Only when reflecting on the availability of these implicatures does one get
a judgement that the implicature is reliably available. In other kinds of task, there is
much evidence in support of an alternative, more fundamentally Gricean, view that
implicatures are derived if and only if warranted by evidence in the utterance situation. So while it is still perhaps an open empirical question whether neo-Gricean
pragmatics offers insights for experimental pragmatics, it is a question which will
require a more thorough engagement with experimental research into utterance interpretation to answer.
With evidence supporting the idea that implicatures are only derived where
warranted by information about the speakers likely intentions in context, we can
ask if there are any theories of pragmatic competence that take a more fundamentally Gricean stance on utterance interpretation. One theory which is notably
Gricean in this respect is proposed in Sperber and Wilson (1986), which argues
that inferring the speakers informative intention is greatly facilitated by the existence of an underlying Relevance mechanism. By contrast, Geurts (2010) is an advocate of a psychologically oriented Gricean position, which adopts Grices original ideas about pragmatic inference based on the Co-operative Principle and
maxims as the starting point of a theory of pragmatic competence. Similarly Clark
(1996) and Tomasello (2008) adopt elements of Grices work in their joint-action
approach to language use.
The viability of more Gricean theories of pragmatic competence has been recently questioned in Sedivy (2003) where it is argued that the rapidity of access to
implicatures on line is difficult to square with the apparent complexity of the reasoning that lies behind these implicatures. Breheny et al (under review a) directly
test whether very rapid access of implicatures in on-line comprehension could be
the result of full-blown pragmatic reasoning on line. The studies involved quantity
implicatures that have no linguistic trigger, as where (6a) below implicates (6b) or
(6c):
6. a. The student put a book on the dusty shelf
b. The student put nothing else on the dusty shelf
c. The student put a book nowhere else
According to both the standard Gricean account and the Relevance Theory account
(see Sperber and Wilson 1995), implicatures such as those in (6b) and (6c) are derived on the basis of assumptions about the purpose (or source of Relevance) of the
utterance and on considering whether the speaker could know alternative, more informative propositions. Breheny et al (under review a) report a series of visual-
568
Richard Breheny
world studies based on the look and listen paradigm first explored in Altmann
and Kamide (1999). In this paradigm, participants are presented with a visual display and listen to spoken discourse. It has been consistently found that eye-movements around the visual scene reflect participants understanding of the text. In
particular, participants looks around the display show that they can anticipate reference to an object on the basis of their understanding of the previous parts of the
utterance but before reference is explicitly made. For instance, in Altmann and Kamide (2007) participants hear about an agent who has drunk or will drink certain things. Where the visual display contains a full glass (of beer) and an empty
wine glass among other items, participants look to the empty glass when the verb is
in perfect tense and at the full glass when the verb is in the future tense. In both
cases, anticipatory looks precede mention of the critical noun, wine or beer.
Breheny et al (under review a) adapt this paradigm to explore whether a quantity
implicature like (6b) is accessed by participants on-line, given the right biasing
context. In a visual display containing a representation of the student putting a
book and another item to one location (a table) and a book and nothing else to the
shelf, participants do anticipate reference to the shelf before they hear the noun
shelf, given the appropriate context that biases the and nothing else quantity
implicature. It is also shown that, at an earlier stage in processing, participants are
sensitive to what the purpose of the utterance is and that they are aware of more informative alternatives. Thus there is evidence that we can and do access quantity
implicatures in incremental interpretation of utterances and that we do so on the
basis of sensitivity to the likely purpose of the utterance and reasoning about the
speakers mental state in the process.
5.
Pragmatic theory meets psycholinguistic theories of the architecture

of on-line comprehension
Independently of any question about the correct theoretical account of the availability and/or derivation of implicatures, we can ask questions about the cognitive
mechanisms that underpin semantic and pragmatic processing, how these mechanisms interact and how the whole architecture of comprehension develops through
infancy and childhood. We will consider developmental research in the next section. In this section, we look at recent research relevant to the architecture of adult
comprehension. As we will see, in this domain, higher level pragmatic theory and
psycholinguistic theory are able to inform one another to their mutual benefit.
As mentioned above, it has been established that we are capable of accessing
quantity implicatures in on-line, incremental interpretation of utterances. While
Breheny et al (under review a) have shown this for quantity implicatures that have
no linguistic trigger, as in (6a-c), Breheny et al (2006) and Breheny et al (under
review b) have demonstrated this for the more familiar case where saying some
569
implicates not all. As we will see in this section, there is already a tradition of
research in psycholinguistics that has looked at another kind of pragmatic implicature, having to do with establishing the reference of a definite expression. While
this research has been relevant to the question of whether any pragmatic implicatures are default, it is equally relevant to architectural issues.
Sedivy et al (1999) have established that, in a visual context where there is
more than one glass, participants already infer that the referent of the tall glass in
(5) is the taller of the glasses before they hear glass, even in the presence of another tall object. Sedivy (2003) reports a study where the scalar adjective tall is
replaced by a material adjective (plastic, wooden etc) and the same effect is
found. Grodner and Sedivy (forthcoming) have established that the effect disappears in an unreliable speaker condition. All of this evidence points to the conclusion that the inference about the referent based on the modifier alone is a
Gricean one: comprehenders assume that speakers will avoid redundancy in their
utterances. When they encounter an adjectival modifier in a definite expression in a
visual context containing two potential referents of the same type, participants can
infer that the likely reason for the presence of the modifier is to distinguish between the two referents.
As participants in these eye-tracking studies seem to be able to access this inference very rapidly (between the onset of tall and the onset of glass), we can
ask how this is possible. As we have seen above, Sedivy and colleagues have ruled
out the theory that rapid access to generalised implicatures only requires formbased cues. As also noted above, one well-motivated alternative hypothesis is that
pragmatic inferences are accessed in on-line interpretation via a constraint-based
processing mechanism. Such a mechanism is interactive, frequency-based and
automatic, meaning that at any point in language processing, the currently favoured hypothesis is based on its likelihood given information from all levels of
processing (including contextual information). While a constraint-based approach
that simply implements neo-Gricean theory would work for common generalised
implicatures, it would have to be extended considerably to work for the ad hoc implicatures studied in Breheny et al (under review a) and in the studies reported in
Huang and Snedeker (2009) and Grodner et al (forthcoming) to be discussed
below. Still, Breheny et al (under review a) argue that a constraint-based architecture could serve to implement more classical Gricean theory, where inferences
about the likely purpose of an utterance, and other aspects of the speakers epistemic state are made automatically and probabilistically.
An alternative to a constraint-based approach to pragmatic processing is one
according to which pragmatic implicatures are accessed only once an interpretation of the linguistic expressions is made, based solely on the compositional semantic meaning of those expressions. This alternative is proposed in Huang and Snedeker (2009), who report on similar visual-world studies to those in Sedivy et al
(1999), but which have different results. In Huang and Snedeker (2009), partici-
570
Richard Breheny
pants are given instructions like in (7) where the visual display has a girl with socks
and a girl with soccer balls:
7. a. Point to the girl that has some of the soccer balls.
b. Point to the girl that has all of the socks.
While participants can disambiguate the referring expression on hearing all in
(7b), they cannot disambiguate the referent for (7a) until after the recognition point
of soccer unless they are able to make a pragmatic inference. Huang and Snedeker found that participants were able to anticipate the correct referent in the
some condition before the recognition point but that, compared to the all condition, they did so at a delay. Huang and Snedeker argue that this delay reflects both
that implicatures are accessed only after the compositional meaning of the sentences is activated and that this is a reflection of the architecture of comprehension
whereby implicatures belong to a different, pragmatic level of representation than
the conventional meaning of a sentence.
The findings in Huang and Snedeker (2009) are difficult to square with those of
Sedivy and colleagues which suggest the pragmatic inference is accessed very
rapidly. So, the question of the time course of accessing implicatures is still open.
This is particularly the case as Grodner et al (2010) found very rapid access to the
pragmatic implicature in a study almost identical to that of Huang and Snedeker.
Breheny et al (under review b) argue that a good Gricean account is available to explain the difference in timing among these studies and present look and listen visual-world studies showing that participants can access scalar implicatures of the
some/not all variety as rapidly as information based on the meaning of all.
Although it is certainly too early to tell for sure, there is mounting evidence that
the results of Gricean pragmatic reasoning can be accessed as rapidly as the results
of deriving information about the interpretation of an utterance from the compositional meaning of the sentence uttered. This evidence suggests that processes that
underpin access to pragmatic implicatures can in some cases operate automatically
or spontaneously. As mentioned above, this kind of evidence is in line with constraint-based approaches to language processing, although other heuristics-based
accounts cannot be ruled out. The evidence does not decide definitively against the
idea that there are two levels of representation, semantic and pragmatic, but only
against the conjunction of that idea and a modular processing architecture according to which all and only Gricean pragmatic processes are set apart from processes
that establish a semantic representation.
Having found evidence that, in some cases, genuine Gricean pragmatic inferences are accessed via what appear to be automatic processing mechanisms, we
can ask to what extent are such inferences operationalised in this way. Gricean
pragmatics explains utterance interpretation as a matter of inferring the speakers
intentions in the utterance situation and much recent research has explored the
question of whether inferences about the speakers intentional state in context can
571
rapidly affect interpretation. Earlier work in this field by Keysar and colleagues
tended to support a strategic egocentrism model according to which first-pass
processes effectively ignored information about the speakers mental state (see
Keysar and Barr 2005). In a typical study, Keysar et al (2000) have participants
play a game where a speaker gives orders to participants to manipulate objects in a
grid and where some objects are obscured from the speaker. In critical items, an instruction like, Pick up the tape is referentially ambiguous if one cannot ignore
items that the speaker is ignorant of, but unambiguous if one only takes the speakers perspective into account. Keysar et al (2000) report severe interference from
the object the speaker cannot see, as predicted if hearers are initially egocentric in
their language processing. Keysar and Barr (2005) argue that the strategic egocentrism strategy is motivated by the apparent difficulty we have in reasoning about
the mental states of others, especially manifest when beliefs diverge from one another (see Apperly and Butterfill 2009).
More recent research into speaker perspective has explored other factors which
may have led to a delay in incorporating speakers perspective. For instance,
Hanna et al (2003) argues that the use of a recording in Keysar et al (2000) is detrimental to the process of monitoring perspective. In Hanna et al (2003), the instructions were given by a confederate who engages the participant in face to face
interaction. The results show much more sensitivity to speaker perspective, virtually without delay. However there is still some small interference. Similar results
are reported in Heller et al (2008) which eliminates a global ambiguity from the
Hanna et al studies that may have encouraged participants to overcome any strategic egocentrism bias. Still better evidence that perspective can be integrated immediately comes from Brown-Schmidt et al (2008) who use a different type of discourse to previous studies. In Keysar et al (2000) and most other perspective
research, the speaker gives orders to a participant. In Brown-Schmidt et als study,
the speaker and hearer are engaged in a joint action and the critical sentences are
questions. That contextual factors such as type of discourse can affect results in
this way supports a middle position between one according to which information
about speakers mental states is completely ignored and one according to which
such information is always automatically incorporated. One version of the middle
position is the Constraint-Based view (Hanna et al 2003, Heller et al 2008, BrownSchmidt et al 2008, among others), according to which cues to mismatched perspective are treated with greater or lesser degrees of certainty and so perspective
information is incorporated only where these cues overcome the very strong constraints provided by the form of the speakers referring expression. One interpretation of this Constraint-Based view is that processing mechanisms are always attempting to integrate the relevant information about the speakers mental state
because interpretation systems are geared toward inferring the speakers intended
meaning. This would be a manifestation of the one-level-of-representation view. A
second interpretation of the C-B view would be to see information from the level of
572
Richard Breheny
speakers meaning interacting to affect a separate semantic level of representation.

Until now, no research has been done to tease these positions apart.
A more modular alternative is offered in Barr (2008) where it is proposed that
while speaker perspective may always be monitored and may immediately affect
anticipation of referents, integration of perspective information with information
from the linguistic input is not automatic. Using a similar design to other studies
mentioned above but a new method of analysis which discriminates anticipation
from integration effects, Barr finds that bottom-up, linguistic information affects
early referential hypotheses independently of whether participants manifest awareness of contrary evidence of speaker perspective. As these studies used a recording
of a speaker and lacked genuine speaker/hearer interactions, and as they involved a
director giving orders, it remains to be seen whether the anticipation-without-integration effect is maintained in more genuinely interactive scenarios across different types of discourse.
Supplementary to research on how a speakers ignorance may affect reference
assignment, recent on-line research has studied the time course of so-called entrainment effects, long documented in off-line interactive research (see Garrod and
Anderson 1987, Brennan and Clark 1996). Entrainment is the phenomenon where
interlocutors fix on a single description for an object or event throughout a conversation. So the same object may be referred to as a metal pipe or silver tube, but once
such a mode of reference is established, subsequent reference among interlocutors
tends to maintain this perspective on the object. The analysis of entrainment
stemming from Brennan and Clark (1996) is that it is the result of a conceptual
pact among interlocutors who (tacitly) agree on a mode of presentation of a referent for the purposes of a conversation. Off-line research has shown that such entrainment effects can last for days (Markman and Makin 1998) and that speakers
do adjust their descriptions according to the knowledge of their audience (Brennan
and Clark 1996).
The off-line research suggests strongly that entrainment is a partner-specific effect but it is still open whether it results from an intentional act by interlocutors, as
Brennan and Clark maintain, or whether some other factors are at play. In an early
on-line study, Metzing and Brennan (2003) had participants play a game with an
experimenter moving objects around a grid. The experimenter (speaker) would fix
on a description for a certain object (silver tube) for three moves and then the experimental conditions manipulate both whether the pact was maintained and who
the speaker was. In measures of visual responses around the grid, it was found that
when the description was changed, significantly more interference was created
with the same speaker than with a new one. However, when descriptions were
maintained, there was no effect of speaker. That is, participants fixed on the correct
referent as rapidly with a new speaker as with the old. While the results of the different description condition support the conceptual pact account, it would be expected that there should be an effect of speaker both in the case where the pact is
573
broken and when it is maintained. Kronmuller and Barr (2007) offer an account of
Metzing and Brennans results according to which the partner-specific effect is the
combined result of an automatic pre-emption of other descriptions due to a link
made in memory between the referent and the description, and a somewhat delayed
application of inference based on pragmatic principles. This account explains the
lack of difference in same-description condition of Metzing and Brennans study
and predicts that the interference effect for the same-speaker in the different-description condition occurs at a delay. This is the result Kronmuller and Barr found
in a visual world study similar to that of Metzing and Brennan. Their conclusion is
that there are two mechanisms at play in the maintenance of descriptions in conversation. However, it is somewhat mysterious why an automatic pre-emption
mechanism would develop, except as a result of something like Gricean pragmatic
principles. So even if Kronmuller and Barr are right, it may still be the case that entrainment is the result of a heuristic that serves joint interaction in discourse. On
the other hand, Brown-Schmidt (2009) argues that the lack of interactivity in Kronmuller and Barrs studies could explain the weakness of their entrainment effects.
Brown-Schmidt employed a slightly different design to the Metzing and Brennan
(2003) study, employing critical items that would encourage entrainment effects
and using more participants. She found both an effect of speaker in the same- and
different-speaker condition. In a second study, she employed the same items but
with a recording rather than a live confederate and found that the effects of speaker
were diminished greatly. The conclusion seems to be that entrainment effects that
are the product of genuine real-time interactions among speakers can occur spontaneously or automatically. Whether, in addition, there exist pre-emption effects,
and if so why they arise, are still open questions.
To sum up this section, we have considered a sample of psycholinguistic research which sheds light on the mechanisms that underpin access to pragmatic inferences. We have seen that there is no strong support for the idea that genuinely
Gricean pragmatic inference processes are somehow set apart from processes that
establish a separate semantic representation. The idea that Gricean inferences
could not automatically or spontaneously be derived because of the complexity of
the mental state reasoning involved also needs to be reconsidered, for two reasons.
Firstly, across a number of different domains, it has been demonstrated that genuine intentional pragmatic inferences are accessed as rapidly as any other information relevant to an interpretation. Secondly, the idea that inferences about
mental states are per se especially difficult compared to any other kind of non-demonstrative inference ought to be re-examined. Indeed, as we will see in the next
section, there is good reason t to think that in the domain of language use and communication, humans develop a keen ability to infer the contents of goal states from
the earliest possible stage.
Given the rapidity of Gricean inference in some cases and the relative open-endedness of their confirmation, the challenge for both psycholinguistics and prag-
574
Richard Breheny
matic theory is to explore the mechanisms that deliver them. We have ruled out the
idea that we only get rapid pragmatic inferences when these can be associated with
particular forms (like some or adjectival modification). In fact we have seen evidence that any ad hoc quantity implicature can be accessed very rapidly. Thus, if
we are to adopt a Constraint-Based model or any other frequency-based model for
pragmatic inference, we need to specify which elements of an utterance situation
are being automatically monitored by hearers. Breheny et al (under review a) suggest that one important place to start is with expectations of the conversational purpose or, more generally, the source of relevance for an utterance. In the case of
quantity implicatures, this could be coupled with a mechanism that monitors alternatives. One way to explore the latter mechanism is via the more general mechanisms that anticipate the completion of an utterance. These are evident in the visualworld research mentioned above (see Altmann and Kamide 1999, Barr 2008) but
much more so in ERP (Event Related Potential) research focussed on the N400 effect.3
Generally speaking, it seems clear that future psycholinguistics research into
the mechanisms that underpin utterance interpretation would benefit from insights
of theoretical pragmatics. At the same time, pragmatic theory (at least the cognitively oriented variety) can greatly benefit from research into on-line processing.
In particular, one important question for recent philosophical and linguistic research is whether there is justification for positing a separate semantic level of representation of the truth conditions of an utterance in context which is independent
of the level at which the speakers intentions are represented. If we treat this theoretical question as also a cognitive question (as do Borg 2004, Carston 2002), then
the type of research reviewed above could be informative to the question. However, while much evidence supports a one-level view, the two-level view cannot be
ruled out given the current state of research.
6.
The development of Gricean pragmatics in language use
Any good theory of pragmatic competence ought not just inform the modelling of
on-line processing in adults but should also be able to explain how the relevant
mechanisms involved develop. This section will review research into the application of intentional pragmatic principles in early language use. We will see that,
while our understanding of child pragmatics has been transformed in the recent
past, the focus has been on getting certain facts right about what, if any, limitations
children have with regards to common pragmatic inferences. The emerging picture
is of children being capable of deriving any pragmatic inferences that adults derive
as early as anyone can ascertain. Where there has been apparent divergence between child responses in experimental studies and adults, this has been partly due
to inadequate methodology. The interesting theoretical question regarding devel-
575
opmental research is how children so rapidly achieve full pragmatic competence

and what lies behind any genuine differences between children and adults regarding the implementation of pragmatic abilities in conversation.
Perhaps in contrast to research into adult comprehension, child language research has, for a long time, focussed its attention on childrens intentional pragmatic abilities. The watchword for most work on childrens syntactic and semantic
abilities has been pragmatic plausibility. For instance, when asking a child whether
she finds a statement about an acted-out situation true or false, a good test for
whether a child is capable of ruling out ungrammatical construals of a string involves scenarios where grammatical construals are false but an ungrammatical
construal is true. For example, in Crain and McKee (1985) experimental scenarios
made He ate pizza and Ninja Turtle danced true only on the ungrammatical construal according to which the turtle ate pizza and danced, while the grammatical
construal is false. The methodological principle of pragmatic plausibility in child
studies is to make the grammatical construal relevant. In this case, by having another character consider eating pizza. The theory is that children in these studies
operate with a principle of charity whereby they are liable to ignore divergences
from proper usage if not doing so does not lead to a plausible construal. Whether
children should operate with such a principle more than adults is something of an
open question. Indeed, as we will see below, the real question now for child pragmatic research is why childrens responses often diverge from adults. Any explanation based on a principle of charity would have to be motivated more fully than
hitherto, while there may be alternative explanations for such divergences.
While early work in syntax and semantics using Truth Value Judgement Tasks
(TVJT) tended to assume some pragmatic competence on the part of toddlers, early
research directly testing childrens access to common pragmatic inferences suggested that pragmatic competence might not be fully developed in some respects.
Here we will focus on childrens understanding of scalar quantity implicatures and
the pragmatic dimension of Principle B (cf. below) of binding theory.
In a landmark paper, Noveck (2001) reports studies using a sentence verification task involving quantificational expressions like All elephants have trunks
or Some elephants have trunks. As mentioned in section 4, adult participants in
these studies tend to favour a response to the some items that is based on an interpretation of the utterance that includes the quantity implicature Not all elephants have trunks. Although numbers vary from study to study, the proportion of
responses that incorporate the implicature rarely falls below two thirds. Noveck
(2001) reports that childrens response rates do not approach that of adults until
7 years in some cases and later with more complex items. Rather, children tend to
favour the logical response, saying that Some elephants have trunks is true. In a
follow-up study, Papafragou and Musolino (2003) used act-out scenarios and a
truth-value judgement task to see if the task demands of sentence verification
(requiring access to long-tem memory) might have been a factor. Indeed, Papafra-
576
Richard Breheny
gou and Musolino (2003) found younger children were performing like adults in
rejecting statements like Some of the horses jumped the fence but still five year
old children showed little sign of having incorporated the implicature into their
understanding of such statements.
In an attempt to determine whether it is the inherent difficulty of the Gricean inference that is at issue, Papafragou and Musolino ran a felicity judgement task
which required the children only to evaluate whether the puppet had described the
scenario well. They found significantly more children were able to evaluate an
underinformative utterance appropriately. The conclusion that could be drawn
from this is that, while children do have expectations of informativity appropriate
to drawing quantity implicatures, something about the TVJT causes the children to
diverge from the adult response pattern. This conclusion is bolstered by the results
of another kind of felicity judgement task (Guasti et al 2005), where children are
given a forced choice between two images, one that depicts a scenario where all of
the characters perform the action and one where some but not all do. In this kind of
study, five year-old children robustly choose the scenario consistent with expectations of informativity. Katsos and Bishop (under review) find that five year old
children succeed at the latter felicity judgement task where the implicature is not
associated with any lexical item, but ad hoc. Kastsos and Bishop also find that
childrens own descriptions of scenarios always respect Gricean expectations of informativeness.
Pouscoulous et al (2007) explore the idea that children do not readily incorporate the implicature into their understanding of the experimental items due to the
task demands resulting from the combination of the metalinguistic nature of the
TVJT and the extra effort required to derive quantity implicatures. In a series of experiments that were aimed at successively lightening task demands, Pouscoulous
et al did find a marked difference in performance. In particular, by changing the
task from truth evaluation to an act-out task, they found that even four-year-old
childrens rate of with-implicature response was above chance. They also found
that childrens with-implicature response rates increased when the items involved
the apparently less complex quantifier quelques rather than the more complex
certains.
One important conclusion from these child studies looking at quantity implicature seems to be that there really is no principled limitation on young childrens
ability to derive quantity implicatures. They seem clearly aware of expectations of
informativeness, they produce appropriately informative utterances and they can
interpret utterances with quantity implicatures. So we could say that the basics of
even four-year-old childrens pragmatic competence is in place. However, even in
the act-out task of Pouscoulous et al (2007) there are still differences across age
groups which look like a trend toward typical adult responses. Pouscoulous et al
(2007), following Noveck (2001), propose that this is due to the cost of the extra inference involved in the enriched interpretation. This view is supported by the adult
577
studies reported in Bott and Noveck (2004) where it is found that adults put under
time pressure tend to respond more often without the implicature. When considering the difference between performance in act out tasks and TVJTs, Katsos and
Bishop (under review) argue that processing cost may not be the whole story but
that children may be more inclined to assent to underinformative utterances in
TVJTs due to the operation of a principle of charity. In light of the fact that even
four-year-old children can derive quantity implicatures, it is clear that they must operate with such a principle, as do adults when faced with a statement that clearly
could be construed as true or false. However, the puzzle with the developmental
trend found in these tasks lies more with adult responses. Note that it is not as
though the adults tend to give a correct response while children deviate from this.
In some respects, the childs logical response is more correct than that of the
adults. If it is true that the principle of charity operates both with adults and
children, then the question really is why adults tend not to see the alternative interpretation. We may be able to find the answer to this question from the adult studies
reviewed above. There we found very rapid access to quantity implicatures in adults
and speculated that adult access to this type of implicature rests on a combination of
automatic, frequency based heuristics. So it may be that what develop between
childhood and adulthood are the heuristic mechanisms that both lighten processing
load and make judgements more certain when it comes to valuation tasks.
The tale of childrens abilities with quantity implicature has played itself out
with only minor modifications when it comes to the development of the so-called
Principle B effects studied by syntacticians. According to the favoured theory of
these effects, due to Reinhart (see Reinhart 1983, 2006), informants judge that the
noun phrases in (8a) cannot be co-referential due to a pragmatic parsimony principle similar to the one we might invoke to account for the contrastive implicatures
studied by Sedivy and colleagues:
8. a. Mama Bear washed her.
b. Mama Bear washed herself.
The idea is that if the speaker wished to express the proposition that Mama Bear
washed Mama Bear, then she would have used the reflexive form as in (8b), as it is
unambiguous. Many child studies of the Principle B phenomena have shown that
children abide by the syntactic principle restricting genuine binding but seem to
fail to appreciate that (8a) should not normally be used to mean (8b) for pragmatic
reasons (see Chien and Wexler 1990, Grodzinsky and Reinhart 1993, Thornton and
Wexler 1999). However, Conroy et al (2009) have challenged certain methodological aspects of these studies and show that 4 year old children are capable of rejecting the pragmatically inappropriate interpretation of (8a) given the right kind of
contextual support. Conroy et al conclude, contra Chien and Wexler (1990) that
children do have the requisite pragmatic abilities to rule out co-reference. But as
with the implicature studies, they do recognise that there is a kind of developmen-
578
Richard Breheny
tal trend. Conroy et al (2009) follow Reinhart (2006) in proposing that the divergence in child responses from the adult norm may have to do with the processing
resources involved. But, again, it may be that the full developmental story turns on
the development of the heuristics which make adults confidently reject the infelicitous utterance.
To sum up this section, we have seen two examples of research on pragmatic
development where, after initial indications to the contrary, it seems children have
full Gricean pragmatic competence as early as anyone can tell. We have also see
that the methods used in child studies have thrown up some intriguing questions as
to why childrens behaviour in some studies deviate from that of adults. Finding
out what develops and how, may shed light on the steady-state adult competence,
while research on the mechanisms that underlie adult processing may in turn shed
light on child pragmatic development.
7.
The ontogenesis of pragmatics Studies in pre-linguistic

communication
One conclusion we may draw from studies of toddlers pragmatic abilities is that as
soon as they have access to the syntactic binding principles, they can rule out pronominal co-reference on pragmatic grounds, and as soon as they grasp the meaning
of indefinite quantifiers like some, they can derive quantity implicatures. We may
then wonder how these pragmatic abilities are ever-present through the development of linguistic abilities. The answer coming from pre-linguistic research seems
to be that it is because infants are fully competent Gricean communicators before
they acquire language.
In a series of interactive pointing production and comprehension tasks, Tomasello and colleagues (see Tomasello 2008, Tomasello et al 2007) have shown that
children treat communicative gestures as involving the specifically reflexive intention first brought to light in Grices work on speaker meaning (see Grice 1957).
That is, not only do infants see communication as involving an intention that another person obtain information but as doing so on the basis of recognising the intention to inform. This overtness condition has been analysed in a number of ways
by philosophers and cognitive scientists but in practice it requires certain cues on
the part of the communicator, such as eye-contact or interactive, contingent behaviour, and certain cues on the part of audience to signal uptake. In elicited production tasks, Liszkowski et al. (2004) elicited infant points to an interesting event
and, in one condition, the adult experimenter simply attended to the event in response, rather than also acknowledge interest to the child. In a baseline condition,
the experimenter both attended to the event and acknowledged uptake with the
child. In the no-uptake condition, as compared to the baseline, children repeated
the communicative act until the experimenter responded appropriately.
579
In a comprehension experiment, Behne et al (2005) engaged infants in a game

where a prize is hidden in one of two containers. In one experimental condition, the
experimenter engages the infant with ostensive cues (particularly, eye contact) and
points to a container. Behne et al found that at 14 months, infants are above chance
in choosing the indicated container, while at 18 months they are virtually at ceiling.
In another condition, the experimenter forms the same hand gesture, fashioned in
the shape of a pointing, but does not engage the child in eye contact or use other ostensive cues. In this condition neither age group favoured the indicated container.
In more recent research, evidence has accumulated that infants understand
communicative acts not simply in terms of what is relevant to them but in terms of
what the communicator would think is relevant (see Liebal et al 2009, and particularly, Southgate et al 2009).
The overall picture is that by the time children approach the one- or two-word
phase, they are already employing sophisticated pragmatic abilities. It now seems
clear that these abilities appear to develop throughout infancy, from around
4 months of age. For instance, Farroni et al (2003) show that 5 month-old infants
tend to follow shifts in eye direction only if the shift is preceded by a period of eye
contact with the infant. Similar research, summarised in Csibra and Gergely (2009)
and Csibra (2010) points to the conclusion that the specifically human mode of
communication involving the recognition of complex intentions develops along
pre-determined pathways prior to the development of fuller social cognitive abilities at the end of the first year. Thus human pragmatic competence may have an innate basis and involve specific neural structures (see Grossman et al 2008). How
this competence relates to more general human social cognitive abilities and to
general linguistic abilities remain intriguing open questions (see Tomasello 2008,
Breheny 2006, Csibra 2010).
8.
Summary
In the past decade, there has been a significant upsurge in experimental research in
the domain of intentional pragmatics, that is, research into our understanding of
speakers meaning based on inferences about intentions in context. This upsurge
has come from different sources, more or less independently. On the one hand, Experimental Pragmatics has attempted to apply experimental methodology to illuminate theoretical debates among linguists.4 On the other hand, psycholinguists
have become increasingly interested in how information about speakers intentions
inform rapid on-line interpretation processes. Meanwhile, there has been a revolution in attitudes towards toddlers linguistic pragmatic abilities; and finally research originating mostly in the infant social cognitive domain has raised the exciting possibility that intentional pragmatic abilities are part of the basic human
makeup.
580
Richard Breheny
Taking this research together, we are by far better informed about intentional
pragmatic abilities now than ten years ago. But there are still very many empirical
questions that remain open. This paper has offered a small sample of recent work
and highlighted some of those open questions.
Notes
* This work was carried out with the support of a grant from the UK Arts and Humanities
Research Council (Ref: AH/E002358/1) and from the Centre for the Study of Mind in Nature (Oslo), Linguistic Agency project.
1. For a general overview of (types of) implicature cf. Huang this volume.
2. Although it is argued in Saul (2002) that Grices account of conversational implicature
should be understood to be a normative account, there is very little textual evidence for
this and far more evidence that Grice would have wished to remain neutral on this issue
(see Petrus 2010).
3. See van Berkum in press for an overview; N400 refers to a negativity that appears in an
ERP reading 400 ms after the offset of a stimulus. It is widely thought of as an indicator of
the processing of meaning. Larger N400 suggests extra semantic processing work (as
when there is semantic anomaly).
4. Here we have reviewed work focussed on quantity implicatures, but there is other work
on metaphor (see Glucksberg 2003, Rubio-Fernandez 2007), presupposition (see
Schwarz 2007, Chemla and Bott 2010) and other areas of pragmatics (see Noveck and
Sperber 2004 for an early collection).
References
Altmann, Gerry T.M. and Yuki Kamide
1999
Incremental interpretation at verbs: restricting the domain of subsequent reference. Cognition 73: 247264.
2007
The real-time mediation of visual attention by language and world knowledge:
Linking anticipatory (and other) eye movements to linguistic processing. Journal of Memory and Language 57: 502518.
Apperly, Ian A. and Stephen A. Butterfill
2009
Do humans have two systems to track beliefs and belief-like states? Psychological Review 116(4): 953970.
Atlas, Jay and Stephen C. Levinson
1981
It-clefts, informativeness and logical form: radical pragmatics (revised standard version). In: Peter Cole (ed.), Radical pragmatics, 161. New York: Academic Press.
Barr, Dale
2008
Pragmatic expectations and linguistic evidence: Listeners anticipate but do not
integrate common ground. Cognition 109: 1840.
581
Behne, Tanya, Malinda Carpenter and Michael Tomasello

2005
One-year-olds comprehend the communicative intentions behind gestures in a
hiding game. Developmental Science 8: 492499.
Berkum, Jos J.A. van
in press The neuropragmatics of simple utterance comprehension: An ERP review.
To appear in: Ulrich Sauerland and Kazuko Yatsushiro (eds.), Semantic and
pragmatics: From experiment to theory. New York: Palgrave.
Borg, Emma
2004
Bott, Lewis and Ira A. Noveck
2004
Some Utterances are Underinformative: The Onset and Time Course of Scalar
Inferences. Journal of Memory and Language 51(3): 437457.
Breheny, Richard
2006
Communication and folk psychology. Mind and Language 21(1): 74107.
Breheny, Richard, Napoleon Katsos and John Williams
2006
Are Scalar Implicatures Generated by Default? Cognition 100: 434463.
Breheny, Richard, Heather J. Ferguson and Napoleon Katsos
under review, a How quantity implicatures are accessed in incremental utterance interpretation: Evidence for full-blown Gricean inferencing.
under review, b On the timecourse of access of conversational implicatures in incremental interpretation.
1996
Conceptual pacts and lexical choice in conversation. Journal of Experimental
Psychology: Learning, Memory, and Cognition 22: 14821493.
Brown-Schmidt, Susan
2009
Partner-specific interpretation of maintained referential precedents during interactive dialog. Journal of Memory and Language 61: 171190.
Brown-Schmidt, Susan, Christine Gunlogson and Michael K. Tanenhaus
2008
Addressees distinguish shared from private information when interpreting
questions during interactive conversation. Cognition 107: 11221134.
Carston, Robyn
2002
Thoughts and utterances. Oxford: Blackwells.
Chemla, Emmanuel and Lewis Bott
2010
Processing presuppositions: dynamic semantics vs pragmatic enrichment. Ms
Institut Jean Nicod (http://www.emmanuel.chemla.free.fr/Material/ChemlaBott-ProcPresProj.pdf)
Chien, Yu-Chin, and Kenneth Wexler
1990
Childrens knowledge of locality conditions in binding as evidence for the
modularity of syntax and pragmatics. Language Acquisition 1: 225295.
Chierchia, Gennaro
2004
Scalar implicatures, polarity phenomena and the syntax/pragmatics interface.
In: Adriana Belletti (ed.), Structures and beyond, 39103. Oxford: Oxford
University Press.
Clark, Herbert H.
1996
582
Richard Breheny
Conroy, Anastasia, Eri Takahashi, Jeff Lidz and Colin Phillips

2009
Equal Treatment for All Antecedents: How Children Succeed with Principle B.
Linguistic Inquiry 45: 446486.
Cooper, Roger M.
1974
The control of eye fixation by the meaning of spoken language: A new methodology for the real-time investigation of speech perception, memory, and language processing. Cognitive Psychology 6: 84107.
Crain, Stephen and Cecile McKee
1985
The acquisition of structural restrictions on anaphora. In: Proceedings of
NELS 16. Ed. by Stephen Berman, Jaewoong Choe and Joyce McDonough,
94110. Amherst: University of Massachusetts, GLSA.
Csibra, Gergely
2010
Recognizing communicative intentions in infancy. Mind and Language 25:
141168.
Csibra, Gergely and Gyorgy Gergely
2009
Natural Pedagogy. Trends in Cognitive Science 13: 148153.
Farroni, Teresa, Eileen M. Mansfield, Carlo Lai and Mark H. Johnson
2003
Infants perceiving and acting on the eyes: Tests of an evolutionary hypothesis.
Journal of Experimental Child Psychology 85: 199212.
Garrod, Simon and Anthony Anderson
1987
Saying what you mean in dialog: A study in conceptual and semantic co-ordination. Cognition 27: 181218.
Gazdar, Gerald
1979
Geurts, Bart
2010
Quantity Implicature. Cambridge: Cambridge University Press.
Geurts, Bart and Nausicaa Pouscoulous
2009
No scalar inferences under embedding. Semantics and Pragmatics 2(4): 134.
Glucksberg, Sam
2003
The psycholinguistics of metaphor. Trends in Cognitive Sciences 7: 926.
Graesser, Arthur C., Morton Ann Gernsbacher Susan R. Goldman
2003
Handbook of discourse processes. Mahwah, NJ: Lawrence Erlbaum.
Grice, H. Paul
1957
Meaning. Philosophical review 66: 377388. Reprinted in Grice (1989,
213223).
Grice, H. Paul
1989
Studies in the way of words. Cambridge, MA: Harvard University Press.
Grice, H. Paul
2001
Aspects of Reason. Oxford: Oxford University Press.
Grodner, Daniel and Julie Sedivy
in press The effects of speaker-specific information on pragmatic inferences. In: Neal J.
Pearlmutter and Edward Gibson (eds.), The Processing and Acquisition of Reference, Cambridge, MA: MIT Press.
Grodner, Daniel J., Natalie M. Klein, Kathleen M. Carbary and Michael K. Tanenhaus
2010
Some, and possibly all, scalar inferences are not delayed: Evidence for immediate pragmatic enrichment. Cognition.
583
Grodzinsky, Yosef and Tanya Reinhart

1993
The innateness of binding and coreference. Linguistic Inquiry 24: 69101.
Grossmann, Tobias, Mark H. Johnson, Sarah Lloyd-Fox, Anna Blasi, Fanni Deligianni,
Clare Elwell and Gergely Csibra
2008
Early cortical specialization for face-to-face communication in human infants.
Proceedings of the Royal Society, London B, 275: 28032811.
Guasti, Maria Teresa, Gennaro Chierchia, Stephen Crain, Francesca Foppolo, Andrea Gualmini and Luisa Meroni
2005
Why children and adults sometimes (but not always) compute implicatures.
Language and Cognitive Processes 20(5): 667696.
Hanna, Joy, Michael K. Tanenhaus and John C. Trueswell
2003
The effects of common ground and perspective on domains of referential interpretation. Journal of Memory and Language 49: 4361.
Heller Daphna, Daniel Grodner and Michael K. Tanenhaus
2008
The role of perspective in identifying domains of reference. Cognition 108:
831836.
Horn, Laurence
1984
A new taxonomy for pragmatic inference: Q-based and R-based implicature.
In: Deborah Schiffrin (ed.), Meaning, Form and Use in Context, 1142.
(GURT 84). Washington: Georgetown University Press.
Horn, Laurence
1989
A Natural History of Negation. Chicago, IL: University of Chicago Press.
Horn, Laurence
2006
The border wars: a neo-Gricean perspective. In: Klaus von Heusinger and Ken
Turner (eds.), Where Semantics Meets Pragmatics, 2148. Amsterdam: Elsevier.
Huang, Yi Ting, and Jesse Snedeker
2009
On-line Interpretation of Scalar Quantifiers: Insight into the Semantic-Pragmatics Interface. Cognitive Psychology 58: 376415.
Katsos, Napoleon and Dorothy Bishop
under review The development of informativeness from the speakers and the comprehenders perspective.
Keysar, Boaz, Dale J. Barr, Jennifer A. Balin and Jason S. Brauner
2000
Taking perspective in conversation: The role of mutual knowledge in comprehension. Psychological Sciences 11: 3238.
Keysar, Boaz, and Dale J. Barr
2005
Coordination of action and belief in communication. In: John C. Trueswell and
Michael K. Tanenhaus (eds.), Approaches to Studying World Situated Language Use: Bridging the Language-as-Product and Language-as-Action,
Kronmuller, Emundo and Dale J. Barr
2007
Perspective-free pragmatics: Broken precedents and the recovery-from-preemption hypothesis. Journal of Memory and Language 56: 436455.
Levinson, Stephen
2000
Presumptive Meanings. Cambridge, MA: MIT Press.
Liebal, Kristin, Tanya Behne, Malinda Carpenter and Michael Tomasello
2009
Infants use shared experience to interpret pointing gestures. Developmental
Science 12(2): 264271.
584
Richard Breheny
Liszkowski, Ulf, Malinda Carpenter, Anne Henning, Tricia Striano and Michael Tomasello
2004
12-month-olds point to share attention and interest. Developmental Science 7:
297307.
MacDonald, Maryellen C. and Mark S. Seidenberg
2006
Constraint satisfaction accounts of lexical and sentence comprehension. In:
Matthew J. Traxler. and Morton Ann Gernsbacher (eds.), Handbook of Psycholinguistics. 2nd ed., 581611. London: Elsevier.
1992
Inference during reading. Psychological Review 99: 440466.
Markman, Arthur B. and Valerie S. Makin
1998
Referential communication and category acquisition. Journal of Experimental
Psychology: General 127: 331354.
Metzing, Charles, and Susan E. Brennan
2003
When conceptual pacts are broken: partner-specific effects on the comprehension of referring expressions. Journal of Memory and Language 49: 201
213.
Noveck, Ira A.
2001
When children are more logical than adults: Investigations of scalar implicature. Cognition 78/2: 165188.
Noveck, Ira, and Andres Posada
2003
Characterising the time course of an implicature. Brain and Language 85:
203210.
2004
Experimental Pragmatics. New York: Palgrave.
Papafragou, Anna and Julien Musolino
2003
Scalar implicatures: Experiments at the Semantics-Pragmatics interface. Cognition 86: 253282.
Petrus, Klaus
2010
Introduction: Paul Grice, Philosopher of Language, but more than that.
In: Klaus Petrus (ed.), Meaning and Analysis: New Essays on H. Paul Grice,
130. New York: Palgrave.
Pouscoulous, Nausicaa, Ira A. Noveck, Guy Politzer, and Anne Bastide
2007
A developmental investigation of processing costs in implicature production.
Language acquisition 14: 347376.
Reinhart, Tanya
1983
Anaphora and semantic interpretation. London: Croom Helm.
Reinhart, Tanya
2006
Interface strategies. Cambridge, MA: MIT Press.
Rubio-Fernandez, Paula
2007
Suppression in Metaphor Interpretation: Differences between Meaning Selection and Meaning Construction. Journal of Semantics 24: 345371.
Saul, Jennifer
2002
Speaker meaning, what is said and what is implicated. Nous 36: 228248.
Sauerland, Ulrich
2004
367391.
585
Sedivy, Julie C.
2003
Pragmatic versus form-based accounts of referential contrast: Evidence for effects of informativity expectations. Journal of Psycholinguistic Research 32:
323.
Sedivy, Julie C., Michael K. Tanenhaus, Craig G. Chambers and Gregory N. Carlson
1999
Achieving incremental semantic interpretation through contextual representation. Cognition 71: 109147.
Schwarz, Florian
2007
Processing Presupposed Content. Journal of Semantics 24(4): 373416.
Southgate, Victoria, Coralie Chevallier and Gergely Csibra
2009
Sensitivity to communicative relevance tells young children what to imitate.
Developmental Science 12(6): 10131019.
1986
1995
Postface to 2nd Edition. In: Dan Sperber and Deidre Wilson, Relevance: Communication and cognition, pl. add+ rede. 2nd ed. Oxford: Blackwell.
Tanenhaus, Michael K., Michael J. Spivey-Knowlton, Kathleen M. Eberhard and Julie C.
Sedivy
1995
Integration of visual and linguistic information during spoken language comprehension. Science 268: 16321634.
Thornton, Rosalind and Kenneth Wexler
1999
Principle B, VP-ellipsis, and interpretation in child grammar. Cambridge,
MA: MIT Press.
Tomasello, Michael
2008
Origins of Human Communication. Cambridge, MA: MIT Press.
Tomasello, Michael, Malinda Carpenter and Ulf Lizskowski
2007
A new look at infant pointing. Child Development 78: 70522.
Warner, Richard
2001
Introduction: Grice on reasons and rationality. In: H. Paul Grice (2001),
viixxxviii.
21.
Corpus-based pragmatics I: qualitative studies1

Gisle Andersen
1.
Contextualising corpus-based pragmatics
This paper concerns the use of computational methods for collection and analysis
of language data as a basis for research in pragmatics. The aim is to describe advantages and problems associated with the use of corpora, and to provide both a
birds-eye view of the field as well as an in-depth account of a limited set of topics
of particular salience. A main claim to be argued here, with reference to several
papers, is that, as a method for empirical research, corpus-based pragmatic analysis
is essentially inductive and interpretative (Mosegaard Hansen 1998: 214) and
can never be performed without the reliance on the analysts intuition (cf. especially section 3). Corpus-based studies are conceptually related to other ways of
doing empirical pragmatics, such as anthropological or ethnological field work, sociolinguistic interviews and the use of experimental techniques and elicitation tests
(Chafe 1992). The insights offered by the studies reviewed in the current survey
have emerged via meticulous analysis of corpus data, as opposed to other empirical
methods. The studies discussed here are further distinguished from those discussed
in Christoph Rhlemanns contribution to the current volume, which focuses on
quantitative studies. However, corpus-based studies are rarely exclusively qualitative or quantitative but commonly describe pragmatic functions as well as quantifiable distributional aspects, involving for instance comparisons between users
and user groups within or across corpora. The current study, then, describes that
body of pragmatics literature that has its main focus on functional aspects but focuses less on measurable distributional differences.
The term discourse linguistics has been used as an umbrella term for discourse analysis, discourse studies, text linguistics, pragmatics, conversation analysis and other related approaches to the study of discourse (Virtanen 2009: 49), a
list of topics which is broadly consistent with the topics to be reviewed in this
chapter. The application of corpus-linguistic method in this knowledge field has
been relatively limited, especially compared to its application in the study of more
formal topics such as syntax, morphology and lexis (cf. McEnery and Wilson
2001; Kennedy 1998). Explorations of corpora have nevertheless been fruitful in
pragmatics, as evidenced by a series of studies which investigate how interaction is
sequentially and interactionally organised, how the attitudes and perspectives of
speakers and hearers are manifested in spoken discourse, how texts are coherently
structured beyond the sentence level, etc. Among the topics studied are turn-taking
phenomena, pauses and repairs, discourse markers, hedging and vague language,
588
Gisle Andersen
small talk, speaker attitudes and so on. (cf. section 2). Despite the variety of topics,
there are many aspects of pragmatics that have not been subjected to corpus-based
investigations to the extent they could have been, and I shall argue in favour of an
extended application of corpora in pragmatics, to capture a wider range of phenomena within the cognitive, social and cultural science of language and communication (Verschueren 2009: 1).
Corpus linguistics and pragmatics can be seen as two related, but historically
distinct research traditions. I endorse the generally accepted view expressed by
corpus linguists such as Leech (1992), Kennedy (1998), McEnery and Wilson
(2001) and Virtanen (2009), that pragmatics and corpus linguistics are ontologically different academic pursuits, in that the former is a branch of linguistics albeit many-faceted and complex, while the latter is not. Rather, corpus linguistics is
regarded as a certain way of performing empirical studies in linguistics, a research approach (Biber 1988) and a methodological basis for doing linguistic
research (Leech 1992: 105; cf. also Kennedy 1998: 7). Empirical pragmatics crucially depends on authentic language data, and corpus linguistics is one methodological approach which may be used for pragmatics research. Like other kinds of
empirical language data, corpora are documentations of the choices made by language users and surface manifestations of the underlying communicative competence of the speakers whose language the corpus represents (de Beaugrande 1999;
cf. also section 3). In the words of Chafe (1992:88), corpora provide a particularly
useful means of access to the language-related capabilities of the mind.
A notable recent development in corpus linguistics has been the increased
focus on corpus-driven research (as opposed to corpus-based research; TogniniBonelli 2001: 8487; Sinclair 2004), which differs from the standard practice of
linguistics in that it makes minimal a priori assumptions regarding the linguistic
constructs that should be employed for the analysis (Biber 2009: 276). In corpusbased research, the researcher approaches the data equipped with a set assumptions
as to its linguistic structure and expected findings. Corpus-driven research, on the
other hand, explores linguistic structures that emerge from the corpus, thus inductively exploiting the potential of a corpus to identify linguistic categories and
units that have not been previously recognised (Biber 2009: 278). One of the main
claims that could be made about pragmatics research using corpora is that it is by
and large corpus-based and not corpus-driven (Sandford 2008). Most studies
have as their starting point a previously recognised structure, where the main value
of the empirical work results from the investigators meticulous analysis of the
range of different functions associated with particular linguistic forms, as evidenced by the naturally occurring discourse contained in the corpus. Examples
would be a structure like the it-cleft in English (e.g. Collins 2005) or an individual
word form like really; see e.g. Stenstrm (1986b), who provides a detailed account
of this form and its various functions in the London-Lund Corpus. The corpus
method principally involves inspecting data via a search term that yields relevant
589
results, i.e. the method known as one-to-one searching (del and Reppen 2008:
2). Often, however, the inspection of extracted corpus data may well result in interesting observations of functional categories not previously accounted for or
even recognised, such as the observation made by Andersen (2001: 139ff) that the
invariant British English form innit (from isnt it or aint it) has an extended function as a marker of mutual manifestness (common ground) directed towards the
previous speakers utterance, in addition to its more generally recognised use as a
tag question which modifies a proposition of the current speaker. Most corpusbased pragmatics studies are thus form-based. The computerised corpus lends
itself well to this kind of approach, in which relevant instances of a type can be
identified as tokens in actual use.
The methodological counterpart is the function-based approach, which takes as
its basis a particular pragmatic function and describes its possible realisations in
actual discourse. Functions that have been investigated include conversational
openings and closing, topic transition, making a request, expressing disagreement
or the like. An example of a study where this method is successfully applied is
Mauranens (2004) study of hedges, items which make propositions more or less
fuzzy and add indeterminacy of one kind or another to an utterance (Mauranen
2004: 194). She shows that in a contextually constrained corpus, the Michigan
Corpus of Academic Spoken English (MICASE), this function is typically performed by a category of items kind of, sort of, or something, or so, somewhat, just,
and a little bit, items which share some functional characteristics as hedges but
which perform specialised functions in different contexts of use.
In both the form and function-based approaches, the role of the corpus is essentially to provide a repository of linguistic examples to be accounted for. Until recently, it has been less customary to carry out corpus-driven studies in pragmatics,
despite the fact that this approach has been applied widely and successfully in related fields, particularly phraseology and lexicography (cf. Sinclair 1991; Partington 1998; McEnery et al. 2006; Biber 2009). Notably, the corpus-based and corpus-driven approaches should be seen as complementary rather than conflicting.
As Biber argues, corpus-driven research is not in any way superior to corpusbased research. The two require radically different methods, and thus they offer the
possibility of uncovering radically different perspectives on language structure and
use (Biber 2009: 279). Studies in pragmatics are likely to gain from using corpora
in manners which go beyond the dominating pattern of one-to-one search used in
corpus-based studies. That this combination may be fruitful can be demonstrated
by a recent paper by Walsh et al. (2008). Their study of vagueness markers in academic discourse is essentially a synthesis of two methodologies not always seen
as complementary, corpus linguistics and conversation analysis. These have much
to offer each other as they provide both quantitative and qualitative insights respectively (Walsh et al. 2008: 16). Using this combinatory approach, the authors
show that vague language varies considerably according to mode of discourse
590
Gisle Andersen
(managerial, classroom, conversational) and that these qualitative differences are

a valuable means of understanding the complex relationship between language and
learning (Walsh et al. 2008:9). Furthermore, discourse functions are often associated with processes of language change, involving the emergence of new discourse
items, realised either as new forms, such as the English phrase yeah right used as
an irony marker in Norwegian (Andersen 2006), or as new uses of old forms, such
as the new quotative marker BE like (Tagliamonte and Hudson 1999; Andersen
2001; Rhlemann, this volume). These cases exemplify processes of change (borrowing vs. structural reanalysis) that are observable using corpus-driven methods.
Studies in pragmatics should investigate changes in vocabulary and in collocational and colligational patterns, with a view to revealing new ways of organising discourse, structural change, categorical reanalysis and emergent grammaticalisation (Traugott and Heine 1991; Lindquist and Mair 2004). In other words,
there are good reasons for complementing the well-tested and approved method of
corpus-based pragmatics with a corpus-driven approach.
The current chapter focuses on studies where corpora have been applied successfully in pragmatics research. The chapter is organised as follows. Section 2
provides an historical overview of the field of corpus-based pragmatics, giving a
brief description of a wide range of studies in order to show the breadth of this
field. Section 3 concerns method and has the two-fold purpose of arguing in favour
of a corpus-based approach as well as problematising the hermeneutical aspects
of this method. Section 4 provides an in-depth account some particularly salient
topics and contributors in corpus-based pragmatics. These include discourse
markers, prosody, listener activities (backchannels) and corpus-based models of
discourse structure. The final section reiterates some of the main claims and observations made in the previous sections.
2.
An overview of the field
Corpora are compilations of naturally occurring spoken or written language that

can be accessed on a computer. Such compilations may be monolingual or multilingual and may represent general language or specific domains (professional/academic corpora). This study will show that corpora form a well-prepared basis for
systematic, descriptive studies of instances of actual speech, of language variation
and of how social context constrains communicative practices. Spoken corpora
have a favoured position in pragmatics. A spoken corpus is a collection of speech
data made accessible via a computer, containing at least transcriptions of speech
but often also audio and/or video files (for problems and types of transcription, cf.
Roger J. Kreuz and Monica A. Riordan, this volume).
Corpora have been available for linguistic research from the 1960s. Since the
early days of corpus linguistics, there has been an interest in using corpora as a
591
means of exploring functional and contextual aspects of language use. However,

following the compilation of the first-generation written corpora, the Brown University Corpus (Francis 1967) and the Lancaster-Oslo/Bergen corpus (Johansson
and Hofland 1994), the bulk of studies were within other fields of linguistics, particularly syntax, semantics, morphology and lexicology (cf. Altenberg 1986 for
an overview). Though few and far between, early pragmatics studies include linguistic stylistics (Enkvist 1973) and aspects of text linguistics, addressing such issues as information structure, fronting and topicalisation (Gustafsson 1982, 1983,
1985; Davidson 1977). The parallel structure and content of the Brown and LOB
corpora enabled reliable comparative studies of American and British English that
focused on lexical and grammatical differences (e.g. Hofland and Johansson 1986;
Algeo 1988).
The ground-breaking compilation of the first spoken corpus, the London Lund
Corpus of Spoken English (LLC, Svartvik and Quirk 1980), triggered a range of
studies in pragmatics (Altenberg 1986). The corpus was based on recordings of
spoken data from the Survey of English usage (SEU), which in its pre-computerised version had resulted in Crystals detailed empirical work on intonation (Crystal 1972, 1975; Crystal and Quirk 1964), Haegemans (1984) empirically based account of pragmatic conditionals like the story, if so it may be termed, is weak and
loose and Kussmauls (1978) early study of Indikatoren von Sprechakten (speech
act markers). LLC inspired research into phenomena relating to speech production
and performance, such as slips of the tongue (Garnham et al. 1981), repairs (Hedstrm 1984) and hesitation (Orestrm 1977).
Several corpus-based studies were collected in what must be seen as a seminal
volume on conversational discourse, namely Enkvist (1982). This volume addresses an array of phenomena relating to the study of speech, including turn-taking, feedback and repair (Faerch and Kasper 1982; Orestrm 1982; Stenstrm
1982), intonation and prominence (Brazil 1982), pragmatic particles (stman
1982), referential expressions (Thavenius 1982) and functional aspects of negation
(Tottie and Paradis 1982). The book contains pioneering work on theoretical and
methodological topics such as transcription, segmentation and annotation of conversational data (Loman 1982; Svartvik 1982) and the validity of contrastive
methods (Sajavaara 1982). Also notable is the variety of languages represented by
the corpus data described in this volume (English, French, German, Hebrew, Russian and Swedish).
A topic which received early attention is discourse markers (dealt with more
extensively in section 4). The early contributions include Ermans work on pragmatic expressions like you know, you see and I mean (1986, 1987), Stenstrms
work on really (1986b) and Svartviks investigation on the various functions of
well (1980). Moreover, Stenstrms (1982, 1984) extensive work on question/response sequences (cf. section 4) dates from this period, as do some accounts that
deal specifically with tag questions (Nsslin 1984; Bublitz 1979). The availability
592
Gisle Andersen
of both spoken and written corpora also inspired comparative work, particularly
Bibers (1986, 1987, 1988) work focusing on the lexical, grammatical and pragmatic differences between these two modes of communication. In addition, LLC
formed the basis of an at least partly corpus-based account of conversational phenomena, Wardhaugh (1985), which has been criticised for its superficiality and
lack of an analytical apparatus and technical terminology (Garnham 1986).
A second wave of corpus-based studies came with the development of additional English language corpora, such as the Kolhapur Corpus of Indian English
(Shastri 1988), the Birmingham Collection of English Texts (Renouf 1986; Sinclair 1987), the Tosca Corpus (van den Heuvel 1988), The Nijmegen Corpus (Aarts
and Meijs 1986) and The Helsinki Corpus of English Texts, representing historical
English (Rissanen 1986; Kyt and Rissanen 1988). Moreover, the introduction of
formal annotation schemes for intonation in speech corpora was a timely and relevant innovation that became highly influential to pragmatics research (Svartvik
1987, 1988; Knowles 1986; Knowles and Lawrence 1987; Taylor and Knowles
1988; Knowles et al. 1992). This sparked research on the pragmatic aspects of
prosody (Altenberg 1987; Nevalainen 1987a) and studies exploring the correlation
between intonation and type of speech act (Geluykens 1986, 1987a, 1988) and between intonation and information structure (Nevalainen 1987b). The annotated
corpora enabled studies of the functional characteristics of pauses and other types
of prosodic segmentation (Stenstrm 1986a, 1988). The range of pragmatic phenomena accounted for increased in the 1980s and came to include investigations of
dislocation (Geluykens 1987b; Aijmer 1989), evidentiality and epistemic modality
(Westney 1986; Holmes 1988; Coates 1987; Biber and Finegan 1989), turn-taking
(Brodda 1988) and direct vs. indirect speech (Aijmer 1986).
There has been a gradual increase in the range of domains of language use that
are represented by corpora, and especially in the period from the 1990s, the growth
in size and coverage has been remarkable. The second generation of corpora includes large, general-purpose corpora like the 100 million word BNC corpus
(Crowdy 1995; Aston and Burnard 1998) and the 20 million word Birmingham
Corpus. A notable coordinated effort is the International Corpus of English (ICE,
Greenbaum 1991, 1992, 1996; Banjo 1996), which has been used for cross-varietal
studies of the English spoken in many Anglophone societies. The Santa Barbara
Corpus of Spoken American English (CSAE, Chafe et al. 1991) constitutes the
American spoken component of ICE. The family of corpora also came to include
corpora for specific purposes, such as foreign learner corpora like the International
Corpus of Learner English (ICLE, Altenberg 1997; Granger 1997) specialised LSP
corpora (Faber and Lauridsen 1991; Fang 1993) such as corpora of doctor-patient
interactions (Thomas and Wilson 1996), and parallel corpora such as the EnglishNorwegian Parallel Corpus (ENPC, Johansson and Hofland 1994) and the EnglishSwedish Parallel Corpus (ESPC, Altenberg and Aijmer 2000). Finally, corpora that
focus on regionally and socially restricted varieties of English include the Bergen
593
Corpus of London Teenage language (COLT, Haslerud and Stenstrm 1995; Stenstrm et al. 1988).
The growth in availability of corpora of various kinds has led to a growth in
corpus-based studies in pragmatics, although the relative growth has been larger in
other fields of linguistics. Several important contributions can be mentioned. With
respect to contrastive studies, Granger and Tyson (1996) and Granger (1997) compare discourse aspects of how native and non-native writers use connectives and
participle clauses in English academic writing. Work on phonology and intonation
advanced greatly through efforts in the SEC/MARSEC corpora (Bald 1990; Bauer
and Holmes 1996; Knowles et al. 1996; Wichmann 1996; Nevalainen 1992). Finally, studies in pragmatics came to cover an ever wider range of topics, including
amplifiers (Altenberg 1991), thanking, apologies and requests (Aijmer 1996), register (Biber 1993, 1994, 1995), conversational humour (Fill 1992), expletives and
taboos (Stenstrm 1991, 1995, 2006) and backchannels and response signals (Bublitz 1988; Tottie 1991).
In its initial phase, corpus linguistics was practically synonymous with English
corpus linguistics. The establishment of the ICAME (International Computer
Archive of Modern and Medieval English) organisation in 1979, and its coordinated development and distribution of many English language corpora, have provided a particularly fruitful environment for corpus-based research. Recently,
however, corpora for a number of other languages have been compiled, including
large, national corpus-building initiatives such as the Croatian National Corpus,
the Czech National Corpus, the Spoken Dutch Corpus, the Gothenburg Spoken
Language Corpus (Swedish), and a number of other, more targeted corpora developed in connection with specific research projects, especially in sociolinguistics
(Andersen 2010). For example, de Haan (2002) uses the Uppsala corpus of contemporary Russian texts to explore different means of expressing modality and negation in Russian and the ambiguity of different markers of modality. Similarly,
Schelfhout et al. (2005) explore the Dutch National Corpus, observing that speakers use of interjections varies significantly with textual genre as well as with
grammatical, functional and prosodic factors of the utterances in which they o
ccur. Among the international corpora available, the C-ORAL-ROM project has
launched comparable spoken corpora in four Romance languages, French, Italian,
Portuguese and Spanish (Cresti and Moneglia 2005). A study like Schneider
(2007) draws on no less than 22 corpora of spoken French, Italian and Spanish in a
comprehensive comparative account of parenthetical clauses like je crois, je dirais,
je pense, je sais pas and je trouve, and its equivalents in the other Romance languages. The Spanish Corpus Oral de Lenguaje Adolescente (COLA) enables comparative studies of the varieties of Spanish used by adolescents in major Hispanophone cities, and has been explored extensively in pragmatics studies (Stenstrm
and Jrgensen 2009). In contrast, the Corpus de Referencia del Espanol Actual
(CREA) represents Spanish spoken by adults. Downing (2006) explores CREA in
594
Gisle Andersen
a contrastive study of the stance marker surely and its Spanish cognates. Furthermore, the Italian spoken corpus Lessico dellitaliano parlato (LIP, De Mauro et
al. 1993) forms the basis of Waltereits (2006) study on the role of speaker motivations in the sense of abstractions of communicative intentions (Waltereit 2006:
62) in the grammaticalisation from lexical items to discourse markers of Italian
forms such as insomma and diciamo. The Lancaster Corpus of Mandarin Chinese
(McEnery and Xiao 2004) is explored by Ming and Chen (2010) in a recent study
of how the discourse function of noun phrases affects the word order in noun
phrases with a relative clause. Factors such as the information status, animacy, and
high versus low discourse salience of head NPs have a bearing on whether a Mandarin Chinese relative clause immediately precedes the head NP or whether it is
separated from its head by a numeral-classifier sequence consisting of yi (one)
and a classifier. Noun phrases representing abstract entities with a low discourse
salience are contrasted with concrete (animate or inanimate) entities with a high
discourse salience, and the relative order of the components are explained with reference to the principle of relevance (Sperber and Wilson 1995), as it seems to indicate the direction in which the relevance is to be sought (Ming and Chen 2010:
188). Used with a fully realised and highlighted discourse function of introducing a
salient participant, the numeral classifier tends to precede the relative clause;
otherwise it tends to follow the relative clause. The authors argue that this shows
that clause level structural analysis does not suffice to explain the underlying factors that govern the choice of alternative means of expressions in a language. By
contrast, corpus data can be tremendously useful in helping us arriving at a realistic
account of the structure of human language (Ming and Chen 2010: 188).
Despite notable studies like the ones mentioned, there is still a considerable
bias towards English-based studies in pragmatics in the literature (McCarthy and
Carter 2004b).2 Importantly, however, it should be pointed out that the availability
of corpora representing different languages, including parallel corpora consisting
of original and translated texts, has triggered significant comparative work. A recent volume that reports a number of multilingual studies is Aijmer and SimonVandenbergen (2006). The authors argue that comparative work is interesting for
its potential to lead to new knowledge about what is universal and what is language-specific (Aijmer and Simon-Vandenbergen 2006: 3), for its potential in investigating etymological or semantic cognates, and for revealing the effects of
translation on pragmatic aspects. These general claims are substantiated by the individual papers in the volume, such as Johansson (2006) and Hasselgrd (2006),
who jointly demonstrate the applicability of parallel corpora in determining the
meaning and distribution of the Norwegian English cognate words vel/well and n/
now, respectively. Moreover, Altenbergs (2006) account of adverbial connectors
shows how cross-linguistic similarities are largely discourse-driven. In the languages compared, Swedish and English, contrastive and exemplifying connectors
often serve to highlight the initial element, syntactically and prosodically, thereby
595
indicating a significant break or shift in the discourse. Hence, despite the syntactic
differences between English and Swedish, there are interesting parallels between
the two languages in the way that connectors adjust to, and signal, the progression
of discourse (Altenberg 2006: 35f). Similarly, Foolen (2006: 70) compares Dutch
toch and German doch, showing that the multilingual approach makes it clear
more than a monolingual study could have done, that the language specific uses of
the same item are indeed rather specific, but at the same time that in all these applications a single schematic meaning was preserved. The distributional differences of the two items are well accounted for; e.g., judging by translators choices
when handling doch/toch as an adversative connective, [a]pparently, German does
not like a bare doch in first sentence position (Foolen 2006: 65). This might be
explained by German dochs status as a conjunction. As modal particles, only toch
can occur in sentence-final tag position, where it functions as a checking device,
inviting an affirmative response from the hearer. As a whole, the collective volume
shows that the contrastive corpus-based approach provides a new perspective on
fundamental properties of discourse markers, such as their multifunctionality and
core meanings.
The most recent advances in corpus linguistics are connected with increased
data processing and storage capacity, the emergence of web-based corpora, and the
(semi-) automatic compilation of large monitor corpora which are constantly updated with new texts representing an unbroken, chronological data flow (Renouf
2007: 36). These recent advances have changed the face of corpus linguistics
(Renouf and Kehoe 2006: 3) to a much richer cycle of events, consisting not just
of small, medium and large primary corpus building, but of specialised and multidimensional secondary corpus building; not just of corpus analysis but also of corpus evaluation; not just of an initial application of theory but of self-reflection and
a new concern with theory in the light of experience (Renouf and Kehoe 2006: 3).
The tangible outputs of these processes are super-corpora containing hundreds of
millions of words, like the Bank of English at the University of Birmingham, The
Corpus of Contemporary American English (COCA; cf. Davies 2009) and the Norwegian Newspaper Corpus (Andersen 2005). These continually growing corpora
are particularly apt for studying change in progress and for revealing innovations
in language use, including the development of new and borrowed lexical items
(Renouf 2007; Furiasso and Hofland 2007). Web-based corpora, aka. cyber-corpora (Renouf 2007) use web crawler technology to harvest data which are subsequently filtered, annotated and transformed into a corpus. This has resulted in a
growing body of corpus-based studies using the web as its prime source of data
(Kilgarriff and Grefenstette 2003; Hundt et al. 2007). Although the exploitation of
these large corpora in pragmatics has so far been limited (but see section 3), the recent developments forecast a promising future for corpus-based pragmatics.
596
3.
Gisle Andersen
Methodological issues
An important event in the history of corpus linguistics was the 1991 Nobel Symposium on Corpus Linguistics, which is comprehensively reported in Svartvik
(1992). This volume does not focus on issues in pragmatics in particular, but it contains an impressive combination of papers that justifies its mention in this context.
It presents a theoretical and methodological basis with great scientific validity for
any type of corpus linguistic work, despite subsequent major technical developments that go well beyond what the pioneers of corpus linguistics could have foreseen (cf. above). The influential contributors, among them Chafe, Nelson Francis,
Fillmore, Halliday, Johansson, Greenbaum, Rissanen and Quirk, discuss a number
of theoretical issues and issues in observational methodology, corpus design and
corpus use. For instance, Chafe provides a strong argument for corpus linguistic
method in cognitive linguistics: the acceptance of the value of corpora is closely
tied to the acceptance of the view that language cannot be separated from what the
mind is and what the mind does and how the mind interacts with the universe in
which it is located. The modular view of language can get along without corpora;
the integrated view requires them (Chafe 1992: 81). Chafe further contextualises
corpus linguistics as an observational technique, placing the use of corpora within
a larger repertoire of observational alternatives (Chafe 1992: 84). Unlike introspective techniques that involve making judgements regarding invented language,
corpus linguistics involves the observation of overt behaviour. Further, corpus linguistics involves natural language, which distinguishes this method from artificial
methods like experimental techniques and elicitation tests. Corpora thus reflect
what people actually say and write. Ontologically, corpus linguistics shares its
characteristics with ethnography, in that it is both a natural and behavioural observation technique.
However, the introspection/observation dichotomy is problematic, especially
from the point of view of pragmatics. As Milroy (1987) points out, [i]n practice,
generalizations based on analyses of corpuses (sic!) are usually derived using an
additional method either introspection or some kind of experimental procedure. entirely corpus-based linguistic description of the kind attacked by
Chomsky (1957: 15; 1965: 2) is extremely rare and probably represents a concept
whose value is largely polemic (Milroy 1987: 4f). The corpus-based approach to
pragmatic phenomena relies on supplementing observations of overt linguistic behaviour with the intuitive knowledge of the analyst. Pragmatics concerns meaning
arrived at on the basis of inferential processes. The empirical researcher uses
spoken data as his/her primary stimuli on the basis of which he/she makes assumptions as regards the communicative intention of the observed interlocutors. Observable forms in a corpus, such as tokens of the word please, represent instantiations of pragmatic acts performed by speakers, such as the speech act function of
requesting somebody to do something, the attitudinal function of signalling tenta-
597
tiveness or the politeness function of face-threat mitigation. Naturally, it can be

hard to judge regarding speaker intentions. This methodological challenge is well
accounted for in Mosegaard Hansens (1998) comprehensive account of discourse
particles in French, where she argues that [a]n interpretative approach of course
has an inescapably subjective dimension, which means that one cannot in any
sense prove the correctness of ones analysis, but one can only hope to demonstrate
it as convincingly as possible (Mosegaard Hansen 1998: 214).
The complexity of spoken language data poses a challenge to the compilers and
users of corpora. The study of linguistic phenomena can be enhanced by annotations of various types of morpho-syntactic and discourse-structural information,
and contributions like Edwards and Lampert (1993) and Leech et al. (1995) offer
systematic accounts on the art of transcription and annotation of spoken discourse.
The availability of prosodically annotated corpora has provided uniform methods
for studies of intonation and its pragmatic effects (cf. e.g. Stenstrm 1984; Aijmer
1996). One objection that has been raised against the first generation of corpusbased studies is the fact that they regard the transcriptions themselves as the primary data, while the sound files are not consulted due to their unavailability to the
analyst. Although the unavailability of audio files can be explained with reference
to historical, practical and technological details (Wichmann 2007), this remains a
justifiable scientific objection for at least two reasons.
Firstly, like phonetics and phonology, many studies in conversation analysis
and discourse analysis need audio data because the interpretation depends on audible phenomena that could never be fully represented in the transcripts: there is far
more in the speech signal to be accounted for than could ever be captured in an orthographic transcription (Wichmann 2007: 76), and [n]ot even the most detailed
transcription can faithfully represent the raw data of spoken interaction (Mosegaard Hansen 1998: 207; cf. also section 4). Prosodic and segmental phenomena
play a more significant role than earlier anticipated in turn-taking, discourse marking, etc. (cf. Couper-Kuhlen 1996; Venditti and Hirschberg 2003; Wichmann 2007).
For example, Walker (2004) uses a variety of (mostly telephone-based) corpora to
explore how features such as pitch movements, glottal aspects, lingual and labial
closures are systematically associated with the signalling of transition relevance.
Secondly, it may be problematic for corpus users to rely fully on the judgements of the transcriber, since transcribers often differ with respect to their interpretations of spoken data. Ensuring inter-annotator agreement is a major challenge
in corpus compilation, and a study such as Raymond et al. (2002) has shown that
even experienced transcribers have different interpretations, particularly when annotating data at phonetic or prosodic levels. This underlines, of course, the analysts need for access to the raw spoken data when using corpora (Andersen 2001:
211ff).
Fortunately, new technology has made it much easier to access audible spoken
data in corpus linguistics. The state-of-the-art spoken corpora contain hypertext
598
Gisle Andersen
links between the transcripts and the audio and/or video data, cf. corpora such as
COLA, COLT, ICE-GB and NoTa (Johannessen and Hagen 2008; Andersen 2010).
There is thus a tendency in corpus building to move away from ambitious paralinguistic annotation schemes aimed at transcribing infinity (Cook 1990: 1) to implementing a more a rudimentary, but nevertheless searchable, prosodic transcription coupled with the actual sound files (Wichmann 2007). Importantly, video data
facilitates the analysis of gestures and facial expressions as part of conversational
behaviour, but the utilisation of video material in corpus-based pragmatics is still
in its infancy. The importance of using video data can be illustrated with reference
to minimal response signals (backchannels). Since active listenership may be realised verbally or non-verbally, typically by a gesture such as nodding, the articulated realisations cannot and should not be seen as separate from non-verbal realisations performing the same discourse function (Bjrge 2009). In other words, a
comprehensive study requires a holistic approach to language data, in which all aspects of utterances are investigated (cf. also Selting and Couper-Kuhlen 2001;
Couper-Kuhlen and Ford 2004). The fruitfulness of such an approach can be
shown with reference to a recent dissertation, Loehr (2004), that explores the functional parallelism between intonation and gestures. Specifically the study demonstrates the ability of gestures like intonation to signal completeness/incompleteness, information status, focus, emphasis, contrast, and even discourse
structure; in fact that hand movements reflect the on-going status of discourse,
while [i]ntonation signalled the discourse status much less subtly (Loehr 2004:
235). The exploration of the interaction between gesture and intonation is a
relatively recent academic pursuit, and studies have looked at temporal alignments
of the two modalities and the mapping between their respective hierarchical structures. Interestingly, there is strong evidence, provided initially by Kendon (1972,
1980) and McNeill (1985, 1992), that gesture and speech in general are two surface
facets of a single underlying thought being expressed (Loehr 2004: 1). Despite the
potential richness of corpus annotation, such a connection could not have been explored using merely transcription-based methods.3
4.
Corpora as basis for the study of topics in pragmatics
In the following, I address topics in pragmatics that have been devoted particular
attention from a corpus-based perspective, namely discourse markers (with a
special focus on Aijmer 2002), prosody (focusing on Wichmanns (1994, 2000,
2004, 2005) work), listener activities/backchannels (Bublitz 1988; Tottie 1991),
and corpus-based models of discourse structure (Sinclair and Coulthard 1975;
Stenstrm 1984, 1994; Sinclair 1996, 2004).
4.1.
599
Discourse markers
It is perhaps not surprising that discourse markers/particles or pragmatic

markers/particles are among the most focused topics in the literature, as this class
lends itself easily to form-based retrieval in corpora.4 Studies include items like
well, like, kind of, sort of, I mean etc. in English (Brinton 1996: 32), eh bien, donc
and allors in French (Mosegaard Hansen 1998), aber, doch, and ja in German (Abraham 1991), etc. With respect to terms and inventory, Fischer (2006) provides a
good overview of conceptual and terminological differences. The meanings of discourse markers are acknowledged to be excruciatingly hard to state (Wierzbicka
1991: 341). They make only a limited contribution to propositional content, but
contribute considerably at the pragmatic level, to the organisation of discourse (for
instance as markers of speaker transition or by marking boundaries in the discourse), to marking a speakers attitude towards the propositional content of the utterance (e.g. the Norwegian particle vel signalling reduced speaker commitment;
Fretheim 1981) and to various functions pertaining to the speaker-hearer relation,
such as you know indicating common ground between speaker and hearer (Jucker
and Smith 1998). Among the substantial contributions are Lenks (1998) study of
connectives and Andersens (2001) sociolinguistic study based on COLT and a
comparable subset of the BNC. The former focuses on anyway/anyhow, however,
still, incidentally, actually and what else in British and American English represented by LLC, CSAE and ICE-GB, whereas Andersen (2001) describes two discourse markers, like and innit, showing their pragmatic function, grammaticalisation and distribution (cf. Rhlemann, this volume). These two studies are
theoretically based in Sperber and Wilsons (1986/1995) relevance theory and analyse the pragmatic function of discourse markers as procedural constraints on
relevance, as does Jucker (1993) in his account of well. Comparative and cross-linguistic aspects of discourse markers are covered by Aijmer and Simon-Vandenbergen (2006) and Fischer (2000), analysing German ja, also, ne, oh and ach and
English yes, yeah, oh and well. Several of the studies in Jucker and Ziv (1998) and
Andersen and Fretheim (2000) are also based on corpora. Furthermore, Mosegaard-Hansen (1998) presents a detailed analysis of the six French markers bon,
ben, eh bien, puis, donc and alors. She bases her study on a collection of recorded
and transcribed corpora obtained from the archives of the UFR de linguistique
francaise of the Universit de Paris III (Mosegaard-Hansen 1998: 201). Among
the markers investigated are donc and alors, forms which both originate in temporal anaphoric expressions (deriving from the Latin forms dunc/illa hora, at that
time/at that hour, respectively). While donc is a monosemous marker of mutual
manifestness, in the sense of Sperber and Wilson (1995), the much more common
marker alors is multiply polysemous and its core use involves a shift in perspective, e.g. a new topic, marking a result or a conclusion, functions which are derived
metonymically from their diachronic sources.
600
Gisle Andersen
Aijmers (2002) contribution is illustrative of both the advantages of the corpus-based approach, as well as some of its problems. She describes a class of
items with fuzzy boundaries as they are used in LLC with occasional comparisons
with LOB and COLT. The inventory of forms includes now signalling emphasis or
transition, oh and ah and their functions as discourse particles signalling topic
shift, clarification or elaboration, just as a marker of speaker involvement, affect or
rhetorical use, the evidential particles (hedges) sort of and kind of, vagueness
markers like and all that, and that sort of thing, etc., and actually used with evidential functions. The value of Aijmers contribution lies first of all in its extensive
qualitative analysis of individual tokens in relevant contexts of use, as well as her
firm reliance on data, in what she considers a bottom-up approach. She develops
very specific criteria needed to distinguish discourse particles from other (often adverbial) uses of the items investigated, and focuses on core meanings and their derivative relation to the discourse functions of individual items, including their
grammaticalisation. A fundamental notion in Aijmers account is that of indexicality, that is to say, the potential of discourse particles to point to the speakers
epistemic attitude to the utterance and affective attitude to the hearer as well as to
the preceding and following discourse (Aijmer 2002: 55). This capacity can be likened to the indexicality of deictic expressions like anaphoric pronouns, in the
sense that they point to the presence of some entity in the immediate situation-athand (Ochs 1996: 414), including the speaker and hearer, spatial and temporal dimensions of the speech situation, epistemic stances like certainty/uncertainty, attitudes towards hearer relationship, social rank, group identity and attitudes towards
the speech act or some of its characteristics. Aijmer argues convincingly that discourse particles must be construed as polysemous items in what can be seen as
a modified meaning minimalist programme, having a single core meaning and
various related senses, as opposed to the proposed homonymy analysis in a meaning maximalist programme, in which particles are seen as having multiple meanings determined by the context of use, explicitly refuted by stman (1995) and
Mosegaard Hansen (1998). It is precisely their polysemy that can explain the grammaticalisation and the diachronic and synchronic relation between a discourse particle and its predecessor. For example, the temporal and topic transitional meanings of the discourse particle now are seen to co-occur as a result of a development
from adverb to discourse particle. Aijmer performs a meticulous qualitative analysis of each individual item, and the treatment of each particle includes descriptions
of a variety of features, including the pragmatic-functional range, collocational
analysis, prosody, grammaticalisation and relation to the words core meaning,
and, where relevant, comparisons with other particles, such as the comparison of
now with well with their respective adverbial functions.
A significant aspect of this contribution is the collocational analysis of each
particle. Specific collocational characteristics are viewed as a functional clue to
assist the hearer when a particle is ambiguous between different senses (Aijmer
601
2002: 28). To exemplify, the pragmatic value of oh is dependent on its co-occurrence with well, as the collocation oh well specifically indicates surprise and acceptance, at the same time belittling or trivialising a previous utterance (Aijmer
2002: 104). This can be contrasted with the collocation ah well, which has an upgrading effect, signalling the importance of the upcoming message. Collocations
also have a potential for distinguishing between marker and non-marker (i.e.
usually adverbial) use of an item, in that now in the contexts of well now, now then
and now look are invariably analysed as discourse markers. However, Aijmer does
not fully explore the possibility that collocations may also be markerhood indicators. It is for example notoriously difficult to draw the line between marker and
non-marker use of just (Tottie 1986), in other words the restrictive adverb with a
propositional meaning of exactly or only. Aijmer devotes relatively limited attention to the collocational patterns of just (but see Erman 1997) and it seems fair
to assume that the collocational analysis could have been fruitfully extended to include descriptions of how specific phrases are (perhaps exclusively) associated
with marker use, while others are not. An objection that could be made towards
empirical studies of discourse markers in general, is that collocational features are
largely underexplored. There are good grounds for claiming, as does Aijmer herself that [a]mong the characteristic properties of discourse particles which have
not received enough attention in the literature is their clustering (collocations
with other particles) (Aijmer 2002: 31; see also Vicher and Sankoff 1989). In
studies of discourse markers, the mention of collocational features is often coincidental rather than systematic. Aijmers work is illustrative in this respect, as her account of just amounts to a mere listing of the top most frequent collocational
patters (as in Aijmer 2002: 161), and the collocational analysis for actually is
ignored altogether. Moreover, the presumably common collocation oh really is
ignored in the section on collocations, albeit acknowledged as a unit with a separate status as a topicaliser in the discussion of pragmatic functions.
Aijmer gives no reason for her selection of the seven items studied. Her only
reference to criteria for inclusion is frequency in LLC, but we are not told whether
this frequency as discourse particles or overall frequency of the forms in question.
One might well ask why a particle like well is not accounted for in her description;
presumably this is because she finds this particle sufficiently covered in the abundant previous literature. Besides, the common use of like as a discourse marker is
only devoted sporadic attention, illustrating a difference between LLC and subsequent English corpora.
4.2.
Prosody
The availability of spoken corpora has prompted many studies on prosody, from
the early LLC-based studies using prosodic transcriptions as primary data, to
studies using instrumental analysis of speech data as a basis for research. In prag-
602
Gisle Andersen
matics, many corpus-based studies fail to explore prosody systematically as part of

the analysis. With respect to the literature on discourse markers outlined above,
this general shortcoming is surprising, since prosody may be crucial for the determination of markerhood status, and since grammaticalisation processes tend to be
accompanied by reduction in form. Aijmers (2002) study is exceptional in that it
utilises the prosodic annotations of LLC extensively, which adds to the credibility
of the analysis.
An important effort in the study of intonation is the Intonational Variation in
English (IViE) Corpus, containing speech data and intonation transcriptions from
nine urban dialects of British English. The data represent five different speaking
styles, including conversation, map task dialogues, retold stories, reading passages
and sentences. Fletcher et al. (2004) use this corpus in a comparative study of the
high-rising terminal (HRT) in four varieties of English. HRT refers to the use of the
high-rising intonation contour generally associated with yes/no questions in English in declarative utterances. With respect to the function of this feature, Guy
and Vonwiller suggest that HRT correlates with the semantic complexity of the
text and therefore the need for checking to see if the audience is understanding
what is being said (1989: 25). Previous studies of Australian and New Zealand
English have shown that HRT is associated with low prestige varieties, is socially
stigmatised, occurs most often in the telling of narratives and is used predominantly by young adolescent females as part of narratives (Allan 1990). Contrary to
this, Fletcher et al. (2004) show that it is the male speakers in IViE who use HRTs
the most. According to the authors, the advantage of using the IViE corpus for this
study is that it provides sufficient coverage of intonational phenomena across a
very large corpus comprising several, very different, varieties of British English
(Fletcher et al. 2004: 407).
Several contributions by Wichmann (1994, 2000, 2004, 2005) exemplify how a
corpus-based method provides a valuable complement to previous conversation
analytic work based on experimentally controlled perception studies (e.g. CouperKuhlen and Selting 1996). Much of her work is based on SEC, while more recent
studies explore the British component of the ICE corpus (ICE-GB). Although corpora which capture speech only in its natural habitat (Wichmann 2004: 1527)
often do not maintain the sound quality needed for instrumental analysis, Wichmanns work firmly demonstrates how an auditory analysis of the data can lead to
enlightening scientific results. Wichmann (2000) analyses how intonation is used
to signal the structure of spoken texts and its role in turn-taking and in signalling
cooperation and clarifying the relationship between utterances in conversational
discourse. For example, major shifts in topic are distinguishable from units with
close cohesive ties on prosodic grounds. Furthermore, Wichmann (2004) shows
the advantages of using a speech corpus in a study of please-requests. She describes the pitch contour and accentual status, focusing on how the realisation
varies according to the speech situation and the speech act function of utterances,
603
as well as the utterance position of please. Her study explores formulaic expressions in which please occur, notably yes please and please do, the isolate
please itself, which is used to reinforce a more explicit request (Wichmann 2004:
1539), as well as the exaggerated segmental realisation [phj li::z] (Wichmann
2004: 1540), where the epenthetic schwa leads to what is in effect a disyllabic realisation, usually expressing scorn or disapproval. The fact that her study is based on
a corpus representing speakers of varying backgrounds and different contexts of
use enables the author to relate her findings of prosodic patterns to the social distance and symmetrical/asymmetrical power relations between interlocutors. This
provides an interpretational perspective which supplements the existing accounts
of the discourse effects of terminal contours, e.g. of a falling contour as contributing a feeling of closure (Croft 1995) and a final rise as signalling openness
(Cruttenden 1997). At the same time she redefines the role of please itself used as a
face-threat mitigator in requests as a gesture of courtesy that contextualises the
accompanying request as occurring within a known set of rights and obligations
(Wichmann 2004: 1544). The importance of intonation for utterance interpretation
is further corroborated in Wichmanns (2005) study of emotional requests, which
also demonstrates the value of investigating low-frequency items. The intonation
of please help yourself with a high onset on please and a falling tone on self functions as a neutral invitation, while please help yourself, with a fall on please, is
much more pleading, functioning as an emphatic, emotional exhortation. Collectively, Wichmanns corpus-based work has contributed consideraby to our understanding of the relation between intonation and speaker attitude, showing that the
attitudinal information carried by intonation is not an inherent feature of specific
intonation contours but a result of context-based pragmatic inference.
Other corpus-based discourse studies focus on the segmental aspects of prosody. Two early contributions by Stenstrm (1990a, 1990b) explore prosodic segmentation and pauses in monologue and dialogue. Stenstrms (1990b) category of
pauses is taken to include so-called verbal fillers like hedges (sort of), softeners (you know), stallers (well) and initiators (anyway). Stenstrm points out that
pauses play important roles in reformulations and as turntakers, turnyielders and
turnholders. Based on the same corpus, Romero (1994) explores the role of pauses
as content-bearing elements, showing that filled and unfilled pauses contribute to
the realisation of Hallidays (1985) theme/rheme structure of utterances. The interpersonal function of the filled pause is to indicate a new move in the conversation,
which may or may not involve a speaker shift. In their textual function, filled
pauses are used as the starting point of the following message. Finally, Romero
(1994) argues that filled pauses also have an ideational function of signalling an element referred to elsewhere in the discourse, typically by anaphora. Though he
may be right in claiming that filled pauses carry meaning of a structural and interactional nature, this ideational function seems dubious. Though the filled pause
admittedly co-occurs with anaphoric referential expressions, it seems unjustified to
604
Gisle Andersen
claim that the pause has to be considered a meaning-carrying particle which conveys semantic information in terms of structure and content (Romero 1994: 502),
as this seems to suggest that pauses are capable of taking on referential meanings
and contributing to the propositional content of utterances.
4.3.
Listener activities and backchannels
Studies in corpus-based pragmatics have become increasingly concerned with the

role of listeners in conversation. Corpus data can be effectively used to demonstrate how interactants actively signal listenership, while allowing the other
speaker to continue to talk. Response signals convey comprehension and agreement or disagreement and are often also furnished with attitudinal meanings, signalling involvement of various types (positive/negative evaluation) and of various
strengths, ranging from weak doubt or scepticism to enthusiastic agreement. Listenership can be expressed by auditory or visual signals. By far, the existing literature has mostly focused on the auditory side, which includes backchannels such as
mm, mhm or uh-huh, attitudinally salient interjections such as wow or damn, discourse markers such as oh, yes/yeah and okay, lexical repetitions, etc. To a much
lesser degree, non-verbal signals such as nods, furrowed brows and gestures have
been studied (Norrick and Fischer 2009).
The term backchannel first appears in the pioneering work of Yngve (1970), referring to the existence of what I call the back channel, over which the person who
has the turn receives short messages such as yes and uh-huh without relinquishing
the turn (Yngve 1970: 568). The LLC has been instrumental in providing an empirical basis for the study of listenership in discourse. Orestrm (1983) provides an
early LLC-based account which describes backchannels in the wider context of
turn-taking which includes a rather narrow set of items (m, yes, yeah, mhm, no,
quite, aha, good) but excludes tokens preceded by the markers you know and you
see. Similarly, Bublitz (1988) uses a subset of the LLC (excluding meetings and interviews) in a study of recipient action which includes hearer activities and certain reactive or recipient speaker contributions (Bublitz 1988: 141). Acknowledging a generally prevailing tendency to agree with, assent to and support the
fellow-speakers attitudes and statements (Bublitz 1988: 153), Bublitz makes a
distinction between primary speakers, secondary speakers and hearers. Unlike a
secondary speaker, who makes a minor contribution to the topic at hand, the hearer
signals that he is taking note of what has been said and meant and uses particular hearer signals taken from a closed set (Bublitz 1988: 153). Bublitzs study is a
bottom-up approach where a careful qualitative analysis of backchannels precedes
generalisations about their basic functions, their high flexibility with respect to discourse placement, and their formal features (usually short and monosyllabic). The
closed paradigm of forms includes yes, yeah, yep, uhuh, h, m, mhm, I see and
really (with a falling tone). Bublitz excludes no from the category on the grounds
605
that it provides an expression of stating the speakers position as to the current

topic (Bublitz 1988:181), thus its function is more than simply taking note of the
previous speakers utterance.
Tottie (1991) provides a comparison between the British and American English
represented by LLC and CSAE. This contribution is important as it addresses
theoretical and methodological problems in connection with corpus-based listenership research. It focuses on differences between supportive and regulative functions of backchannels, that is, the dual function of signalling understanding/agreement and encouraging the speaker to continue. Tottie points out that [t]he term
backchannel has been extended to designate backchannel message (Tottie
1991: 256). In fact, in its current use, the term is multiply polysemous, as it is also
used to refer to forms that can realise the supportive and regulative functions of listenership signals, as well as actual instantiations (tokens) in real data (Andersen
2009). Tottie also problematises the fact that the two datasets may not necessarily
be comparable in all respects, for example since the LLC transcription records
tone in detail but does not account for lengthening, which is shown in the CSAE
transcription (Tottie 1991: 260). This illustrates that the information available to a
researcher does not always make corpora fully equivalent. A more substantial difference concerns the data themselves, not the transcription conventions. Tottie suggests that, in order to draw justifiable conclusions in a comparative corpus-based
study, it is necessary to control for a variety of non-linguistic factors in order to
avoid that emerging differences are artefacts of biased data. She explicitly discusses factors such as the speakers age, gender and affinity with each other, subject matter and degree of consensus, as well as whether or not the conversations are
surreptitiously recorded. Ideally, in comparative studies, the data should match on
all of these points, certainly if one is to make any sense of quantitative comparisons. Tottie acknowledges that her corpus data match in some of these respects, but
not all. Nevertheless, she is able to demonstrate convincingly the qualitative differences in intonation, syllabicity, articulation and length of the British and American variety of mm.
Tottie raises the issue of what distinguishes a backchannel from a turn, stating
that the most important principle is that backchannel status can be determined
only on the basis of the following utterance (Tottie 1991: 260). It is problematic
that Tottie links backchannelling so closely to observable speech production, without taking speaker/listener intentions into account. A backchannel such as mm may
be intended as mere supportive device without the listener intending to take the
floor. But it may well be that the primary speaker chooses not to continue his talk,
in which case the backchanneller cedes the opportunity and there is a change of
speaker. It is problematic to argue that this changes the functional characteristics of
the backchannel item to any great extent. Therefore, a sharp division between an
mm that does not involve change of speaker and one that does so does not seem justified. Totties study concludes with a valuable proposal for a future research pro-
606
Gisle Andersen
gramme on backchannelling, proposing the need to investigate whether subject

matter, sex, and number of participants determine the use of backchannelling and
how their prosodic realization is linked to discourse function (1991: 270). Although many of her questions still remain unanswered, subsequent corpus-based
research has provided important insights into this field of study. For example,
McCarthy (2002; cf. also 2003) studies response tokens used in everyday conversation, focusing on differences between the British CANCODE corpus and the
Cambridge-Cornell Corpus of Spoken North American English. Moreover, the
conversation analytical work of Gardner (1998) argues that research-based findings on receipt tokens should have bearings on how language is taught, concluding
that it would be worth considering exposing learners to these interactional exponents, and providing opportunities for practice and feedback (Gardner 1998:
220221). Heintz (2003) compares the backchannel behaviour of German and
American English speakers, demonstrating differences between Germans who are
proficient in English and those that are not. Concerned with the use of English as a
lingua franca (ELF), Bjrges (2009) recent investigation is valuable as it sees verbal and non-verbal aspects of backchannelling in context. She investigates video
recordings of ELF interaction involving students partaking in simulated business
negotiations, showing that backchannel behaviour varies according to conflict
level. Some of these results are corroborated by Andersen (2009, based on a similar
dataset), who considers backchannels in polylogic discourse, with a particular
focus on the interactional placement of backchannels, that is, the direction of backchannelling between two teams representing a seller and buyer. Finally, Kjellmer
(2009) considers the insertion point of backchannels, observing that, although
most often inserted at the end of a completed utterance, backchannel insertions
often occur when the utterance has not been completed. What this signifies is that
the information-content of an utterance can be conveyed to the listener before it
has been expressed in toto (Kjellmer 2009: 107).
4.4.
Models of discourse structure
Given the short history of corpus-based pragmatics, special attention should be

paid to scholarly contributions which are less concerned with micro-phenomena
but which focus on the theoretical basis of this research domain, addressing basic
issues such as the units of discourse and their sequential organisation, the mapping
of lexical, morphological and syntactical form with illocutionary function, participant relations, textual co-production, etc. Inspired by theoretical precursors like
Firth (1935) and Sinclair (1966), the seminal contribution by Sinclair and Coulthard (1975) was original in its exploration of naturally occurring speech in an academic atmosphere which massively discarded the notion that studying real data
might have any significance. Their choice of the classroom as the venue for
collecting samples of speech was motivated by their wish to control as many of
607
the potential variables as possible (Sinclair and Coulthard 1975: 6) and overcome
the many problems inherent in natural conversation and opt for a more simple
type of spoken discourse, one which has a much more overt structure, where one
participant has acknowledged responsibility for the direction of the discourse,
for deciding who shall speak when, and for introducing and ending topics (Sinclair and Coulthard 1975: 6). Based on a restricted corpus of transcribed speech
data, they develop an analytical apparatus which, despite subsequent revisions
(e.g. Sinclair and Brazil 1982; Sinclair and Coulthard 1992) and criticism (Stubbs
1984), still remains influential and has been inspirational for a wealth of research.
They propose a discourse model of exchange and transaction structure which
shows that spoken discourse can be systematically described according to its structure, just as syntax can be. The model, which is closely modelled on Hallidays
Categories of a Theory of Grammar (sic.; Sinclair and Coulthard 1975: 24; cf.
Halliday 1961), encompasses a hierarchy of discourse components, specified as a
ranking scale containing the four components transaction, exchange, move and act.
In its original version, the model also included a top-level of lesson which has in
later revisions of the theory been dismissed as a stylistic type specific to the usage
domain of classroom discourse. At each level, discourse units are divided into
classes which realise an element of a structure, and the internal structure of the unit
is described. Exchanges have a three-part structure containing the elements initiation, response, feedback (IRF-structure), with the corresponding classes of move:
opening, answering and follow-up (Sinclair and Coulthard 1975: 26). Despite later
criticism for example concerning the unnatural nature of classroom discourse (in
which teachers ask pseudo-questions) upon which the model is based, this contribution effectively involves no less than the inception of a research field, discourse analysis.
A decade later, Stenstrm (1984) develops this model further in a comprehensive analysis of question/response strategies. Her contribution is valuable in that it
uses LLC as a test bed for the discourse model in an analysis of authentic conversational data. She adds empirical support to the theory by accounting for how the
different levels defined in the model of analysis were manifested in the corpus
(Stenstrm 1984: 94). As its title suggests, the primary focus is on exchanges. Although there is a bias towards the analysis of responses, as opposed to the initiation
and feedback moves, Stenstrm provides a detailed account which represents a
much wider user domain than the original empirical basis for the model, selected
to represent the greatest possible variety of speech situations (Stenstrm 1984:
262), ranging from informal conversation to procedural dialogue (planned, public discourse; e.g. courtroom interaction). She extends the model in various ways
by introducing notions such as topic-bound transactions (anaphorically related to
the previous context) versus proposition-bound transactions (forward-oriented)
and by specifying optional and obligatory moves in exchange sequences. Her study
incorporates categories like hesitation, pauses, hedges, reformulations and the like,
608
Gisle Andersen
and shows how they are linguistically realised in a corpus which, for its time, was
both extensive and varied. The Stenstrmian model (Kirk 2002) is further developed into what the author herself refers to as a synthesis of much of the work that
has been done so far in the analysis of conversation (Stenstrm 1994: xii). Kirk
(2002) provides a good illustration of how the model can be applied in the analysis
of real corpus data, arguing that although there are many competing analytic models available (e.g. Edwards and Lampert 1993; Eggins 1996; Langford 1994; etc.),
none is as rigorous or as syntactically systematic as that of Anna-Brita Stenstrm
(Kirk 2002: 144). Despite Kirks praise, the model has raised criticism as regards
the conceptual nature of discourse units and their relation to each other, and to the
hierarchy of units, theoretical difficulties [which] are of course problems of the
subject in general, and are not specific to the book (Knowles 1995: 122). The relation of the turn (not part of Sinclair and Coulthards model) to the rest of the hierarchy is particularly problematic, as an exchange boundary may come in the
middle of a turn, and since a turn boundary may come in the middle of a move, for
instance if a speaker finishes off someone elses word in joint utterance production.
In his more recent work, Sinclair has introduced the notion of extended units
of meaning, first in the seminal paper Sinclair (1996), and more comprehensively
in Sinclair (2004). He presents the hypothesis that the notion of a linguistic item
can be extended so that units of meaning are expected to be largely phrasal
(Sinclair 1996: 82). Although the fixedness of word strings is most obvious in idiomatic/proverbial usage and phrasal verbs, Sinclairs model extends the concept of
phrasehood to words in general. Corpus evidence shows that each word in the language has its own unique lexico-grammatical profile and that different senses of
for instance a noun such as bank have their own distinctive sets of collocates. Sinclair argues that these observations force the linguist to reassess the notion of
meaning. His model includes structural categories with an increasing level of abstraction: collocation (a frequent co-occurrence of words, such as his example
naked eye), colligation (the co-occurrence of grammatical choices, e.g. choice of
preposition with/to in front of the naked eye), semantic preference (the collocation of a lexical item with items from a specific (more or less general) semantic
subset Bednarek 2008: 119) and semantic prosody (persistent collocation with
items with specific connotations; Louw 1993; cf. also Hunston 2009). Despite the
term and importantly for the current study Sinclair views semantic prosodies as
attitudinally significant and places them on the pragmatic side of the semantics/
pragmatics continuum (Sinclair 1996: 87). Indeed, speakers choice of a semantic
prosody is what links meaning to communicative purpose: Without it, the string
of words just means it is not put to use in a viable communication all subsequent choices within the lexical item relate back to the prosody (Sinclair 1996:
88). This can explain the pragmatic difference between semantically equivalent
sentences like I will make no promises and I will not make any promises. Hence,
Sinclairs analysis has an explanatory power to account for the creation of new
609
pragmatic expressions like make no mistake, which turns out strikingly more common than any other collocation of make no in a large corpus such as COCA. Sinclair (2004) develops the analytical model further by specifying the core (the invariable component of a lexical item) and the semantic prosody as the two
obligatory components of a lexical item, and the remaining three (collocation, colligation and semantic preference) as optional components, by relating the categories of co-selection to the paradigmatic and syntagmatic dimensions of language.
Furthermore, Stubbs (1996, 2001) is another linguist who argues that meaning
is a largely a phraseological phenomenon and whose corpus-based accounts also
contribute to the rigour and maturity of the field of corpus-based discourse analysis. Like the previous models, he draws on the British neo-Firthian tradition by
Halliday and Sinclair, but this account is more sociologically oriented than the
ones referred to above. It takes into account the status of speakers and listeners and
the importance of social institutions for the interpretation of the meaning of texts.
Stubbss linguistic analysis, largely based on collocation patterns, focuses on how
users lexis and grammatical choices (e.g. active vs. passive voice) contribute to
different meanings that can be interpreted in a social context.
The rapidly growing interest in corpus-based pragmatics is underlined by several collective volumes which for reasons of space have to be mentioned in passing. Adolphs (2008), del and Reppen (2008) and Biber et al. (2007) are significant contributions that are exclusively or predominantly corpus-based and which
address the main theoretical and methodological tenets of this field, while Baker
(2006) provides an enlightening evaluation of a variety of methodologies for corpus-based discourse analysis.
5.
Concluding remarks
The overview presented above has demonstrated the breadth and complexity of the
field of corpus-based pragmatics. Most studies consider overt and easily retrievable features such as discourse markers, connectives, hedges etc., and are formbased (cf. section 1). There are fewer studies where corpora are used to explore a
particular discourse function and its possible realisations, although such studies
appear to be on the increase. Relevant examples would be Deutschmanns (2006)
study which explores particular speech act function apologies and its various
realisations, and Stenstrm et al. (2002) who demonstrate how quotations are
marked by different structures, including BE + like, GO and SAY as reporting verbs,
voice modulation and zero-quotations.
There appears to be an ongoing increase in the different topics that are explored
by corpora. Schmid (2001) studies various patterns for expressing presuppositions.
Corpora are also used in the study of tropes such as metaphor (Sandford 2008) and
610
Gisle Andersen
hyperbole in recent contributions such as Norrick (2004) and McCarthy and Carter
(2004b), who argue that this approach can add substance to the already healthy
existing literature on rhetorical strategy, conventional and creative metaphor, irony
and over- and understatement. (2004b: 158). These contributions illustrate the
problem associated with one-to-one searching, as it assumes an a priori understanding of the forms that may realise a particular discourse function. This is especially the case in studies of creative language processes such as the use of tropes.
I have argued in favour of supplementing corpus-based pragmatics studies with
a corpus-driven approach. Corpus-based pragmatics studies require a hermeneutical and holistic approach to language data, because face-to-face interaction involves multiple layers of information. Video-based research methods are therefore
valuable. Modern multimodal corpora developed with state-of-the-art technology
provide a promising outlook for such studies in the future.
Notes
1. I am grateful to the editors and the anonymous reviewer for valuable comments on the
first draft of this paper.
2. This can for example be seen from the studies included in three recent special issues of
Journal of Pragmatics focusing on corpus linguistics (e.g. McCarthy and Carter 2004a),
of which a vast majority is based on English.
3. Another shortcoming of some corpora is that the data are only available as short snippets
of sound and not as longer stretches of spoken dialogue. For many research purposes, access to longer text passages is a prerequisite, e.g. studies of the structure and development
of narratives in conversation.
4. In the current account, I opt for the term discourse markers, which seems the most commonly used, especially in the most recent literature. For discussions on terminology,
cf. Brinton (1996); Andersen (2001), Fischer (2006) and Mosegaard Hansen (2006), who
argues that discourse particle is misleading on formal grounds.
References
Aarts, Jan and Willem Meijs (eds.)
1986
Corpus Linguistics II. New Studies in the Analysis and Exploitation of Computer Corpora. Amsterdam: Rodopi.
Abraham, Werner (ed.)
1991
Discourse Particles. Amsterdam: Benjamins.
del, Annelie and Randi Reppen (eds.)
2008
Corpora and Discourse: The Challenges of Different Settings. Amsterdam:
Benjamins.
Adolphs, Svenja
2008
Corpus and Context: Investigating Pragmatic Functions in Spoken Discourse.
611
Aijmer, Karin and Anne-Marie Simon-Vandenbergen (eds.)

2006
Pragmatic Markers in Contrast. Amsterdam: Elsevier.
Aijmer, Karin
1986
Why is actually so frequent in spoken English? In: Gunnel Tottie and Ingegerd
Bcklund (eds.), English in Speech and Writing: A Symposium, 119129.
Stockholm: Almqvist and Wiksell.
Aijmer, Karin
1989
Themes and tails: The discourse functions of dislocated elements. Nordic
Journal of Linguistics 12: 137154.
Aijmer, Karin
1996
Conversational Routines in English. London: Longman.
Aijmer, Karin
2002
English Discourse Particles: Evidence from a Corpus. Amsterdam: Benjamins.
Algeo, John
1988
British and American grammatical differences. International Journal of Lexicography 1: 131.
Allan, W. Scott
1990
The rise of New Zealand intonation. In: Allan Bell and Janet Holmes (eds.),
New Zealand Ways of Speaking English, 115128. Clevendon: Multilingual
Matters.
Altenberg, Bengt and Karin Aijmer
2000
The English-Swedish Parallel Corpus: a resource for contrastive research and
translation studies. In: Christian Mair and Marianne Hundt (eds.), Corpus Linguistics and Linguistic Theory, 1534. Amsterdam: Rodopi.
Altenberg, Bengt
1986
ICAME bibliography. ICAME News 10: 6279.
Altenberg, Bengt
1987
Prosodic patterns in spoken English. Studies in the Correlation between Prosody and Grammar for Text-To-Speech Conversion. Lund: Lund University
Press.
Altenberg, Bengt
1991
Amplifier collocations in spoken English. In: Stig Johansson and Anna-Brita
Stenstrm (eds.), English Computer Corpora, 127147. Berlin: Mouton de
Gruyter.
Altenberg, Bengt
1997
Exploring the Swedish component of the International Corpus of Learner English. In: Barbara Lewandowska-Tomaszcyk and Patrick James Melia (eds.),
PALC97: Practical Applications in Language Corpora, 119132. Ldz: Ldz
University Press.
Altenberg, Bengt
2006
The function of adverbial connectors in second initial position in English
and Swedish. In: Karin Aijmer and Anne-Marie Simon-Vandenbergen (eds.),
Pragmatic Markers in Contrast, 1138. Amsterdam: Elsevier.
Andersen, Gisle and Thorstein Fretheim
2000
612
Gisle Andersen
Andersen, Gisle
2001
Pragmatic Markers and Sociolinguistic Variation: a Relevance-Theoretic Approach to the Language of Adolescents. Amsterdam: Benjamins.
Andersen, Gisle
2005
Assessing algorithms for automatic extraction of anglicisms in Norwegian
texts. In: Proceedings from The Corpus Linguistics Conference Series, Vol. 1,
no. 1, Corpus Linguistics 2005.
Andersen, Gisle
2006
Engelske importord i norsk. Sprknytt 3/2006: 2732.
Andersen, Gisle
2009
Responses in academic and professional discourse. Paper presented at the
11th International Pragmatics Conference, Melbourne, 1217 July 2009.
Andersen, Gisle
2010
How to use corpus linguistics in sociolinguistics. To be published in Anne
OKeeffe and Michael McCarthy (eds.), Routledge Handbook of Corpus Linguistics, 547562. Londin: Routledge.
Aston, Guy and Lou Burnard
1998
The BNC Handbook. Exploring the British National Corpus with SARA. Edinburgh: Edinburgh University Press.
Baker, Paul
2006
Using Corpora in Discourse Analysis. London: Continuum.
Bald, Wolf-Dietrich
1990
An example of phonological reduction in English. In: S. Ramsaran (ed.),
Studies in the Pronunciation of English. A Commemorative Volume in Honour
of A.C. Gimson, 317322. London: Routledge.
Banjo, Ayo
1996
The sociolinguistics of English in Nigeria and the ICE project. In: Sidney
Greenbaum (ed.), Comparing English Worldwide: The International Corpus of
English, 239248. Oxford: Clarendon Press.
Bauer, Laurie and Janet Holmes
1996
Getting into a flap! /t/ in New Zealand English. World Englishes 15: 115124.
Beaugrande, Robert de
1999
Linguistics, sociolinguistics, and corpus lingistics: Ideal language versus real
language. Journal of Sociolinguistics, 3: 128139.
Bednarek, Monika
2008
and Lingustic Theory 4: 119139.
Biber, Douglas (ed.)
1994
Sociolinguistic Perspectives on Register. Oxford/New York: Oxford University Press.
1989
Styles of stance in English. Lexical and grammatical marking of evidentiality
and affect. Text 9: 93124.
Biber, Douglas, Ulla Connor and Thomas A. Upton
2007
Discourse on the Move: Using Corpus Analysis to Describe Discourse Structure. Amsterdam: Benjamins.
613
Biber, Douglas
1986
Spoken and written textual dimensions in English: Resolving the contradictory
findings. Language 62: 384414.
Biber, Douglas
1987
A textual comparison of British and American writing. American Speech 62:
99119.
Biber, Douglas
1988
Press.
Biber, Douglas
1993
An analytical framework for register studies. In: Douglas Biber and Edward
Finegan (eds.), Sociolinguistic Perspectives on Register, 3156. New York:
Biber, Douglas
1995
Dimensions of Register Variation. A Cross-Linguistic Comparison. Cambridge: Cambridge University Press.
Biber, Douglas
2009
A corpus-driven approach to formulaic language in English: Multi-word patterns in speech and writing. International Journal of Corpus Linguistics 14.3:
275311.
Bjrge, Anne-Kari
2009
Conflict or cooperation: the use of backchannelling in ELF negotiations. English for Specific Purposes 29: 191203.
Brazil, David
1982
Impromptuness and intonation. In: Nils Erik Enkvist (ed.), Impromptu Speech:
a Symposium, 277290. bo: bo Akademi.
Brinton, Laurel J.
1996
Pragmatic Markers in English: Grammaticalization and Discourse Functions.
Brodda, Benny
1988
Tracing turns in the London-Lund Corpus with BetaText. Literary and Linguistic Computing 3: 94104.
Bublitz, Wolfram
1979
Tag questions, transformational grammar, and pragmatics. Papers and Studies
in Contrastive Linguistics 9: 521.
Bublitz, Wolfram
1988
Supportive Fellow-Speakers and Cooperative Conversations. Amsterdam:
Benjamins.
Chafe, Wallace L
1992
The importance of corpus linguistics to understanding the nature of language.
In: Jan Svartvik (ed.), Directions in Corpus Linguistics. Proceedings of Nobel
Symposium 82, 7997. Berlin: Mouton de Gruyter.
Chafe, Wallace L., John W. Du Bois and Sandra A. Thompson
1991
Towards a new corpus of Spoken American English. In: Karin Aijmer and
Bengt Altenberg (eds.), English Corpus Linguistics. Studies in Honour of Jan
Svartvik, 6482. London: Longman.
614
Gisle Andersen
Chafe, Wallace
1992
In: Jan Svartvik (ed.), Directions in Corpus Linguistics. Proceedings of
Nobel Symposium 82 Stockholm, 48 August 1991, 7997. Berlin: Mouton de
Gruyter.
Chomsky, Noam
1957
Syntactic Structures. The Hague: Mouton.
Coates, Jennifer
1987
Epistemic modality and spoken discourse. Transactions of the Philological Society 85(1): 110131.
Collins, Peter
2005
It-clefts and wh-clefts: Prosody and pragmatics. Journal of Pragmatics: 38:
17061720.
Cook, Guy
1990
Transcribing Infinity: Problems of Context Presentation. Journal of Pragmatics 14: 124.
Couper-Kuhlen, Elizabeth and Cecilia E. Ford (eds.)
2004
Sound Patterns in Interaction: Cross-linguistic Studies from Conversation.
Couper-Kuhlen, Elizabeth and Margret Selting (eds.)
1996
Prosody in Conversation. Cambridge: Cambridge University Press.
1996
The prosody of repetition: on quoting and mimickry. In: Elizabeth CouperKuhlen and Margret Selting (eds.), Prosody in Conversation: Interactional
Studies, 366405. Cambridge: Cambridge University Press.
Cresti, Emanuela and Massimo Moneglia (eds.)
2005
C-ORAL-ROM: Integrated Reference Corpora for Spoken Romance Languages. Amsterdam: Benjamins.
Croft, William
1995
Intonation units and grammatical structure. Linguistics 33: 839882.
Crowdy, Steve
1995
The BNC spoken corpus. In: Geoffrey Leech, Greg Myers and Jenny Thomas
(eds.), Spoken English on Computer. Transcription, Mark-up and Application.
224234. London: Longman.
Cruttenden, Alan
1997
Intonation. Cambridge: Cambridge University Press.
Crystal, David and Randolph Quirk
1964
Systems of Prosodic and Paralinguistic Features in English. The Hague: Mouton.
Crystal, David
1972
The intonation system of English. In: Dwight Bolinger (ed.), Intonation,
110136. Harmondsworth: Penguin.
Crystal, David
1975
The English Tone of Voice. Essays on Intonation, Prosody and Paralanguage.
London: Arnold.
Davidson, Benjamin D.
1977
Aspects of information structuring in modern spoken English. In: Wolf-Diedrich Bald and Robert Ilson (eds.), Studies in English Usage: The Resources of
615
Present-day English Corpus for Linguistic Analysis, 157170. Frankfurt am

Main: Peter Lang.
Davies, Mark
2009
The 385+ Million Word Corpus of Contemporary American English (1990
2008+): Design, Architecture, and Linguistic Insights. International Journal
of Corpus Linguistics 14: 15990.
Deutschmann, Mats
2006
Social variation in the use of apology formulae in the British National Corpus.
In: Antoinette Renouf and Andrew Kehoe (eds.), The Changing Face of Corpus Linguistics, 205222. Amsterdam/New York: Rodopi.
Downing, Angela
2006
The English pragmatic marker surely and its functional counterparts in Spanish. In: Karin Aijmer and Anne-Marie Simon-Vandenbergen (eds.), Pragmatic
Markers in Contrast, 3958. Amsterdam: Elsevier.
Edwards, Jane A. and Martin D. Lampert
1993
Talking Data: Transcription and Coding in Discourse Research. Hillsdale:
Lawrence Erlbaum.
Eggins, Suzanne
1996
Analysing Casual Conversation. London: Continuum.
Enkvist, Nils Erik (ed.)
1982
Impromptu Speech: a Symposium. bo: bo Akademi.
Enkvist, Nils Erik
1973
Linguistic stylistics. Janua Linguanun Series Critica 5. The Hague: Mouton.
Erman, Britt
1986
Some pragmatic expressions in English conversation. In: Gunnel Tottie and
Ingegerd Bcklund (eds.), English in Speech and Writing: a Symposium, 131
147. Stockholm: Almqvist and Wiksell.
Erman, Britt
1987
Pragmatic Expressions in English: a Study of You Know, You See, and I Mean
in Face-To-Face Conversation. Stockholm: Almqvist and Wiksell.
Erman, Britt
1997
Guys just such a dickhead; the context and function of just in teenage talk.
In: Ulla-Brite Kostsinas, Anna-Brita Stenstrm and Anna-Malin Karlsson
(eds.), Ungdomssprk i Norden, 8295. Stockholm: Institutionen fr nordiska
sprk.
Faber, Dorrit and Karen M. Lauridsen
1991
The compilation of a Danish-English-French corpus in contract law. In: Stig
Johansson and Anna-Brita Stenstrm (eds.), English Computer Corpora. Selected Papers and Research Guide, 235243. Berlin: Mouton de Gruyter.
1982
Phatic, metalingual and metacommunicative functions in discourse. Gambits
and repairs. In: Nils Erik Enkvist (ed.), Impromptu Speech: a Symposium,
71104. bo: bo Akademi.
Fang, Alex C.
1993
Building a corpus of computer science. In: Jan Aarts, Pieter de Haan and Nelleke Oostdijk (eds.), English Language Corpora: Design, Analysis and Exploitation. 7378. Amsterdam: Rodopi.
616
Gisle Andersen
Fill, Alwin
1992
Joking in English and German: A contrastive study. In: Christian Mair and
Manfred Marcus (eds.), New Departures in Contrastive Linguistics. Vol. 2,
2131. Innsbruck: Institut fr Anglistik, Universitt Innsbruck.
Firth, John R
1935
The technique of semantics. Reprinted in 1957 in Papers in Linguistics,
193451,723. Oxford: Oxford University Press.
Fischer, Kerstin
2000
From Cognitive Semantics to Lexical Pragmatics: The Functional Polysemy of
Discourse Particles. Berlin: Mouton de Gruyter.
Fischer, Kerstin
2006
Approaches to Discourse Particles. Amsterdam: Elsevier.
Fletcher, Janet, Esther Grabe and Paul Warren
2004
Intonational variation in four dialects of English: the high rising tune. In:,
Sun-Ah Jun (ed.) Prosodic Typology. The Phonology of Intonation and Phrasing, 390410. Oxford: Oxford University Press.
Foolen, Ad
2006
Polysemy Patterns in Contrast: The Case of Dutch toch and German doch. In:
Karin Aijmer and Anne-Marie Simon-Vandenbergen (eds.), Pragmatic Markers
in Contrast, 5972. Amsterdam: Elsevier.
Francis, W. Nelson
1967
The Brown University Standard Corpus of English: Some implications for
TESOL. In: Betty Wallace Robinett (ed.), On Teaching English to Speakers of
Other Languages, 13135. Washington, D.C.: TESOL.
Fretheim, Thorstein
1981
Ego-dempere og alter-dempere. Maal og minne 12: 86100.
Furiasso, Cristiano and Knut Hofland
2007
The retrieval of false anglicisms in newspaper texts. In: Roberta Facchinetti
(ed.), Corpus Linguistics 25 Years on, 347364. Amsterdam/New York: Rodopi.
Gardner, Rod
1998
Between speaking and listening: the vocalisation of understandings. Applied
Linguistics 19.2: 204224.
Garnham, Alan
1986
Review of R. Wardhaugh, How Conversation Works. Journal of Literary Semantics 15: 153154.
Garnham, Alan, Richard Shillcock, Gordon Brown, Andrew Mill and Anne Cutler
1981
Slips of the tongue in the London-Lund Corpus of spontaneous conversation.
Geluykens, Ronald
1986
Questioning intonation: An empirical study into the prosodic feature rising
intonation and its relevance for the production and recognition of questions.
Antwerp Papers in Linguistics 48.
Geluykens, Ronald
1987a
Intonation and speech act type: An experimental approach to rising intonation
in declaratives. Journal of Pragmatics 11: 483494.
617
Geluykens, Ronald
1987b
Tails (right-dislocations) as a repair mechanism in English conversation. In:
Jan Nuyts and Georges de Schutter (eds.), Getting Ones Words into Line: On
Word Order and Functional Grammar, 119129. Dordrecht: Fortis.
Geluykens, Ronald
1988
On the myth of rising intonation in polar questions. Journal of Pragmatics 12:
467485.
Granger, Sylviane
1997
On identifying the syntactic and discourse features of participle clauses in academic English: native and non-native writers compared. In: Jan Aarts, Inge
de Mnnink and Herman C. Wekker (eds.), Studies in English Language and
Teaching in Honour of Flor Aarts, 185198. Amsterdam/Atlanta: Rodopi.
Granger, Sylviane and Stephanie Tyson
1996
Connector usage in the English essay writing of native and non-native EFL
speakers of English. World Englishes 15: 1727.
Greenbaum, Sidney (ed.)
1996
Comparing English Worldwide: The International Corpus of English. Oxford:
Clarendon Press.
Greenbaum, Sidney
1991
ICE: the International Corpus of English. English Today 28: 37.
Greenbaum, Sidney
1992
A new corpus of English: ICE. In: Jan Svartvik (ed.), Directions in Corpus Linguistics. Proceedings of Nobel Symposium 82, 171179. Berlin: Mouton de
Gruyter.
Gustafsson, Marita
1982
Textual aspects of topicalization in a corpus of English. ICAME News 6:
4676.
Gustafsson, Marita
1983
Fronting of adverbials in four genres of English. In: Sven Jacobson (eds.),
Papers from the Second Scandinavian Symposium on Syntactic Variation,
Gustafsson, Marita
1985
Topicalizations revisited. In: Jorma Tommola and T. Vinanen (eds.), Working
Papers in English Studies 6, 4350. Publications of the Department of English,
Turku: University of Turku.
Guy, Gregory and Julia Vonwiller
1989
The meaning of an intonation in Australian English. Australian Journal of Linguistics 4: 117.
Haan, Ferdinand de
2002
Strong modality and negation in Russian. In: Randi Reppen, Susan M.
Fitzmaurice and Douglas Biber (eds.), Using Corpora to Explore Linguistic
Variation, 91110. Amsterdam: Benjamins.
Haegeman, Lillian
1984
Pragmatic conditionals in English. Folio Linguistica 13: 485502.
1961
Categories of the theory of grammar. Word 17: 24192.
618
Gisle Andersen

1985
An Introduction to Functional Grammar. London: Edward Arnold.
Hardy, Donald E. and Amy Leuchtmann
1996
Topic versus cohesion in the prediction of causal ordering in English conversation. Discourse Processes 21: 237254.
Haslerud, Vibeke and Anna-Brita Stenstrm
1995
The Bergen Corpus of London Teenager Language (COLT). In: Geoffrey
Leech, Greg Myers and Jenny Thomas (eds.), Spoken English on Computer.
Transcription, Mark-up and Application, 235242. London: Longman.
Hasselgrd, Hilde
2006
Not now on Non-Correspondence between the Cognate Adverbs now and
n. In: Karin Aijmer and Anne-Marie Simon-Vandenbergen (eds.), Pragmatic
Hedstrm, Karl Erik
1984
A study of repairs in speech. In: Stockholm Papers in English Language and
Literature 4, 69101. Stockholm: Department of English, Stockholm University.
Heinz, Bettina
2003
Backchannel responses as strategic responses in bilingual speakers conversations. Journal of Pragmatics 35: 11131142.
Hofland, Knut and Stig Johansson
1986
The Tagged LOB Corpus: KWIC Concordance. Microfiche. Bergen: Norwegian Computing Centre for the Humanities.
Holmes, Janet
1988
Doubt and certainty in ESL textbooks. Applied Linguistics 9: 2144.
Hundt, Marianne, Carolin Biewer and Nadja Nesselhauf
2007
Corpus Linguistics and the Web. Amsterdam: Rodopi.
Hunston, Susan
2009
Semantic prosody revisited. In: Rosamund Moon (ed.), Words, Grammar,
Text Revisiting the Work of John Sinclair, 85103. Amsterdam: Benjamins.
Johannessen, Janne B. and Kristin Hagen (eds.)
2008
Sprk i Oslo: Ny forskning omkring talesprk. Oslo: Novus.
Johansson, Stig and Knut Hofland
1994
Towards an English-Norwegian parallel corpus. In: Udo Fries, Gunnel Tottie
and Peter Schneider (eds.), Creating and Using English Language Corpora,
2537. Amsterdam: Rodopi.
Johansson, Stig
2006
How well can well be translated? On the English discourse particle well and its
correspondences in Norwegian and German. In: Karin Aijmer and Anne-Marie
Simon-Vandenbergen (eds.), Pragmatic Markers in Contrast, 115138. Amsterdam: Elsevier.
Jucker, Andreas and Sara Smith
1998
And people just you know like Wow: discourse markers as negotiating strategies. In: Andreas Jucker and Yael Ziv (eds.), Discourse Markers: Descriptions and Theory, 171201. Amsterdam: Benjamins.
Jucker, Andreas and Yael Ziv (eds.)
1998
Discourse Markers: Descriptions and Theory. Amsterdam: Benjamins.
619
Jucker, Andreas
1993
The discourse marker well: a relevance-theoretical account. Journal of Pragmatics 19.5: 43552.
Kendon, Adam
1972
Some relationships between body motion and speech: An analysis of an
example. In: Aron Siegman and Benjamin Pope (eds.), Studies in Dyadic Communication, 177216. New York: Pergamon Press.
Kendon, Adam
1980
Gesticulation and speech: Two aspects of the process of utterance. In: Mary
Ritchie Key (ed.), The Relationship of Verbal and Nonverbal Communication,
Kennedy, Graeme
1998
An Introduction to Corpus Linguistics. London: Addison Wesley Longman.
Kilgarriff, Adam and Grefenstette, Gregory
2003
Introduction to the Special Issue on Web as Corpus. Computational Linguistics
29: 115.
Kirk, John
2002
Applications of the Stenstrm model of discourse structure. In: Angela Hasselgren and Leiv Egil Breivik (eds.), From the COLTs Mouth and Others,
Kjellmer, Gran
2009
Where do we backchannel? On the use of mm, mhm, uh huh and such like. International Journal of Corpus Linguistics 14.1: 81112.
Knowles, Gerry and Lita Lawrence
1987
Automatic intonation assignment. In: Roger Garside, Geoffrey Leech and
Geoffrey Sampson (eds.), The Computational Analysis of English. A Corpusbased Approach, 139148. London: Longman.
Knowles, Gerry, Briony Williams and Lita Taylor (eds.)
1996
A Corpus of Formal British English Speech. The Lancaster/IBM Spoken-English Corpus. London: Longman.
Knowles, Gerry, Lita Taylor and Briony Williams
1992
A Corpus of Formal British English Speech. London: Longman.
Knowles, Gerry
1995
Review of Stenstrm 1994. ICAME Journal 19: 120123.
Knowles, Gerry
1986
Tone grouping by numbers. In: Lancaster Papers in Linguistics 44. Lancaster:
Department of Linguistics and Modern English Language, University of Lancaster.
Kussmaul, Paul
1978
In fact, actually, anyway : Indikatoren von Sprechakten im informellen gesprochenen Englisch. Die Neueren Sprachen 27: 357369.
Kyt, Merja and Matti Rissanen
1988
The Helsinki Corpus of English Texts: Classifying and coding the diachronic
part. In: Merja Kyt, Ossi Ihalainen and Matti Rissanen (eds.), Corpus Linguistics Hard and Soft, 169179. Amsterdam: Rodopi.
Langford, David
1994
Analysing Talk. London: Macmillan.
620
Gisle Andersen
Leech, Geoffre
1992
Corpora and theories of linguistic performance. In: Jan Svartvik (ed.), Directions in Corpus Linguistics. Proceedings of Nobel Symposium 82, 105122.
Leech, Geoffrey, Greg Myers and Jenny Thomas (eds.)
1995
Spoken English on Computer. London: Longman.
Lenk, Uta
1998
Marking Discourse Coherence: Functions of Discourse Markers in Spoken
English. Tbingen: Narr.
Lindquist, Hans and Christian Mair (eds.)
2004
Corpus Approaches to Grammaticalization in English. Amsterdam: John Benjamins.
Loehr, Daniel P.
2004
Gesture and Intonation. PhD. Dissertation, Georgetown University, Washington, DC.
Loman, Bengt
1982
The segmentation problem in the study of impromptu speech. In: Nils Erik
Enkvist (ed.), Impromptu Speech: a Symposium,105130. bo: bo Akademi.
Louw, Bill
1993
Irony in the text or insincerity in the writer? The diagnostic potential of semantic prosodies. In: Mona Baker, Gill Francis and Elena Tognini-Bonelli (eds.),
Text and Technology. In Honour of John Sinclair, 157176. Amsterdam: Benjamins.
Mauranen, Anna
2004
Theyre a Little Bit Different Observations on Hedges in Academic Talk.
In: Karin Aijmer and Anna-Brita Stenstrm (eds.), Discourse Patterns in
Spoken and Written Corpora, 173197. Amsterdam: Benjamins.
Mauro, Tullio de, Federico Mancini, Massimo Vedovelli and Miriam Voghera
1993
Lessico di Frequenza dellItaliano Parlato. Milan: Etas.
McCarthy, Michael and Ronald Carte
2004a
Introduction. Journal of Pragmatics 36: 147148.
McCarthy, Michael and Ronald Carter
2004b
Theres millions of them: hyperbole in everyday conversation. Journal of
Pragmatics 36: 149184.
McCarthy, Michael
2002
Good listenership made plain: British and American non-minimal response
tokens in everyday conversation. In: Randi Reppen, Susan M. Fitzmaurice and
Douglas Biber (eds.), Using Corpora to Explore Linguistic Variation, 7390.
McCarthy, Michael
2003
Talking Back: Small Interactional Response Tokens in Everyday Conversation. Research on Language and Social Interaction 36.1: 3363.
McEnery, Tony and Andrew Wilson
2001
Corpus Linguistics. 2nd ed. Edinburgh: Edinburgh University Press.
McEnery, Tony and Richard Xiao
2004
The Lancaster Corpus of Mandarin Chinese: A Corpus for Monolingual and
Contrastive Language Study. Paper presented at LREC 2004. May 2004, Lisbon.
621
McEnery, Tony, Richard Xiao and Yukio Tonio

2006
Corpus-based Language Studies: An Advanced Resource Book. New York:
Routledge.
McNeill, David
1985
So you think gestures are nonverbal? Psychological Review 92.3: 350371.
McNeill, David
1992
Hand and Mind: What Gestures Reveal about Thought. Chicago: University of
Chicago Press.
Mey, Jacob L.
2004
FoI Corpus Linguistics Part III Introduction. Journal of Pragmatics 36: Editorial. 1519.
Milroy, Leslie
1987
Observing and Analysing Natural Language. Oxford: Blackwell.
Ming, Tao and Liang Chen
2010
A discourse-pragmatic study of the word order variation in Chinese relative
clauses. Journal of Pragmatics 42: 168189.
Mosegaard Hansen, Maj-Britt
1998
The Function of Discourse Particles. Amsterdam: Benjamins.
2006
A dynamic polysemy approach to the lexical semantics of discourse markers.
In: Kerstin Fischer (ed.), Approaches to Discourse Particles, 2141. Oxford:
Elsevier.
Nsslin, Siv
1984
The English Tag Question: A Study of Sentences Containing Tags of the Type
Isnt It?, Is It?. Stockholm: University of Stockholm.
Nevalainen, Terttu
1987a
A quantitative survey of nuclear tone variation in English. In: Proceedings of
the Eleventh International Congress of Phonetic Sciences. Vol 3, 180183.
Tallin: Institute of Language and Literature, Academy of Sciences of the Estonian S.S.R.
Nevalainen, Terttu
1987b
Adverbial focusing and intonation. Lingua 73: 141165.
Nevalainen, Terttu
1992
Intonation and discourse type. Text 12: 397427.
Norrick, Neal and Kerstin Fischer
2009
Listener activities and responses: Panel Introduction at the 11th International
Pragmatics Conference, Melbourne, 1217 July 2009.
Norrick, Neal
2004
Hyperbole, extreme case formulation. Journal of Pragmatics 36.9: 1727
1739.
Ochs, Elinor
1996
Linguistic resources for socializing humanity. In: John J. Gumperz and
Stephen Levinson (eds.), Rethinking Linguistic Relativity, 407438. Cambridge: Cambridge University Press.
Orestrm, Bengt
1977
Why /dhi/ Book? SSE Report, Survey of Spoken English, Lund University.
622
Gisle Andersen
Orestrm, Bengt
1982
When is it my turn to speak?. In: Nils Erik Enkvist (ed.), Impromptu Speech:
Orestrm, Bengt
1983
Turn-taking in English Conversation. Lund: Liber/Gleerup.
stman, Jan-Ola
1982
The symbiotic relationship between pragmatic particles and impromptu
speech. In: Nils Erik Enkvist (ed.), Impromptu Speech: a Symposium, 147
stman, Jan-Ola
1995
Pragmatic particles twenty years after. In: Brita Wrvik, Sanna-Kaisa, Hiltunen, and Risto Hiltunen (eds.), Organization in Discourse: Proceedings from
the Turku Conference, 95108. Turku: University of Turku.
Partington, Alan
1998
Patterns and Meanings: Using Corpora for English Language Research and
Teaching. Amsterdam: Benjamins.
Raymond, William D., Mark Pitt, Keith Johnson, Elizabeth Hume, Matthew Makashay,
Robin Dautricourt and Craig Hilts
2002
An Analysis of Transcription Consistency in Spontaneous Speech from the
Buckeye Corpus. ICSLP-02, Denver.
Renouf, Antoinette
1986
The exploitation of a computerised corpus of English text. In: Michele Rivas
(ed.), Actes du VIIIeme Colloque G.E.R.A.S, 123136. Paris: Universite de
Paris-Dauphine.
Renouf, Antoinette
2007
Corpus development 25 years on: from super-corpus to cyber-corpus. In: Roberta Facchinetti (ed.), Corpus Linguistics 25 Years on, 2750. Amsterdam/
New York: Rodopi.
Renouf, Antoinette and Andrew Kehoe (eds.)
2006
The Changing Face of Corpus Linguistics. Amsterdam: Rodopi.
Rissanen, Matti
1986
Variation and the study of English historical syntax. In: David Sankoff (ed.),
Diversity and Diachrony. 97109. Amsterdam: Benjamins.
Romero, Jesus
1994
Ahm, ehm you call it theme? A thematic approach to spoken English. Journal of Pragmatics 22: 495509.
Sajavaara, Kari
1982
Contrastive analysis and impromptu speech. In: Nils Erik Enkvist (ed.), Impromptu Speech: a Symposium, 195206. bo: bo Akademi.
Sandford, Daniel
2008
Discourse and metaphor: A corpus-driven inquiry. Corpus Linguistics and
Lingustic Theory 4: 209234.
Schelfhout, Carla, Petro-Arno Coppen, Nelleke Ostdijk and Frans Van Der Slik
2005
Interjections in Dutch: a corpus-based approach, available at
http://www.neerlandistiek.nl/publish/articles/000100/article.pdf.
Schmid, Hans-Jrg
2001
Presupposition can be a bluff: How abstract nouns can be used as presupposition triggers. Journal of Pragmatics. 33.10: 15291552.
623
Schneider, Stefan
2007
Reduced Parenthetical Clauses as Mitigators. Amsterdam: Benjamins.
2001
Studies in Interactional Linguistics. Amsterdam: John Benjamins.
Shastri, S.V.
1988
The Kolhapur Corpus of Indian English and work done on its basis so far.
ICAME Journal 12: 1526.
Sinclair, John M. (ed.)
1987
Looking up. An Account of the COBUILD Project in Lexical Computing. London: Collins ELT.
Sinclair, John M.
1966
Indescribable English. Inaugural lecture, University of Birmingham, 1966
(unpublished).
Sinclair, John M.
1991
Corpus, Concordance, Collocation. Oxford: Oxford University Press.
Sinclair, John M.
1996
The search for units of meaning. Textus 9: 75106.
Sinclair, John M.
2004
Trust the Text: Language, Corpus and Discourse. London: Routledge.
Sinclair, John M. and Malcolm Coulthard
1975
Towards an Analysis of Discourse: the English used by Teachers and Pupils.
1992
Towards an analysis of discourse. In: Malcolm Coulthard (ed.), Advances in
Spoken Discourse Analysis, 134. London: Routledge.
Sinclair, John M. and David Brazil
1982
Teacher Talk. Oxford: Oxford University Press.
1986
1995
Relevance: Communication and Cognition. 2nd ed. Oxford: Blackwell.
Stenstrm, Anna-Brita and Annette Myre Jrgensen
2009
Youngspeak in a Multilingual Perspective. Amsterdam: Benjamins.
Stenstrm, Anna-Brita, Gisle Andersen and Ingrid Kristine Hasund
2002
Trends in Teenage Talk: Corpus Compilation, Analysis and Findings. Amsterdam: Benjamins.
Stenstrm, Anna-Brita, Gisle Andersen, Kristine Hasund, Kristine Monstad and Hanne Aas
1998
Users Manual to Accompany The Bergen Corpus of London Teenage Language (COLT). Bergen: Department of English at the University of Bergen.
Stenstrm, Anna-Brita
1982
Feedback. In: Nils Erik Enkvist (ed.), Impromptu Speech: a Symposium,
1984
Questions and Responses in English Conversation. Malmo: Gleerup/Liber.
1986a
A study of pauses as demarcators in discourse and syntax. In: Jan Aarts and
Willem Meijs (eds.), Corpus Linguistics II. New Studies in the Analysis and
Exploitation of Computer Corpora, 203218. Amsterdam: Rodopi.
624
Gisle Andersen
1986b
What does really really do? Strategies in speech and writing. In: Gunnel Tottie
and Ingegerd Bcklund (eds.), English in Speech and Writing: A Symposium,
1988
Adverbial commas and prosodic segmentation. In: Merj Kyt, Ossi Ihalainen and Matti Rissanen (eds.), Corpus Linguistics Hard and Soft, 221236.
Amsterdam: Rodopi.
1990a
Adverbial commas and prosodic segmentation. In: Jan Svartvik (ed.), The London-Lund Corpus of Spoken English: Description and Research, 253266.
(Lund Studies in English 82.) Lund: Lund University Press.
1990b
Pauses in monologue and dialogue. In: Jan Svartvik (ed.), The London-Lund
Corpus of Spoken English: Description and Research. 211252. (Lund Studies
in English 82.) Lund: Lund University Press.
1991
Expletives in the London-Lund Corpus. In: Karin Aijmer and Bengt Altenberg (eds.), English Corpus Linguistics. Studies in Honour of Jan Svartvik,
1994
An Introduction to Spoken Interaction. London: Longman.
1995
Taboos in teenage talk. In: Gunnell Melchers and Beatrice Warren (eds.),
Studies in Anglistics, 7179. Stockholm: Almqvist and Wiksell.
2006
Taboo words in teenage talk: London and Madrid girls conversations compared. In: Clare Mar-Molinero and Miranda Stewart (eds.), Language Variation and Change, 115138. Amsterdam: Benjamins.
Stubbs, Michael
1984
Review of Sinclair and Brazil 1982. Applied Linguistics 5: 7174.
Stubbs, Michael
1996
Text and Corpus Analysis. London: Blackwell.
Stubbs, Michael
2001
Words and Phrases: Corpus Studies of Lexical Semantics. Oxford: Blackwell.
Svartvik, Jan (ed.)
1992
Directions in Corpus Linguistics. Proceedings of Nobel Symposium 82 Stockholm, 48 August 1991. Berlin: Mouton de Gruyter.
Svartvik, Jan and Randolph Quirk (eds.)
1980
A Corpus of English Conversation. Lund: Gleerup/Liber.
Svartvik, Jan
1980
Well in conversation. In: Sydney Greenbaum, Geoff Leech and Jan Svartvik
(eds.), Studies in English Linguistics, 167177. London: Longman.
Svartvik, Jan
1982
The segmentation of impromptu speech. In: Nils Erik Enkvist (ed.), Impromptu
Speech: a Symposium, 131145. bo: bo Akademi.
625
Svartvik, Jan
1987
Graphic English prosody. In: Ishrat Lindblad and Magnus Ljung (eds.), Proceedings from the Third Nordic Conference for English Studies, 335346.
Svartvik, Jan
1988
A computerized corpus of spoken English. In: Vinca Liedloff (ed.), Standardization in Computerized Lexicography, 3352. Saarbrcken: Institut der Gesellschaft zur Frderung der angewandten Informationsforschung e.V. an der
Universitt des Saarlandes.
Tagliamonte, Sali and Rachel Hudson
1999
Be like et al. beyond America: The quotative system in British and Canadian
youth. Journal of Sociolinguistics 3: 147172.
Taylor, Lita and Gerry Knowles
1988
Progress report on the Spoken English Corpus. In: Merja Kyt, Ossi Ihalainen and Matti Rissanen (eds.), Corpus Linguistics Hard and Soft, 237244.
Amsterdam: Rodopi.
Thavenius, Cecilia
1982
Exophora in English conversation: a study of third person pronominal reference. In: Nils Erik Enkvist (ed.), Impromptu Speech: a Symposium, 291305.
bo: bo Akademi.
Thomas, Jenny and Andrew Wilson
1996
Methodologies for studying a corpus of doctor-patient interaction. In: Jenny
Thomas and Mick Short (eds.), Using Corpora for Language Research, 92
109. London Longman.
2001
Tottie, Gunnel
1986
The importance of being adverbial. Adverbials of focusing and contingency in
spoken and written English. In: Gunnel Tottie and Ingegerd Bcklund (eds.),
English in Speech and Writing: A Symposium, 93118. Stockholm: Almqvist
and Wiksell.
Tottie, Gunnel and Carita Paradis
1982
From function to structure: Some pragmatic determinants of syntactic frequencies in impromptu speech. In: Nils Erik Enkvist (ed.), Impromptu Speech: a
Symposium, 307317. bo: bo Akademi.
Tottie, Gunnel
1991
Conversational style in British and American English: The case of backchannels. In: Karin Aijmer and Bengt Altenberg (eds.), English Corpus Linguistics.
Studies in Honour of Jan Svartvik, 254271. London: Longman.
Traugott, Elizabeth Closs and Bernd Heine (eds.)
1991
Van den Heuvel, Theo
1988
TOSCA: An aid for building syntactic databases. Literary and Linguistic Computing 3: 147151.
Venditti, Jennifer J. and Julia Hirschberg
2003
Intonation and discourse processing. In: Maria Josep Sol, Daniel Recasens
and Joaqun Romero (eds.), Proceedings of the 15th International Congress of
Phonetic Sciences, 107114. Barcelona, Spain.
626
Gisle Andersen
Verschueren, Jef
2009
Introduction: The pragmatic perspective. In: Jef Verschueren and Jan-Ola stman (eds.), Key Notions for Pragmatics, 127. Amsterdam: Benjamins.
Vicher, Anne and David Sankoff
1989
The emergent syntax of pre-sentential turn openings. Journal of Pragmatics
13: 8197.
Virtanen, Tuija
2009
Discourse linguistics meets corpus linguistics: theoretical and methodological
issues in the troubled relationship. In: Antoinette Renouf and Andrew Kehoe
(eds.), Corpus Linguistics, Refinements and Reassessments, 4966. Amsterdam: Rodopi.
Walker, Gareth
2004
The phonetic design of turn endings, beginnings, and continuations in conversation. PhD thesis, Department of Language and Linguistic Science, University of York.
Walsh, Steve, Anne OKeeffe and Michael McCarthy
2008
post-colonialism, multi-culturalism, structuralism, feminism, post-modernism and so on and so forth: A comparative analysis of vague category
markers in academic discourse. In: Annelie del and Randi Reppen (eds.),
Corpora and Discourse: The Challenges of Different Settings, 930. Amsterdam: Benjamins.
Waltereit, Richard
2006
The rise of discourse markers in Italian: a specific type of language change. In:
Kerstin Fischer (ed.), Approaches to Discourse Particles, 6176. Amsterdam:
Elsevier.
Wardhaugh, Ronald
1985
How Conversation Works. Oxford: Basil Blackwell, London: Andre Deutsch.
Westney, Paul
1986
How to be more or less certain in English: Scalarity in epistemic modality. International Review of Applied Linguistics 24: 311320.
Wichmann, Anne
1994
F0 troughs and prosodic phrasing: an investigation into the linguistic information contained in a speakers base-line when reading. Hermes 13:
205212. Aarhus School of Business.
Wichmann, Anne
1996
Prosodic style: a corpus-based approach. In: Gerry Knowles, Anne Wichmann
and Peter Alderson (eds.), Working with Speech. Perspectives on Research into
the Lancaster/IBM Spoken English Corpus, 168188. London/New York:
Longman.
Wichmann, Anne
2000
Intonation in Text and Discourse. London: Longman.
Wichmann, Anne
2004
The intonation of Please-requests: a corpus-based study. Journal of Pragmatics 36: 15211549.
Wichmann, Anne
2005
Please: from courtesy to appeal: the role of intonation in the expression of attitudinal meaning. English Language and Linguistics 9.2: 229253.
627
Wichmann, Anne
2007
Corpora and spoken discourse. In: Roberta Facchinetti (ed.), Corpus Linguistics 25 Years on, 7388. Amsterdam/New York: Rodopi.
Wierzbicka, Anna
1991
Cross-cultural Pragmatics: The Semantics of Human Interaction. Berlin:
Mouton de Gruyter.
Yngve, Victor
1970
On getting a word in edgewise. In: Papers from the 6th Regional Meeting, Chicago Linguistic Society, 567577. Chicago: Chicago Linguistic Society.
22.

Christoph Rhlemann
Introduction
Due to the massive dependence of pragmatic phenomena on context, corpora, as a
unfit for use in pragmatic research. Nonetheless, corpus linguistic analyzes, both
into key pragmatic phenomena. The aim of this paper is to illustrate key quantitative corpus studies into phenomena of pragmatic interest. The paper is divided
into six sections. The opening section addresses the question of how context-sensitive corpora are. Section 2 presents a case study into semantic prosody, an attitudinal phenomenon at the semantics/pragmatics interface. Section 3 presents a
relevance-theoretic study of the pragmatic marker like. Section 4 is concerned with
quantitative studies on reference. Section 5 introduces corpus research into speech
acts. The concluding section looks to the future, outlining recent attempts at building multimodal corpora.
1.
Pragmatics and quantitative corpus linguistics: a troubled

relationship?
Pragmatics is concerned with meaning in context. Because speakers can mean

more than they say, pragmatics is the art of the analysis of the unsaid (Mey 1991:
245). The relationship between pragmatics, thus understood, and corpus linguistics
is seen by some as a troubled one. The reason is simple: corpora record text, not
meaning, and they record context only crudely, particularly in spoken corpora.
Spoken corpora are based on transcriptions made from audio recordings. Since
tape recorders cannot filter out non-speech noises, thus selectively listening to
speech as humans can many, and often large, stretches of corpus transcripts
may be inaudible or unintelligible. Further, the information provided in spoken
corpora about the contexts in which the spoken texts were produced is fairly minimal; we get to know who (sex, age, class, etc.) was talking to whom when. Moreover, other types of context are recorded only abstractly: we learn in what kinds of
setting the talk occurred, or what type of interaction it was. Yet another type of abstract context, which is considered so crucial that in most large corpora the data
are categorized according to it, is type of situation, or register, such as academic
writing, fiction, and conversation. The types of context largely missing from
630
Christoph Rhlemann
spoken corpora include the almost infinite wealth of concrete situational, nonverbal, and social context that conversationalists in their specific contexts of situation
are connected to (cf. Cook 1990). In sum, we may have impoverished textual evidence, and only rudimentary or abstract contextual evidence. Therefore, corpora
have long been seen by some researchers as unfit for use in pragmatic research.
Indeed, some pragmatic features inevitably escape corpus linguistic analysis.
This is because in part-of-speech (POS)-tagged corpora, only those phenomena
can be studied fully whose lexical form(s) and pragmatic function(s) display a
straightforward one-to-one relationship. This relationship is found, for example, in
the words sorry and pardon, which are regularly included in apologies (cf. Jucker
2009). The form-function match is already weaker in compliments which need not
necessarily be realized using typical conventionalized patterns (cf. Jucker et al.
2008). Where there is a complete form-function mismatch, as in cases of conversational implicature, a quantitative corpus study will be useless: what listeners take
to be implicated in an utterance cannot be retrieved exhaustively from a corpus but
can only be inferred (post hoc) with varying degrees of confidence.
What a corpus can do even in those cases where the form-function mismatch of
a phenomenon prevents exhaustive searches, is to provide the analyst with illustrative examples that are not only attested and, in this sense, authentic but also embedded in their co-texts, thus giving some evidence of the context in which they
were used. Such corpus illustrations can usefully complement, or even replace, the
invented and often completely decontextualized examples that have formed the
basis of much pragmatic enquiry.
Another approach to studying pragmatics corpus-linguistically is to use pragmatically-annotated corpora. For example, a small subcorpus of the Michigan Corpus of Academic Spoken English (MICASE) is tagged for some speech acts (cf.
Maynard and Leicher 2007) and in the Corpus of Verbal Response Mode (VRM)
Annotated Utterances, all utterances are coded twice: once for their literal meaning
and once for their pragmatic meaning, using a principled taxonomy of speech acts
(cf. Stiles 1992). However, the number of such pragmatically-annotated corpora is
still small and they are coded for selected aspects of pragmatic interest only.
Given these limitations, it may be surprising that quantitative corpus analyses
of pragmatic phenomena have grown into a large body of literature and produced a
wealth of new insights. In the following sections I will present some corpus linguistic studies into pragmatic units. The approach will be selective: only a few key
studies can be presented. References to related studies, however, will be provided.
Unless otherwise indicated, the illustrative examples will be taken from the British
National Corpus (BNC), a 100-million-word corpus of contemporary British English (cf. Hoffmann et al. 2008).
2.
631
The semantics/pragmatics interface: semantic prosody
One of the clearest strengths of corpus linguistics is the analysis of what Sinclair
(1991) refers to as the idiom principle, that is, broadly, how language patterns at
the level of phraseology. A very large body of corpus linguistic literature points to
idiom phenomena such as collocation, collostruction, colligation and so forth as
important building blocks of phraseological patterning. Another, crucial, phraseological phenomenon is semantic prosody, which I will focus on in this section.
Semantic prosodies probably best illustrate Channells (2000) claim that some
pragmatic phenomena can only be revealed in studies of large corpora. This is because semantic prosodies are normally hidden not only from introspection but also
from observation of small numbers of examples. I will first briefly outline essential
characteristics of semantic prosody and then illustrate the phenomenon in a case
study of BREAK out.
Semantic prosody is closely related to semantic preference, which is defined
by Stubbs as the relation, not between individual words, but between a lemma or
word-form and a set of semantically related word forms (2002: 65). Stubbs cites
the example of the adjective large, which often collocates with nouns denoting
quantity and size such as number, scale, part, amounts, etc. Semantic prosody,
on the other hand, can, as a rule of thumb (Partington 2004: 149), be seen as a subcategory of semantic preference. The distinguishing feature of semantic prosody is
that the type of semantic preference is related to polarity: the collocates of a word
are either typically positive or, more often, negative. Because of the inherently
evaluative nature of semantic prosody, alternative terminologies have been proposed, such as pragmatic meaning (Channell 2000), emotive prosody (Bublitz
2003) or evaluative prosody (Morley and Partington 2009). Further, because of the
evaluation they convey, semantic prosodies are said to be on the pragmatic side of
the pragmatics/semantics continuum (Sinclair 2004: 34).1 Bublitz (1996) and Morley and Partington (2009) also note the cohesive role of semantic prosody, which the
latter describe as the mechanism which shows how one elemental type of meaning evaluative meaning is frequently shared across units of discourse and, by ensuring consistency of evaluation or evaluative harmony, plays a vital role in keeping
the discourse together, in its cohesion (2009: 139). An illustrative example from
the Corpus of Contemporary American English (COCA) (cf. Davies 2009) is (1):
(1) Hello, everyone. I am Dr. Sanjay Gupta. Your health, there is nothing more important. # And now that winter weather has set in, there are certain health and
safety issues tat [sic] you need to be more concerned about, including how to
avoid succumbing to sickness. (COCA: CNN_YourHealth 2002)
This example involves the phrasal verb SET in, of which Sinclair (1991) observed
that its subjects refer to unpleasant states of affairs (Sinclair 1991: 74). Sinclair
cites as main vocabulary for SET in words like rot, decay, malaise, despair, ill-will,
632
Christoph Rhlemann
etc. In (1), the subject is winter weather, a rather neutral phrase. However, winter
weather is negatively evaluated. This becomes clear when we consider the context:
health and safety issues, need to be concerned about, and succumbing to sickness
are clear indices of the presence of negative evaluation. The evaluative harmony, to
use Partingtons term, in the excerpt is thus ensured by the accumulation of
negative items. See also Louws (1993: 173) observation that in many cases semantic prosodies hunt in packs and potentiate and bolster one another, an observation Bublitz (2003: 387) refers to as the need to establish a common emotive
ground by accumulating equi-polar means.
Excerpt (1) is illustrative of yet another crucial aspect of semantic prosody: it
can serve as an effort-saving device lighten[ing] the burden on the listener and
free[ing] the speaker of the tedium of labouring a point (Morley and Partington
2009: 144). This is, in Louws (1993: 157) words, because semantic prosody is defined as an aura of meaning with which a form is imbued by its collocates. In other
words, because SET in typically co-selects negative subjects, the negativity of its
collocates colours the meaning of SET in in such a way that it is perceived as
negative even if it co-occurs with a subject that is not clearly negative, such as
winter weather. Louws notion of imbuing has met with criticism. Whitsitt (2005),
for example, argues that there is no evidence for assuming that we can see the results of a diachronic process of imbuing (2005: 296; emphasis in original; see also
Bednarek 2008; Morley and Partington 2009). Whitsitt cites as evidence words
such as alleviate and heal, which habitually co-occur with clearly unpleasant words
and yet certainly [do] not come to have an unpleasant meaning because of that
company (2005: 297). On the other hand, the fact that speakers/writers can achieve
certain rhetorical effects such as irony by deliberately deviating from typical collocational patternings, as in outbreak of sanity, shows that the underlying semantic
prosody clearly persists, even and especially, in collocational clashes (Morley and
Partington 2009: 150). How can this evaluative persistence be explained?
An answer that has recently been given is the theory of priming (Hoey 2005).
As a word, such as SET in, is learnt through encounters with it in speech and writing, it is loaded with the cumulative effects of those encounters such that it is part of
our knowledge of the word that it co-occurs with other words (Hoey 2003), such
as, in the case of SET in, heavily negative items like rot, decay and so on. That is, in
(1), the speaker need not spell out that winter weather is taken as bad, for example
by adding a negative adjective such as cold, severe, or harsh. S/he can trust the addressee to know the primings of SET in, that is, a set of instructions on how to use
it, how it normally interacts with other items. Among these primings is a words
semantic prosody, an instruction, which tells the reader when you find me in a
text, read the surrounding discourse in a favourable/infavourable light, unless
theres something else around which tells you not to (Partington forthcoming).
The evaluation conveyed by semantic prosodies may be obvious or hidden to
the naked eye. Morley and Partington (2009: 151) suggest that it is best con-
633
sidered as a cline, with items such as murder and venerable on the overt-evaluation end expressing fairly clearly unfavourable and favourable evaluation, items
such as peddle and fraught with somewhere in the middle, and items such as SET in
and not + BUDGE (see below) occupying the covert-evaluation end where the
evaluation can only be identified using corpus linguistic methods.
The linguistic analysis and description of semantic prosody faces several challenges. One problem is that while it may be easy to achieve a consensus that words
such as rot, decay etc. describe negative states of affairs, the detection of negative
or positive semantic prosody may be much less straightforward in other cases.
Whether a word is used and interpreted positively or negatively depends in part on
the wider context in which it is used and in part on the speaker who uses it. Stubbs
(1996) cites the example of intellectuals, whose collocates e.g., activist, student,
leftwing, liberal may have a negative ring in some circles but be positively
evaluated in others (cf. Bednarek 2008: 122). Semantic prosody therefore cannot
be objectively derived from corpus data and requires a lot of inference on the part
of the analyst (Bednarek 2008: 132).2
A second problem is the evaluative inconsistency of semantic prosodies. Prosodies can be switched off or even reversed. Morley and Partington (2009: 149)
discuss the example of not + BUDGE, which usually carries a negative prosody.
However, under certain circumstances, for example when the subject is first person, it can be used favourably. An example of positive evaluation in the use of not
+ BUDGE is given in Hunston (2007):
(2) The Prime Minister rejected resounding calls for the resignation of the government, I will not budge, he said
Further, the distribution of semantic prosody and word form may be asymmetrical
(Bublitz 1996). That is, it may be misleading to claim that verb X or noun Y has
positive or negative prosody because in many cases different forms of one and the
same word have different evaluative tendencies. Consider the phrasal verb BREAK
out: when used transitively, as in Just as linotype operators at the Sun were breaking
out the champagne to celebrate the arrival of 1,000 pay packets (CHU 308), its
prosody is clearly positive (cf. Louw 2000); when used intransitively, as in Fires keep
breaking out (A18 1206), its prosody is decidedly negative (cf. the discussion below).
Another problem is that semantic prosodies can be subject to variation across
registers. OHalloran (2007: 15) demonstrates that, for example, the past tense
form erupted has largely positive associations in the sports report register, but
largely negative ones in the hard news register. That is, the non-register specific
concept of semantic prosody may, in some cases, have to be replaced by the concept of register prosody, which indicates that some prosodies have probabilistic
relationships to register (OHalloran 2007: 4; see also Bublitz 1996).
Finally, semantic prosody cuts across the traditional lexis/grammar dichotomy.
The GET-passive has been shown to have a negative prosody in that it typically at-
634
Christoph Rhlemann
tracts past participles sharing an adversative core meaning (Carter and McCarthy
1999; Stubbs 2002; Rhlemann 2007b). Typical examples are GET stuck, GET
caught, and GET killed. Further, semantic prosody may also be observed at a level
below the word. Rhlemann (2006) demonstrates that the prefix dys-, as in dyslexia, dysfunction, dysplasia, etc., as well as the prefix dis-, when used as a productive morpheme, as in disease, disappear, disabled, etc. form words that have
negative prosody.
In the remainder of this section a case study into the pattern (inanimate) N +
(intransitive) BREAK out will bring into relief some of the above said.
A search for the verbal lemma BREAK immediately followed by out finds
1,126 occurrences in the BNC. (3) shows a random excerpt from the concordances.
As is suggested by the inanimate nouns (underlined in (3)) preceding the string,
negative things such as hostilities, war, mayhem etc. break out.
(3) if hostilities ever broke out. Weve
When war broke out he returned to
complete mayhem broke out. Hats were flung
a fight broke out at the Zuwaya
resembling panic broke out among the ladies
Is this association with negative items systematic? Table 1 presents the 15 most
frequent nouns preceding BREAK out in the BNC; right hand column gives the
Z-scores, which measure the strength of attraction between collocates.
Table 1.
Top 15 nouns collocating with BREAK out
No.
Collocate
Freq.
Z-score
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
war
fire
fighting
fight
row
world
violence
blaze
riot
argument
hostility
scuffle
rioting
way
sweat
259
89
46
39
36
32
20
17
17
17
16
14
13
12
11
167.8
82.8
113.8
75.3
48.5
12.6
29.4
72.4
44.7
16.3
43.5
132.3
82.1
1.1
37.2
635
As can be seen from Table 1, BREAK out most commonly collocates with inanimate nouns that share physical violence and, to a lesser degree, fire meanings. War,
the most extreme manifestation of violence, is not only by far the most frequent
collocate but also has the highest Z-score; note also that the 259 occurrences of war
+ BREAK out account for more than a fifth of all occurrences of BREAK out. War is
thus a very strong collocate of BREAK out. Other nouns in the violence group include fighting (rank 3), fight (4), row (5) etc.; world, ranked 6th, occurs in combination with war forming the compound world war. The fire group includes the
nouns fire (2) and blaze (8). Inspection of concordance lines suggests that it is invariably destructive, not warming or cosy, fire that breaks out. Since destructive
fire is an inevitable concomitant of war, we might subsume the fire meanings under
the heading of the violence meanings, defining the negative prosody of BREAK out
thus: BREAK out co-selects inanimate nouns that typically express violence (including destructive fire) meanings. BREAK out illustrates that the concepts of semantic prosody and semantic preference are not mutually exclusive in that the
phrasal verb is a semantic prosody with a clear semantic preference.
Interestingly, there are six occurrences of peace + BREAK out in the BNC. Inspection of the concordance lines suggests that covert evaluation is the driving
force behind these uses. Consider (4) (from the Economist, 1991):
(4) Where will the dollar head when peace breaks out? (ABH 2961)
In (4), the prospect of peace (at the time of the Gulf war) is taken as a threat to the
dollar, which has risen before the war but, it is suggested, might fall when the war
is over. That peace may have bad consequences for the dollar is not made explicit.
Rather, the negative aura of meaning surrounding BREAK out is tacitly projected
onto peace.
In sum, prosodies are a useful resource for the diagnosis of overt and covert
speaker evaluation. Further, as Louw argues, there is the same well-calculated collocational deviance in instances of irony (Louw 1993). The prospect, then, is that
semantic prosodies may help to computationally uncover irony, a pragmatic phenomenon that has so far escaped quantitative empirical study. Moreover, semantic
prosodies may assist in the study of persuasion: Propaganda, advertising and promotional copy will now be gradable against the semantic prosodies of the whole
language (Louw 1993: 173). Exploiting the diagnostic potential of semantic prosodies for these kinds of speaker meaning has only just begun and more insights
may be expected from this promising avenue of research.
3.
Pragmatic markers: like in a relevance-theoretic perspective
Pragmatic markers are words or phrases that do not add so much to the propositional content of utterances as they metalingually flag how discourse relates to
636
Christoph Rhlemann
other discourse. Pragmatic markers thus play a crucial role in facilitating processes
of pragmatic inference.
Because they are lexically relatively fixed and thus relatively easily retrievable
from a corpus, corpus research into pragmatic markers has been extremely productive.3 This section presents Andersen (2001), a corpus study on a marker that is
particular in that it attends to several discourse functions and, not surprisingly, has
attracted a wealth of recent research: like. Andersens study, which is based on the
Bergen Corpus of London Teenage Language (COLT), is notable because the approach to like is informed by relevance theory (Sperber and Wilson 1995). This
theory holds that human communication is geared towards the maximisation of relevance such that from the wealth of information and stimuli in our cognitive environments we tend to select that information and those stimuli that are most relevant to us in any given communicative situation.
Andersen identifies five broad functions of like used as a pragmatic marker.
They are illustrated in (5)-(9). In (5), like carries out an approximation function indicating that the numerical information is approximate rather than precise; in (6),
like is in an exemplification function. The common denominator of uses of like classified as approximation and exemplification is that they involve non-identical resemblance between the encoded and the communicated concepts (Andersen 2001:
237). The third function Andersen identifies is metalinguistic use, which concerns
the speakers relation to the proposition in terms of its formal linguistic characteristics (Andersen 2001: 243); in example (7), like marks a particular word choice
(reminder) as potentially inexact or inappropriate. Another function like can carry
out is as interpretive use, a term Andersen prefers over the more commonly used
term quotative on the grounds that like may be used to preface not only presentations of speech, as in (8), but also gestures and facial expressions that can be seen
as metarepresentations of speaker attitude (Andersen 2001: 254; see also Buchstaller 2008). Finally, Andersen stresses the role of like as a hesitational/linking device, that is, its function in accompanying false starts, self repairs and cut-off utterances and in providing a discourse link between syntactically distinct units of
discourse (Andersen 2001: 259). In (9), for example, like acts as a pause filler granting the speaker planning time (note the co-occurrence with the hesitation form er):
(5) I mean Ive been in two shops now theres fifty pound difference like, you
know (KB2 2401)
(6) think its the way he looks, like, if you know what I mean, you know like []
pull his face and like, look over glasses (KB1 768)
(7) PS007 >: Oh yes, I was assuming that erm a I do, I didnt really set it
out as a formal agenda just as a
PS002 >: Just as a note
PS007 >:a reminder, like. (KB0 1337)
637
(8) Yeah thats what I, why, thats what I said to Susanna and she was like dont be
ridiculous! (KP5 2163)
(9) PS03T >: well you should wipe the outside of the frame didnt you?
PS03S >: er, yes, like, you like just flick round it (KBB 6472)
The functional distribution of like in the COLT is shown in Table 2:
Table 2.
Functional distribution of like (cf. Andersen 2001: 266)
Function
Approximation
Exemplification
Metalinguistic focus
Interpretive use (quotative)
Hesitational/discourse link
21
19
19
7
35
As is shown in Table 2, the hesitational/linking functions in the corpus are most

salient, followed by approximation, exemplification, and metalinguistic use. The
quotative function, by contrast, is relatively infrequent. This may be contrary to expectations, given the higher rates for quotative like in Tagliamonte and Hudson
(1999). The discrepancy between Andersen and Tagliamonte and Hudson, however, may be due to the fact that quotative like is a recent development that has
been spreading to many regional varieties of English (cf. Buchstaller 2008) and
that Tagliamonte and Hudsons data were collected in 1996 whereas the COLT data
stem from the early 1990s.
Do all of the different functions of discourse marker like have a common root?
Andersen (2001) interprets discourse marker like as a looseness marker, that is,
as a signal that the relation between an utterance and its underlying thought is
not a one-to-one relationship, but a relation of non-identical resemblance (2001:
230). As such, it is a highly co-constructive item inviting recipients to collaborate
in the negotiation of meaning.
4.
Reference
Reference undoubtedly touches on one of the most fundamental questions concerning language: how can we speak and, in speaking, communicate something
meaningful about things, people, and states of affairs? The answers linguists and,
more importantly, philosophers have suggested are by far too varied and complex
to go into in sufficient detail in this short section (see Schwarz-Friesel and Consten
this volume). A few remarks must suffice.
638
Christoph Rhlemann
It is often claimed that referring expressions refer to their referents, such that,
for example, the expression the present U.S. American president refers to
Barack Obama. On this view, the expression denotes an object in the real world;
see Russells (1905) influential paper, entitled On denoting. Denotation, in
Russell, is taken to be a relation between an expression, considered in abstraction, and the thing that is the expressions referent or denotatum (Lycan 2008:
19). Russell distinguishes between meaning and denotation. An expression such
as the present U.S. American prime minister would have to be assigned meaning because it could be perfectly understood, but, since the US political system
does not provide for a prime minister, it would fail to denote anything. In Russells view, the sentence The present U.S. American prime minister is wise
would be considered to be false. A number of objections have been brought
against Russells theory of definite descriptions, chiefly by Strawsons (1950) article aptly entitled On referring. Strawson thought of referring not as an abstract relation between an expression and an object but argued that referring is
not something an expression does; it is something that someone can use an expression to do (1950: 326). The variable that he brings into the equation is the
context of an utterance [] and by context I mean, at least, the time, the place,
the situation, the identity of the speaker, the subjects which form the immediate
focus of interest, and the personal histories of both the speaker and those he is
addressing (1950: 336). Whether the present U.S. American president is used
to refer to Mr Obama or his predecessor or his successor or any other U.S.
American president crucially depends on the circumstances of its use. Also, the
sentence The present U.S. American prime minister is wise would not be seen
as lacking truth-value; since it builds on a false presupposition, the question
whether the statement is true or false simply does not arise. On this view, which
links reference intimately to context and, hence, inference, reference is deeply
pragmatic.
The following two subsections section present corpus-based research into two
types of referring expressions: definite noun phrases and deixis.
4.1.
Reference through definite noun phrases
One type of material that corpus linguists standardly work with are frequency lists,
that is, lists in which the words in a corpus are ranked in order of their frequency in
that corpus.
Table 3 displays what are, according to Kilgarriff (1998), the 20 most frequent
items in the conversational subcorpus and the written subcorpus of the BNC.
Table 3.
639
20 most frequent items in the conversational subcorpus (C) and the written
subcorpus (W) of the BNC
C
Rank
Frequency
Item
Tag
Frequency
Item
Tag
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
167,640
135,217
128,165
115,247
92,239
90,886
77,611
68,846
62,382
58,810
48,322
47,391
43,977
42,241
41,654
38,515
35,156
34,901
34,837
34,477
i
you
it
the
s
and
nt
a
that
yeah
he
to
they
do
oh
what
is
of
was
in
pnp
pnp
pnp
at0
vbz
cjc
xx0
at0
dt0
itj
pnp
to0
pnp
vdb
itj
dtq
vbz
prf
vbd
prp
5,776,384
2,789,403
2,421,302
1,939,617
1,695,860
1,468,146
1,892,937
1,845,350
1,839,964
1,834,957
1,768,898
1,606,027
1,605,749
1,603,178
1,590,305
1,580,267
1,561,041
1,490,673
1,435,574
1,426,207
the
of
and
a
in
to
is
to
was
it
for
with
he
be
on
i
that
by
at
you
at0
prf
cjc
at0
prp
to0
vbz
prp
vbd
pnp
prp
prp
pnp
vbi
prp
pnp
cjt
prp
prp
pnp
Table 3 reveals clear differences in the ways that language users refer in conversation (C) and writing (W) respectively. Among the 20 most frequent items in C,
there are (i) more personal pronouns (the pronoun they is not included in the top 20
in W) and (ii) far less NP-related items (prepositions and the definite and the indefinite article) than in W: there are ten NP-related items in W but only four in C.
Table 3 further shows that the definite article the is by far the most common word
in writing (roughly twice as common as the next one, the preposition of). Indeed,
the is not only the most frequent word in writing but also the most frequent word in
the whole of the BNC (cf. Kilgarriff 1998) and in most other general corpora, such
as the Cambridge International Corpus (CIC) (cf. McCarthy 1998) or the Bank
of English (BoE) (cf. Sinclair 1999). With its roughly 6 million occurrences in
the 90-million-word written subcorpus of the BNC, the accounts for more than six
percent of all word tokens in that subcorpus. It is ironic that, a good 100 years on,
640
Christoph Rhlemann
Russells verdict that to the philosophical mathematician [the] is a word of great

importance (1919: 167) should be empirically confirmed by corpus frequency
counts.
This evidence from the BNC suggests two major interpretations: (i) the most
common referring technique in writing is the use of both the definite NP and, less
importantly, the indefinite NP, while (ii) in conversation, reference is expressed
most commonly by means of pronouns. While this latter finding will be discussed
in the next subsection (5.2), the remainder of this subsection takes a closer look at
the use of the definite article both in conversation and writing.
The overriding function of the definite article is to specify that the referent of
the noun phrase is assumed to be known to the speaker and the addressee (Biber et
al. 1999: 263). That is, broadly speaking, the functions as a marker of given information. It does so in a variety of ways. In a study carried out on the Longman
Spoken and Written English (LSWE) Corpus, Biber et al. (1999) identify the following reference patterns of definite NPs marked by the.4 They are illustrated by
relevant examples:
(10) Anaphoric: A MAN died and a girl was badly injured when fire ripped
through a house yesterday. The girl, who had been clinging to a
third-floor window ledge, fell just as firemen were about to grab
her. (CBF: 2921)
(the girl is understood as the girl that was badly injured when fire ; reference is backward-looking)
(11) Indirect anaphoric:
A woman died yesterday after being knocked down by a shoplifter fleeing with a 2.58 descaler. Mrs Lillian Amelia Smith,
81, sustained a fractured skull at the store at Newham, east London, on Friday. Police are treating the incident as murder.
(AKH: 260)
(the incident can be identified via inference as the events described in the
preceding report)
(12) Cataphoric: The next main point is about the complexity of the system.
(HHW: 3638)
(the complexity in question is identifiable via the of-phrase following it; reference is forward-looking)
(13) Situational: I think there might be parcel for you at the door is it? (FPU: 358)
(based on frame knowledge the reference of door is understood as the
front door where parcels are normally delivered)
(14) Generic:
Just for once, a Frenchie has conceded that the Brits do something better. (A0C 1386)
641
(the denotes the whole class of Brits without specific reference to particular Brits)
(15) Idiom:
But, quite frankly, whats the point? (A5Y: 230)
(the is an integral part of the idiom whats the point?; it has no referring or
denoting function; see also Searle 1969: 72)
Intriguing results emerge from the analysis of the extent of use of these patterns in
different registers by Biber et al. (1999). The registers considered include three
written registers (fiction, news, and academic writing) and conversation. The results are summarized in Table 4:
Table 4.
Percentage use of reference patterns in four registers (C: conversation;

F: fiction; N: news reportage; A: academic writing) (cf. Biber et al. 1999: 266)
Situational
Anaphoric
Indirect anaphoric
Cataphoric
Generic
Idiom
Uncertain
55
25
5
5
<2.5
<2.5
10
10
30
10
15
<2.5
<2.5
15
10
30
15
30
5
<2.5
5
10
25
15
40
5
<2.5
5
The findings presented in Table 4 allow for a number of observations. Generally,

there are marked cross-register differences in the use of the reference patterns.
Specifically, anaphoric reference, which may intuitively be seen as the major reference type of definite NPs, accounts for less than a third in all four registers. Cataphoric reference represents, respectively, 30 percent in news reportage, 40 percent in academic writing, but only five percent in conversation. Situational
reference, conversely, accounts for 55 percent in conversation while it is found in
only ten percent in the written registers.
The analysis in Biber et al. (1999) also enables us to understand more deeply
the relative frequencies of the definite article in the word frequency list discussed
above. We saw that the is by far the most frequent word in the written part of the
BNC but less highly ranked in the conversational part of the BNC. Applying the
Biber et al. analysis to these frequency-based rankings, we can now see that definite NP reference is not only relatively less important in conversation, where it is
second to reference through personal pronouns, than in writing, where it is by far
the most important reference type. We can also see that the difference in the use
of the definite NP between conversation and writing is in fact much more dramatic because reference through definite NP fulfills different functions in the two
modes: while it is largely (indirect) anaphoric and cataphoric and, hence, endo-
642
Christoph Rhlemann
phoric in writing, it is overwhelmingly situational and, hence, exophoric in conversation.

Another type of exophoric reference is through the use of deixis. Its use in
discourse presentation will be the focus of the next section. That section will also
facilitate a more detailed understanding of the overrepresentation of personal pronouns in conversation we observed in Table 3 above.
4.2.
Deixis and discourse presentation5
Deictic reference is a prime example of exophoric reference because establishing

referents of deictic words necessarily requires extra-linguistic context. For
example, the references of the person-deictic word I and the time-deictic word tonight are completely dependent on the speech situation in which the two words are
uttered. Deictic words presuppose a deictic centre (origo) relative to which they
are computed. The deictic centre is in most cases associated with the current
speaker. It can also be seen as the viewpoint from which the speaker refers to the
dimensions of person, time, and place (Lyons 1977: 638). This viewpoint constrains the use of deictic words such that person, time, and place deictics normally
conform to this, one, viewpoint, thus forming a deictic system (Levinson 1983:
68).
One area of language use in which speakers can use deictic words that do not
conform to their own deictic system is discourse presentation, alternatively referred to as speech reporting and constructed dialogue. This will be explained
with reference to corpus research carried out by McIntyre et al. (2004), a study
based on the Lancaster Speech, Writing and Thought Presentation Spoken Corpus,
a small, balanced corpus of contemporary spoken English drawn from the conversational subcorpus of the BNC and oral history archives from Lancaster University
containing elicited interviews. The authors propose categories for the presentation
of not only speech but also thought and writing. For space limitations, only the categories as well as the studys initial results for speech presentation will here be reported on.
McIntyre et al.s (2004) model of speech presentation provides for six major
categories. With reference to the examples listed below, the categories include
the two direct categories Direct, as in (16), and Free Direct, as in (17). Both
modes represent speech in the form in which it is directly manifest to a listener
(Leech and Short 1981: 345) in an anterior situation. A distinction between the
two is that Direct requires a preceding reporting clause, whereas Free Direct is not
accompanied by a reporting clause. The categories further include Indirect, as in
(18), and Free Indirect, as in (19). Like Free Direct, Free Indirect has no accompanying reporting clause. In (Free) Indirect mode the propositional content of the
original speech is specified, but no claim is made to present the words and structures originally used to utter that proposition (McIntyre et al. 2004: 61). Further,
643
the categories include Representation of Speech Act, as in (20). This category

presents the illocutionary force of an utterance or text (part) with an optional
noun or prepositional phrase indicating the topic (McIntyre et al. 2004: 61) but
does not claim to represent the propositional content or the original wording of
that content. The final category is Representation of Voice, as in (21), which
captures minimal references to speech with no indication of the illocutionary
force, let alone the propositional content or form of the utterance (part) (McIntyre
et al. 2004: 62).
(16) Direct (D):
Yeah she said erm you have it my love dont you

worry but I said no Ill give you some no no I dont
want no money for it at all she said, you take it my
love.
(KB6 1355)
(17) Free Direct (FD):
[Speaker is reporting how someone asked him/her for

change for a fiver].
I said no! [] only. So well can you lend me a
pound? I said no!
(KD5 7945)
(18) Indirect (I):
And I thought she might have come today cos she said
shed bring up the money but she hasnt, not yet.
(KB6 474)
(19) Free Indirect (FI):
Father said can my girls come? No they couldnt come

(McIntyre et al. 2004: 60)
(20) Representation of
Speech Act (RSA):
The only, the only er thing I complained about with

you is, is the butter.
(KBC 6240)
(21) Representation
of Voice (RV):
the children didnt even know he was talking to them.

(KB0 375)
Which of these modes is the most frequent in everyday speech? The answer
McIntyre et al.s analysis suggests is straightforward:
644
Christoph Rhlemann
Table 5.
Proportions of speech presentation categories (McIntyre et al. 2004)
Category
Frequency
Rank
D
FD
I
FI
RSA
RV
38 %
4%
12 %
2%
27 %
17 %
1
5
4
6
2
3
As is shown in Table 5, Direct is by far the most common speech presentation

mode in McIntyre et al.s spoken corpus, accounting for more than a third of all
presentations. The second most frequent mode is RSA, accounting for more than a
quarter, followed by RV, accounting for less than a fifth. Free Direct is relatively
underrepresented in the corpus (ranking fifth), a fact which may be due to the nonconversational interview sections in the corpus. In spontaneous conversation, Free
Direct may be more common (cf. Stenstrm et al. 2002: 110 ff.).
In order to make sense of the dominance of the direct mode in speech it is important to consider this mode in terms of reference and deixis. A fundamental difference between the direct modes and all other modes lies in the speaker perspective (Coulmas 1986: 2). In Direct and Free Direct, the perspective is that of the
presentee: all deictic features are appropriate to the speaker in the anterior situation. By using words which are ostensibly marked as the words of a non-present
speaker, the presenter assumes the role of the presentee: he/she enacts the presentee and his/her speech. By contrast, in the non-direct categories (Free) Indirect,
RSA, and RV, the perspective is that of the presenting speaker: all deictic features
are appropriate to the speaker in the posterior, discourse presenting, situation (cf.
McIntyre 2004: 60).
The switch into the presentees perspective and deictic system that characterizes (Free) Direct has important implications on the referential plane. Each time
the perspective and hence the deictic system changes, so do the referents of the
same deictics. Prime examples of such referential changes are the pronouns I and
you. The two pronouns are core person deictics referring to people immediately
present in the speech situation. (As such, they are distinguished from third-person
pronouns such as he or she which typically refer anaphorically or cataphorically to
entities in the text.) Reconsider example (16), reprinted here as (22). The speaker is
presenting three utterances: two by a speaker who is not present in the posterior,
presenting, situation (utterances 1 and 3) and one she made herself in that situation
(utterance 2). For illustration purposes let us call the presenter speaker A and the
presentee speaker B. The letters A and B in the excerpt indicate co-referential
deictics (presented speech is in bold and deictics are underlined):
(22) 1
2
3
645
Yeah she said erm you(A) have it my love dont you(A) worry
but I(A) said no I(A)ll give you(B) some
no no I(B) dont want no money for it at all she said, you(A) take it my
love.
In the excerpt, a number of switches in deictic reference can be observed. In utterance 1, speaker A uses you twice, the person deictic appropriate to speaker Bs
deictic system to refer to an addressee, in that case, speaker A. Then, in utterance
2, speaker A uses I twice to refer to herself (speaker A), and you to refer to
speaker B thus deploying the deictics appropriate to her own deictic system. Finally, in utterance 3, speaker A switches back into speaker Bs deictic system using
I, whose referent is speaker B, and you, whose referent she is herself (speaker A).
That is, because in Direct presentation, speakers assume different perspectives
and the corresponding deictic systems, reference in Direct mode jumps back and
forth between referents: deictic-system oscillation leads to reference oscillation.
Further, we need to be aware that such discourse presentational changes in perspective are not isolated events occurring here and there but pervasive not only
in longer conversations but often within one and the same utterance, particularly
in narrative.
The study by McIntyre et al. is thus significant on two counts. First, it suggests
one answer (out of several) to the question why I and you are among the most frequent words in conversation, as seen in the previous section.6 I and you are so common in conversation because the most frequent discourse presentation mode is Direct: when presenting discourse, conversationalists mostly use the mode in which
deictic references such as I and you used by speakers in anterior situations are
copied and pasted from that anterior situation into a posterior presenting situation.
Second, McIntyre et al. underscore a view of deixis as a far more flexible system than is often thought. The study provides empirical evidence to suggest that
presenters are by no means confined to presenting discourse and the references
therein from their point of view, with all deictic references conforming to their
origo and the corresponding deictic system. Presenters are free, and make ample
use of that freedom, to shuttle between various origos, deictic systems and perspectives thus creating a continuous oscillation of reference. To judge by the commonness of direct speech presentations, presenters perform this referential oscillation
with great ease. What about the listeners? How can they resolve the ambiguities in
reference that the constant oscillation is likely to bring with it? Clearly, different
voices may be marked off by intonational means, speaker change may be signalled by use of reporting clauses such as he said, I says, etc., utterance openers
such as oh and well may flag the start of a direct speech presentation. But in many
cases no such additional processing instructions are being used. Still, listeners
seem to process switches in perspective and deictic system equally easily as pres-
646
Christoph Rhlemann
enters do. Precisely what enables listeners to resolve reference, which presenters
handle with such flexibility, is still underresearched. Therefore, investigating reference resolution in discourse presentation might add valuable insights to pragmatic theories of inference.
5.
Speech acts
As noted in the introductory section, the relation between quantitative corpus

analysis and speech act analysis is not a one-to-one match because corpora record
surface forms while speech acts are defined on the basis of their function, not their
form (Jucker 2009: 7). However, speech acts are often realized using illocutionary force indicating devices (Searle 1969: 30) or largely routinized forms. Such
features have proven useful anchors for corpus searches for speech acts. The
number of corpus studies into speech acts is to date still small. Two such studies
will be presented in this section, one on the speech act of complimenting and one
on the indirect speech act of suggesting.7
5.1.
Compliments
Compliments have received a fair amount of attention in pragmatic research

(for an overview see Jucker 2009). In a pioneering study, Manes and Wolfson
(1981), using a notebook method (cf. Jucker 2009), claim specific frequencies
for the compliment patterns they found in their data. The study that will be
briefly reviewed here, Jucker et al. (2008), aims to assess the accuracy of Manes
and Wolfsons findings with the help of the BNC. The study is also intended to
highlight methodological problems involved in approaching speech acts via corpora.
The two studies report strikingly different frequencies of compliments/patterns. Using search strings which correspond to the compliment patterns established by Manes and Wolfson, hand-searching subsets of matches and extrapolating their frequencies to the entire sets, Jucker et al. conclude that there are
approximately 343 compliments in the 100-million-word BNC; Manes and Wolfson, by contrast, collected 686 compliment sequences. Further, Manes and Wolfson established nine compliment patterns; the frequencies Jucker et al. found for
these patterns diverge considerably from the frequencies reported by Manes and
Wolfson. The frequencies are shown in Table 6:

Table 6.
647
Frequencies in Manes and Wolfson (1981) and Jucker et al. (2008)
Pattern
Manes & Wolfson
Jucker et al.
1
2
3
4 and 6
5
7
8
9
53.6 %
16.1 %
14.9 %
5.7 %
2.7 %
1.6 %
1.6 %
76.4 %
3.2 %
2.3 %
5.0 %
6.4 %
3.5 %
3.2 %
1.0 %
The differences are most marked with regard to patterns 13. While pattern 1 accounts for slightly more than half of all compliments in Manes and Wolfsons data,
this pattern accounts for more than two thirds in the BNC. Even more striking, in
relative terms, are the differences for patterns 2 and 3: the frequencies reported in
Manes and Wolfson are more than five times as high as in Jucker et al.8 The three
patterns are illustrated in (23)-(25) (optional elements in brackets):
(23) NP+copula+(intensifier)+adjective:
You look so beautiful, he whispered.
(FSF 1954)
(24) I+(intensifier)+verb of liking+NP
You know I really like you!
(KE1 1859)
(25) Pronoun+ copula+(intensifier)+(indefinite article)+adjective+(NP)
Another two pounds a week for that? He looked anxiously at her. Oh,
thats very generous, thank you. That would be marvelous.
(CCM 1399)
The significance of the Jucker et al. corpus study lies not in an assumed superiority
of the corpus method over the notebook (or any other non-corpus linguistic)
method and thus in the claim that the Jucker et al. frequencies were more reliable
than Manes and Wolfsons. By contrast, Jucker et al. demonstrate and discuss in
great detail the merits and, more importantly, shortcomings of the corpus method
as an alternative approach to speech acts. As the authors point out, the surface patterns used are crude (Jucker et al. 2008: 290). This is because the hits either include too many irrelevant matches (a problem of precision) or they fail to retrieve
all relevant examples (a problem of recall). Indeed, precision and recall were
very low, ranging between less than 1 and 20 percent. For example, poor recall may
be the reason why a much lower number of compliments was found in the BNC: al-
648
Christoph Rhlemann
though highly differentiated search strings were used, not all compliments could be
retrieved because compliments, particularly in spoken sections, may include some
minor deviation from the search string, such as pauses or repairs.
Indeed, as the authors point out, [a]lmost every query method fails to have
complete precision and recall (Jucker et al. 2008: 276). To solve this problem,
large pragmatically-annotated corpora would be needed. These, however, are still
in very short supply. Given that the notebook (and any other non-corpus-based)
method have their drawbacks too (see Jucker 2009 for a balanced discussion), the
corpus approach, even in its present imperfect form, is nonetheless a welcome addition to the variety of methods used to study speech acts.
5.2.
Suggestions
Indirect speech acts are cases in which one illocutionary act is performed indirectly by way of performing another (Searle 1975: 60), as in Can you pass the
salt? A much debated question in pragmatics is how listeners resolve the ambiguity between, in this case, question and request. It is often assumed that the adequate interpretation of indirect speech acts requires a complex chain of inferences
by which a recipient first decodes the literal force (in this case, that a question has
been posed to him/her about his/her ability to pass the salt), then realizes that
something is up with this question (for example, that it is not in accordance with
the Cooperative Principle because it fails to be relevant) and only then infers that
the question is not to be taken literally but rather indirectly, as a request by the
speaker to be passed the salt. Corpus studies, by contrast, emphasize the role less of
inferences but more of collocational patterning. This section will highlight one
such study, namely Adolphss (2008), which is based on the Cambridge and Nottingham Corpus of Discourse in English (CANCODE). The focus in this study is
on speech act expressions introducing suggestions; the expressions investigated
include, among others, why dont you.
Why dont you can be used both in direct and indirect speech acts: when used to
introduce a genuine question, a direct speech act is performed; when used to make
a suggestion, a question form is used to put forward a proposal for consideration by
the addressee (in Searles taxonomy of speech acts, suggestions would count as
directives):
(26) PS52K>: I dont believe that.
PS52C>: Why dont you believe it? Its a survey
(KP6 1737)
(27) PS1C1>: but Ive got nobody to go with!
PS1JA>: Oh! Why dont you come with us?
(KDW 2752)
649
The majority of uses of why dont you, Adolphs observes, perform a suggestion.
She elaborates a functional profile of why dont you, consisting of its collocations,
the discourse factors bearing upon its use and its contextual distribution. Thus, she
observes that suggestion-why dont you, henceforth S-WDY, is often preceded by a
form of SAY introducing speech presentation, as in (28). No such association with
speech presentation is reported for question-why dont you, henceforth Q-WDY.
(28) oh I think its because I was saying why dont you come up like this week
(KDM: 7460)
As to right-hand collocates, at N+1 (that is, in the slot immediately after why
dont you) S-WDY is regularly followed by a group of transitive verbs including
ask, get, tell, and use. The set of verbs that Q-WDY collocates with is distinctly
different: they include like, want, and have to. Another key collocate is also found
at N+1: the marker just which, in association with S-WDY, serves to down-tone the
imposition implied in suggestions (remember that suggestions are a type of directives). Consider:
(29) Why dont you just sit down somewhere?
(KBH 4400)
Further, S-WDY and Q-WDY are distinguished by the type of response they typically trigger: while Q-WDY require a more detailed answer (Adolphs 2008: 62),
the responses to S-WDY range from minimal acknowledgment tokens to agreement, or evaluations of the suggestion (ibid.). Finally, Adolphs found that
S-WDY was most frequent in the intimate speaker relationship category in the
CANCODE, that is, it occurs most frequently in interactions between people
whose social closeness is maximal (partners, family, very close friends). In this
context type, speakers are most off-guard and the imposition implied in S-WDY
as a directive is felt as less threatening.
Thus, we see that the uses of Q-WDY and S-WDY are, indeed, idiomatic in
the sense that a large number of distinctly different co-textual and contextual factors bear upon their use and it appears plausible to assume that these idiom factors are salient enough for communicative partners to disambiguate the two speech
acts.
This is not to say that all indirect speech acts are idioms and do not require inferential processes for their interpretation. What Adolphss case study shows is
merely that some indirect speech acts may be less indirect than rather idiomatic,
and it is as yet by no means clear whether the share of such idiomatic speech acts in
all possible types of (indirect) speech act is large or small. No doubt, it is still early
days for corpus linguistic research into speech acts. However, the beginnings are
promising and the prospects are that corpora can make important contributions to
speech act theory.
650
6.
Christoph Rhlemann
Looking to the future
Current corpora facilitate fascinating observations of how words are actually used.
However, they fail to represent communication beyond the word. A particular
challenge for current research is therefore to integrate corpus linguistic methods
and theories of multimodal linguistic research (Carter and Adolphs 2008: 276). At
present, corpora targeted on aspects of multimodal communication are both small
in size and number. Their central purpose is to facilitate explorations of how meaning is made through multimodal patterns, that is, patterns of interaction between
verbal and nonverbal choices (Carter and Adolphs 2008: 281).
Research faces major challenges before this goal is reached. Given that nonverbal meaning seems to make up a very large chunk of overall meaning
(Birdwhistell 1970: 1578 estimates that probably no more than 30 to 35 percent
of the social meaning of a conversation or an interaction is carried by the words) it
will not be surprising that the nonverbal semiotic systems may be as highly differentiated as the verbal system. For example, Rim and Schiaratura (1991: 248) present a taxonomy of speech-related hand gestures which includes six broad variables: speech markers, ideographs, iconic gestures, pantomimic gestures, deictic
gestures, and symbolic gestures. Ekman and OSullivan (1991: 176) discuss evidence for the cross-cultural recognition of at least six emotions expressed via facial
actions: happiness, anger, fear, sadness, surprise, and disgust. Variables of prosody
include rhythm, volume, tempo, voice quality, and intonation with its manifold
subvariables. Obviously, the task of developing a coding scheme to capture such a
wealth of individual variables corpus-linguistically is daunting. It becomes even
more daunting considering that the variables interact rather than act independently
of one another. Therefore, an important goal is the development of tools that provide an integrated approach to the representation of the data (Carter and Adolphs
2008: 283; emphasis in original). Another complicating factor is that verbal and
nonverbal choices are hard to align since within any sequence a substantial
number of utterances and gestures made by speaker and hearer overlap (Carter
and Adolphs 2008: 284). That is, unlike speaking turns which are taken orderly in
the sense that normally one speaker speaks at a time, nonverbal turns are much
less restricted: while a speaker is speaking (and acting nonverbally) the listener(s)
may produce, in response to the speakers unfolding utterance, nonverbal signals
and actions themselves.
Given these challenges, it is small wonder that current multimodal corpus analyses are decidedly selective, focussing on narrow multimodal phenomena rather
than trying to study multimodal patterning in its (at present overwhelming) complexity. One such work in progress is Baldry and Thibault (2006), who take a systemic-functional approach to analyze gaze in a corpus of TV car advertisements.
Another work in progress is Carter and Adolphss (2008) Headtalk project. This
project, which is based on a small corpus of several hours of video-taped MA and
651
PhD supervision sessions at Nottingham University, is intended to explore the patterning of multimodal backchannels, focussing specifically on head nods as a type
of nonverbal backchannel in co-occurrence with verbal backchannels.
To conclude, multimodal corpus linguistics is very much in its infancy
(Baldry and Thibault 2006: 181). However, since this strand of research offers intriguing prospects for an enhanced description of how speakers mean more than
they say, the construction, annotation, and exploitation of multimodal corpora may
in future become a major site of corpus pragmatic research.
Notes
1. Corpus linguistic explorations into evaluative language are numerous. They include,
among many others; Aijmer (1989) on tails, that is, postponed items succeeding the core
of the clause, as in Pathetic behaviour that is, innit?, which primarily seem to fulfill an affective-stance function; Biber and Finegan (1988) on adverbial stance types across text
clusters; and Norrick (2009) on interjections in narrative.
2. An initial attempt at quantifying good and bad prosodies without relying on the
researchers subjective, evaluative judgments is Dilts and Newman (2006), who use a
method based on experimentally measured judgments of goodness and badness obtained prior to, and independently of, corpus-based studies (2006: 240; emphases in
original).
3. A few selected corpus papers on discourse markers include: Aijmer (1987) on the mental
processes signaled by oh and ah; Stenstrm (1998) on cos as a continuation (rather than a
cause) marker; Lenk (1998) on the coherence inducing function of discourse markers. For
a recent collection of papers on discourse markers see Jucker and Ziv (1998).
4. Definite NPs can be marked not only by the (by far the most frequent definite determiner)
but also by possessive determiners (e.g., his, her), and demonstrative determiners (e.g.,
this, that) (Biber et al. 1999:26910). In Biber et al.s study, definite NPs were identified
using an interactive programme which looked for NPs co-occurring with the definite article only (Biber, personal communication).
5. For deixis and indexicality cf. Hanks this volume.
6. The reasons why I and you are so highly common in spontaneous speech are undoubtedly
manifold: they include planning-induced repetition, as in But I cert I I I I I ju I it it just
sounds [] (KB7: 3681), use of I and you in high-frequent discourse markers such as I
know, I see, you know etc. and the fact that, in conversation, speaker and addressee are in
immediate contact, and the interaction typically focuses on matters of immediate concern (Biber et al. 1999: 333) (for a more detailed discussion see Rhlemann 2007a).
7. Further corpus studies into speech acts include Aijmer (1996) on routinized speech act
expressions based on the London-Lund Corpus (LLC) and Kohnen (2000), a pilot study
into explicit performatives based on the Lancaster Oslo/Bergen (LOB) Corpus, the London-Lund Corpus (LLC), and the (historical) Helsinki Corpus (HC); for speech acts in
general see Collavin this volume.
8. Note that the differences in patterns 57 are more important, in relative terms, than the
differences in 1, 4 and 6; however, since the percentage values for patterns 57 are low,
the differences cannot be assigned much significance.
652
Christoph Rhlemann
References
Adolphs, Svenja
2008
Corpus and context. Investigating pragmatic functions in spoken discourse.
Aijmer, Karin
1987
Oh and Ah in English conversation. In: Meijs, Willem (ed.), Corpus Linguistics
and Beyond. 6168. Amsterdam/Atlanta, GA: Rodopi.
Aijmer, Karin
1989
Aijmer, Karin
1996
Conversational Routines in English. London: Longman
Andersen, Gisle
2001
Pragmatic Markers and Sociolinguistic Variation. A Relevance-Theoretic Approach to the Language of Adolescents. Amsterdam: Benjamins.
2006
System and Corpus: Exploring Connection, 16483. London/Oakville: Equinox.
Bednarek, Monika
2008
and Linguistic Theory 4(2): 119139.
1988
Adverbial stance types in English. Discourse Processes 11: 134.
Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad and Edward Finegan
1999
Longman Grammar of Spoken and Written English. Harlow: Pearson.
Birdwhistell, Ray L.
1970
Kinesics and Context: Essays on Body Motion Communication. Philadelphia:
Bublitz, Wolfram
1996
Semantic prosody and cohesive company: somewhat predictable. Leuvense
Bijdragen 85: 132.
Bublitz, Wolfram
2003
Buchstaller, Isabelle
2008
The localization of global linguistic variants. English World-Wide 29(1): 1544.
Carter, Ronald and Svenja Adolphs
2008
Linking the verbal and the visual: new directions for corpus linguistic. Language and Computers 64: 275291
Carter, Ronald and Michael J. McCarthy
1999
The English get-passive in spoken discourse: description and implications for
an interpersonal grammar. English Language and Linguistics 3(1): 4158.
Channell, Joanna
2000
Corpus-based analysis of evaluative lexis. In: Susan Hunston and Geoff
Thompson (eds.), Evaluation in Text: Authorial Stance and the Construction of
Discourse, 3855. Oxford: Oxford University Press.
653
Cook, Guy
1990
Transcribing infinity: Problems of context presentation. Journal of Pragmatics
14: 124.
Coulmas, Florian (ed.)
1986
Direct and Indirect Speech. Berlin: Mouton de Gruyter.
Davies, Mark
2009
The 385+ million word Corpus of Contemporary American English (1990
2008+). Design, architecture, and linguistic insights. International Journal of
Corpus Linguistics 14(2): 159190.
Dilts, Philip and John Newman
2006
A note on quantifying good and bad prosodies. Corpus Linguistics and
Linguistic Theory 2(2): 233242.
Ekman, Paul and Maureen OSullivan
1991
Facial expression: Methods, means, and moves. In: Robert S. Feldman and
Bernard Rim, Fundamentals of Nonverbal Behaviour, 163199. Cambridge:
Hoey, Michael
2003
Lexical priming and the properties of text. Available at: http://www.monabaker.
com/tsresources/LexicalPrimingandthePropertiesofText.htm (last visited October 2009)
Hoey, Michael
2005
Lexical priming. A new theory of words and language. London: Routledge.
Hoffmann, Sebastian, Stefan Evert, Nicholas Smith, David Lee and Ylva Berglund Prytz
2008
Corpus Linguistics with the BNCweb a practical guide. Frankfurt am Main:
Peter Lang
Hunston, Susan
2007
Semantic prosody revisited. International Journal of Corpus Linguistics 12
(2): 249268.
Jucker, Andreas H.
2009
Speech act research between armchair, field and laboratory. The case of compliments. Journal of Pragmatics doi:10.1016/j.pragma.2009.02.004
Jucker, Andreas H. and Yael Ziv (eds.)
1998)
Discourse Markers. Descriptions and Theory. Amsterdam: Benjamins.
Jucker, Andreas H., Gerold Schneider, Irma Taavitsainen and Barb Breustedt
2008
Fishing for compliments. Precision and recall in corpus-linguistic compliment
research. In: Andreas H. Jucker and Irma Taavitsainen (eds.), Speech Acts in
the History of English, 316341. Amsterdam: Benjamins.
Kilgarriff, Adam
1998)
BNC database and word frequency lists. Available at:
http://www.kilgarriff.co.uk/bnc-readme.html (last visited October 2009).
Kohnen, Thomas
2000
Corpora and speech acts: The study of performatives. In: Christian Mair and
Marianne Hundt (eds.), Corpus Linguistics and Linguistic Theory. Papers from
the Twentieth International Conference on English Language Research on
Computerized Corpora (ICAME 20). Freiburg im Breisgau 1999, 17786.
Amsterdam: Rodopi.
654
Christoph Rhlemann
Lampert, Andrew, Robert Dale and Ccile Paris

2006
Classifying Speech acts using Verbal Response Modes. Proceedings of 2006
Australian Language Technology Workshop (ALTW 2006), 3441.
Leech, Geoffrey and Michael H. Short
1981
Style in Fiction. London: Longman
Lenk, Uta
1998
Marking discourse coherence. Functions of discourse markers in spoken English. Tbingen: Narr.
1983
Louw, Bill
1993
Text and Technology, 15792. Amsterdam: Benjamins.
Louw, Bill
2000
Contextual prosodic theory: Bringing semantic prosodies to life. In: Chris
Heffer and Helen Saunston (eds.), Words in context. Discourse Analysis Monograph 18 [CD Rom]. Birmingham: University of Birmingham.
Lycan, William G.
2008
Philosophy of language. A contemporary introduction. New York: Routledge.
Lyons, John
1977
Semantics. Volumes I and II. Cambridge: Cambridge University Press.
Manes, Joan and Nessa Wolfson
1981
The compliment formula. In: Florian Coulmas (ed.), Conversational Routine.
Explorations in Standardized Communication Situations and Prepatterned
Speech, 115132. The Hague: Mouton.
Maynard, Carson and Sheryl Leicher
2007
Pragmatic annotation of an academic spoken corpus for pedagogical purposes.
In: Eileen Fitzpatrick (ed.), Corpus Linguistics beyond the Word. Corpus Research from Phrase to Discourse, 107115. Amsterdam: Rodopi.
McCarthy, Michael J.
1998
Spoken Language and Applied Linguistics. Cambridge: Cambridge University
Press.
McIntyre, Dan, Carol Bellard-Thomson, John Heywood, Tony McEnery, Elena Semino and
Mick Short
2004
Investigating the presentation of speech, writing and thought in spoken British
English: A corpus-based approach. ICAME Journal 28: 4976
Mey, Jacob L.
1991
Pragmatic gardens and their magic. Poetics 20: 233245
Morley, John and Alan Partington
2009
A few Frequently Asked Questions about semantic or evaluative prosody.
International Journal of Corpus Linguistics 14(2): 139158.
Norrick, Neal
2009
Using large corpora of conversation to investigate narrative. The case of interjections in conversational storytelling performance. International Journal of
655
OHalloran, Kieran
2007
Critical discourse analysis and the corpus-informed interpretation of metaphor
at register level. Applied Linguistics 28(1): 124.
Partington, Alan
2004
Utterly content in each others company. Semantic prosody and semantic
preference. International Journal of Corpus Linguistics 9(1): 131156.
Partington, Alan
forthc.
Semantic prosody, evaluation, priming and phrasal irony. In: Karin Aijmer and
Christoph Rhlemann, Corpus Pragmatics. Exploring Speaker Meaning Computerized Corpora.
Rim, Bernard and Loris Schiaratura
1991
Gesture and speech. In: Robert S. Feldman and Bernard Rim. Fundamentals
of Nonverbal Behaviour, 239281. Cambridge: Cambridge University Press.
Rhlemann, Christoph
2006
Coming to terms with conversational grammar: Dislocation and dysfluency. International Journal of Corpus Linguistics 11(4): 385409.
Rhlemann, Christoph
2007a
Conversation in Context. A Corpus-driven Approach. London: Continuum.
Rhlemann, Christoph
2007b
Lexical grammar: the GET-passive as a case in point. ICAME Journal 31:
111127.
Russell, Bertrand
1905
On denoting. Mind 14: 479493. Reprinted in Mind (2005) 114: 873887.
Russell, Bertrand
1919
Introduction to Mathematical Philosophy. London: Routledge.
Searle, John R.
1969
Speech acts. An essay in the philosophy of language. New York: Cambridge
University Press.
Searle, John R.
1975
Indirect Speech Acts. In: Peter Cole and Jerry L. Morgan (eds.), Syntax and
Semantics III, 5982. New York: Academic Press.
Sinclair, John McH.
1991
Corpus, concordance, collocation. Oxford: Oxford University Press.
Sinclair, John McH.
1999
A way with common words. In: Hilde Hasselgard and Signe Oksefjell (eds.),
Out of corpora: Studies in honour of Stig Johansson, 157179. Amsterdam:
Rodopi.
Sinclair, John McH
2004
Trust the text. London: Routledge.
1995
Relevance. communication and cognition. 2nd ed. Oxford: Blackwell.
1998
From Sentence to Discourse: Cos (because) in Teenage Talk. In: Andreas H.
Jucker and Yael Ziv (eds.), Discourse Markers. Descriptions and Theory.
Stenstrm, Anna-Brita, Gisle Andersen and Ingrid K. Hasund
2002
Trends in teenage talk. Amsterdam: Benjamins.
656
Christoph Rhlemann
Stiles, William B.
1992
Describing talk. A taxonomy of verbal response modes. Newbury Park/CA:
Sage Publications.
Strawson, Peter F.
1950
Stubbs, Michael
1996
Text and corpus analysis. Oxford: Blackwell.
Stubbs, Michael
2002
Words and phrases. Corpus studies of lexical semantics. Oxford: Blackwell.
1999
youth. Journal of Sociolinguistics 3(2): 147172.
Whitsitt, Sam
2005
A critique of the concept of semantic prosody. International Journal of Corpus
Linguistics 10(3): 283305.
23.
The transcription of face-to-face interaction1

Roger J. Kreuz and Monica A. Riordan
1.
Introduction
The transcription of face-to-face interaction presents formidable challenges for researchers who study pragmatics. Those venturing into this domain are confronted
by a wide variety of transcription systems that have been devised by researchers
from a diversity of disciplines over several decades. Although these systems are
not mutually exclusive, they often possess large differences in scope, emphasis,
and nomenclature, as well as in the symbols used to transcribe these dimensions.
It is beyond the scope of this chapter to reconcile these systems, and it is far
from clear that such a reconciliation would be desirable. Instead, we will provide
an overview of a number of these systems, and make suggestions about their suitability for various transcription needs. In addition, we will address transcription
issues with regard to speech in diverse populations (e.g., children, aphasics, and
cognitively impaired individuals). Finally, we will consider the challenges of transcribing the non-acoustic (i.e., facial and gestural) signals that are of special interest to pragmatics researchers. It is our hope that this overview and discussion will
provide some guidance to those who wish to practice the art of transcription.
2.
Some preliminaries
2.1.
Purpose of transcription
Researchers who study pragmatics create transcriptions in order to test their theories about discourse. In other words, no one creates transcriptions of face-to-face
interactions as an end in themselves; rather, transcription is always performed as a
means to an end. For example, President Richard Nixons Oval Office tapes (New
York Times 1974) were transcribed for the purpose of determining what the president knew about the Watergate cover-up. (The frequent use of the term expletive
deleted by the transcriber provided the public with some unintended pragmatic insight into the speech of the head of the Executive Branch.) Other researchers have
gone on to use Nixons recordings for their own research purposes (e.g., Novick,
Walton and Ward 1996), but the original transcriptions were not created with any
other purpose in mind.
The fidelity of transcription is a matter of some importance in the fields of
medical and legal transcription. However, in these applied settings, the focus is pri-
658
marily on what was said or intended as opposed to how it was uttered. Although
there may be important commonalities between transcription in applied and research domains, they will not be reviewed here.
2.2.
Transcription versus coding
Many issues in pragmatics may be profitably explored without recourse to transcription. Because transcription is extremely time intensive (see section 2.5), it
may in fact amount to overkill for many purposes. For instance, if one simply
wanted to compare the number of specific discourse markers (e.g., Schiffrin 1987)
in two stretches of discourse, it might in fact be easier to tally them directly from an
audio- or videotape of the interaction. This approach might also be more accurate,
because it would avoid many of the complications that are inevitable when an interaction is reduced to a transcription. Such coding should be possible for many
phenomena that are conceptually well-defined and macroscopic.
In other cases, a researcher may be interested in classifying utterances according to a particular taxonomy, such as illocutionary speech acts (e.g., Searle 1975).
In such cases, it would probably be easiest to work from a relatively broad transcription of the interaction. The focus of this review will be the broad range between counts and coding schemes, which require the use of a particular transcription system that produces a faithful record of an interaction.
2.3.
Issues of terminology
Phonetic transcriptions are frequently characterized as broad or narrow, in

which narrow transcription documents the allophonic variation in speech sounds.
International standards exist for transcribing phonetic information at both the
broad and narrow levels (Pullum and Ladusaw 1996), although not surprisingly,
reliability is higher for broad transcriptions than narrow ones (Shriberg and Lof
1991). In a similar way, the coding of non-phonetic dimensions, such as pauses or
gestures, can be characterized as broad or narrow. A paradox of transcription is
that, as a transcript becomes narrower (and theoretically, more faithful to the discourse it represents) it becomes more difficult to read and interpret. In other
words, there may be a tradeoff between the fidelity of a transcript and its intelligibility.
Consider, for example, a few turns from the relatively narrow transcription provided by Kyratzis (2001) in her analysis of the interactions of preschool friends:
1. Speaker1: if someone comes, then we **hi::de, really//
2. Speaker 2: ==yeah/
3.
were *shy:: wizards// (Kyratzis 2001: 363)
659
The symbols provide important information about intonation, lengthened segments, and latching, although most readers will need to refer to the authors list of
transcription conventions to fully decode these features.
It may also be helpful at this point to define the specific dimensions that are of
interest to transcribers of discourse. OConnell and Kowal (1995a, 1999, Kowal
and OConnell 2004) suggest that transcribed behaviors can be categorized as belonging to one of four classes of features. Verbal features refer to the words themselves (i.e., what was said), whereas prosodic features correspond to the ways in
which the words were spoken (e.g., pitch, duration, and loudness). Paralinguistic
features (such as laughter, breathing, sighing, or crying) may accompany the
spoken words, or they may occur independently. Finally, extralinguistic features
are behaviors defined as nonvocal and nonverbal (OConnell and Kowal 1999:
109), but which are nevertheless germane to the discourse being described (e.g., facial expression, gaze, and gesture). The purpose of a transcription system is to subdivide these features into a number of dimensions, which can then be labeled with a
particular code or set of codes. Transcription systems differ in their coverage of
these four features, the number of dimensions employed within each feature, and in
the specific codes that are used.
2.4.
Transcription layout
Edwards (1993b) has reviewed how transcription systems differ in the spatial arrangement of information as well as the type and level of description. Vertical arrangement, in which speakers turns are arranged sequentially (as in a script) may
be the most common, although other formats may be more informative. Representing the interaction between interlocutors in columns may be helpful in some cases,
and a partiture format, as in Ehlichs (1993) HIAT system (see section 4.6), is effective for capturing interactions with lots of simultaneous speech.
There are several different ways in which a transcriber can choose to arrange
prosodic, paralinguistic, and extralinguistic features within a transcript. One
choice, referred to by Edwards (1993b) as running text, places such information
following the words (e.g., the code laughter immediately after an utterance containing laughter). This method preserves the temporal contiguity of the information in the transcript. However, many transcription systems use an interspersed format for recording prosodic information (Edwards 1993b). Changes in
pitch, for example, can be directly mapped onto the syllables themselves, or indicated by specific codes. A third approach is to use a segment-plus-specification
(SPS) format, in which one tier provides the verbal dimension, and other tiers or
rows below the first tier provide syntactic, semantic, or pragmatic codes (Edwards
1993b). A fourth choice is referred to as utterance-plus-clarification. In this
format, utterances are broken apart and nonverbal or contextual information (e.g.,
gesture, gaze, or behavior of the speaker) appears below each speakers turn. As
660
Edwards (1993b) notes, even something as mundane as the arrangement of speakers turns can have important implications for how a transcript is analyzed and interpreted.
2.5.
Procedural issues
One issue all researchers in pragmatics can agree on is that transcription is very
labor intensive. MacWhinney (2001), for example, has estimated that it can take
over ten hours to transcribe one hour of discourse. Consequently, it makes sense to
transcribe as few dimensions as possible for the purposes at hand. For example, if
one wanted to study the use of discourse markers such as uh and um (Clark and
Fox Tree 2002), it might make little sense to transcribe head movements of the listener. However, such choices are crucial, for as Ochs (1979) and many others have
argued, a transcription system is not theory neutral: it already reflects the beliefs
and biases of the researcher. By choosing not to transcribe a particular dimension,
the researcher has implicitly decided that the dimension plays no role in the phenomenon in question. Considering the infinity of pragmatically salient dimensions
that the researcher could transcribe, the choice of the dimensions themselves is already something of a compromise. To put it another way, transcription is always
subjective and interpretive to some degree, and transcription inevitably leads to
data reduction. These issues are of fundamental importance, but are beyond the
scope of this chapter to do them justice.2
2.6.
The role of context
Unless it is occurring under unnatural laboratory conditions, face-to-face interaction is deeply embedded within a social context, involving dimensions that may
not be readily apparent to the transcriber. The relationship that exists between the
conversational participants may be crucial to the interpretation of the interaction
and should be specified when possible. However, assessing the relevant parameters, such as the amount of shared common ground (Clark 1996), may be quite
difficult. Most transcribers have not paid much attention to specifying context, although some of these issues have been explored by Cook (1995), Norris (2004),
and Bucholtz (2007b). Sensitivity to these issues can make for a more informative
transcript: acoustically, there is no difference in transcribing a brief absence of
speech as a pause or as shocked silence, but clearly the latter provides more information (perhaps at the expense of complete objectivity).
2.7.
Technological considerations
Technological developments have greatly impacted the way in which researchers

record conversational interaction. Most researchers now employ digitally recorded
661
video, which makes it possible to record acoustic, facial, and gestural information
(a helpful review of these issues may be found in Bavelas, Kenwood, and Phillips
2002).
An important technological issue with regard to transcription is the conversion
of transcription systems into mark-up languages that can be interpreted by computer programs. Such conversion is necessary to allow efficient data storage and
retrieval, but as with other aspects of transcription, a number of tradeoffs are involved (see Leech, Myers, and Thomas 1995, for a variety of such examples).
A variety of software packages exist for use in the creation of a transcript. These
programs can lessen some of the tedium involved in repeatedly playing a troublesome section of speech, or temporally aligning verbal and extralinguistic features.
Examples include VoiceWalker, described as a discourse transcription utility (Du
Bois 2006: 1), Anvil, an annotation program (Kipp 2003), and Transana (Woods
2007), an analysis package available at http://www.transana.org. Other widely used
and free software packages include ELAN, http://www.lat-mpi.eu/tools/elan/ and
Praat, http://www.praat.org. Programs that can assist in the coding of gestural components of language also exist (Neidle, Sclaroff and Athitsos 2001).
3.
Design principles
A number of theorists and researchers have made proposals concerning what an

ideal transcription system should encompass. These ideas will be reviewed chronologically in order to show their development over time.
Du Bois (1991) proposed five maxims for designing transcription systems and
23 design principles that follow from these maxims. The maxims are category definition (e.g., use categories that are explicit, necessary, and sufficient); accessibility
(e.g., use familiar and easily learned notations); robustness (e.g., avoid invisible or
fragile contrasts); economy (e.g., avoid verbose notations, use space meaningfully); and adaptability (e.g., allow for seamless integration of user-defined codes
and transcription categories).
Edwards (1993b) argued that the creators of transcripts should keep in mind issues of category design, readability, and computational tractability. By category
design, she means that the dimensions must be systematically discriminable,
exhaustive, and contrastive. Readability refers to placing related events in close
proximity, the visual separation of unlike elements, time-space iconicity, logical
priority, mnemonic marking, and efficiency and compactness. Computational tractability refers to systematicity and predictability in encoding dimensions. Failure
to consider computational tractability can lead to the underselection or overselection of instances from a transcription. For example, a search of a transcript for instances of going to would not necessarily identify instances of gonna, while a
search for the verb bear might also snare the ursine variety. Edwards (1995) later
662
expanded on these concerns, and also discussed issues of validity and reliability
when computerized language archives are used.
OConnell and Kowal (1994, 1995a, 1999, 2008) point out that the ultimate
purpose of a transcription (i.e., the purpose of the researchers and the readership of
a transcript) must be kept in mind. Second, these authors argue against inserting
transcription symbols within words to indicate prosodic features, because doing so
impairs the legibility of the transcript. A third point concerns the consistency of
notation: each symbol should encode only one dimension, and conversely, each dimension should be encoded by only one symbol. In addition, conventional typographic elements such as ampersands and ellipses should not be used as transcription symbols, because their primary purpose is already deeply ingrained in the
minds of readers. Finally, the authors argue that measures of continuous variable
dimensions, such as amplitude and time, must be made with accurate equipment,
and not subjectively.
Dressler and Kreuz (2000) encouraged the developers of transcription systems
to keep in mind seven design principles: (1) specificity (the tradeoffs of broad versus narrow transcription); (2) universality (not linking the conventions to a particular language, such as s for softly); (3) consensus (using symbols as others have
used them in the past); (4) transparency (using intuitive symbols, such as a rising
line, /, to indicate rising intonation); (5) parsimony (the use of a small number of
codes); (6) conventionality (using codes that can be easily located on a keyboard);
and (7) extensibility (the system should be open-ended to allow new dimensions to
be transcribed).
Mller and Damico (2002), approaching these issues from the perspective of
clinical linguistics and phonetics, proposed six guiding principles that share much
in common with the points already raised. However, for dealing with the complexities of disordered speech, they stress the importance of flexibility to ensure authenticity and individuality (Mller and Damico 2002: 312).
4.
Review of transcription systems
Many researchers choose to reinvent the wheel and create their own notational systems when transcribing their data. Although this may be justifiable in some cases,
it is rarely necessary. There are, in fact, many disadvantages to this approach. An
ad hoc system is likely to be less comprehensive and may be employed inconsistently. In addition, if examples are included in published research, the use of a new
system requires some mastery on the part of editors, reviewers, and readers. Finally, a transcription key must be provided, which wastes the resource of journal
space (Dressler and Kreuz 2000).
The list of transcription systems provided here is not intended to be comprehensive; as noted above, many systems have been developed and used only once.
663
In addition, some well-known systems are not particularly comprehensive, whereas others have been employed in relatively small geographic regions or for only
one language. Therefore, the goal has been to describe briefly the systems that are
(a) widely employed, (b) in current use or historically important, and (c) reasonably comprehensive. Most of these systems have been developed to transcribe
prosodic and paralinguistic features; the transcription of extralinguistic features,
such as gesture and eye gaze, will be reviewed in section 5. Finally, each transcription system is listed by name (if it has one), or by reference to researchers and publication(s) that describe the system in detail.
4.1.
Jeffersonian Transcription System
Gail Jeffersons transcription system (Sacks, Schegloff, and Jefferson 1974; Schenkein 1978; Atkinson and Heritage 1984; Jefferson 2002, 2004) has been widely employed and refined over a thirty-year period, and has become a de facto standard in
the field of conversation analysis (often referred to as simply CA). The journal
Research on Language and Social Interaction, for example, uses the Jeffersonian
notation as its default transcription system.
In its 1978 formulation (Schenkein 1978), the Jeffersonian system provided
about 18 codes for tracking seven different categories of conversational phenomena: simultaneous utterances, overlapping utterances, contiguous utterances, intervals within and between utterances, characteristics of speech delivery, transcriptionist doubt, and other transcript symbols. In a later formulation (Atkinson and
Heritage 1984), about eight codes, such as shifts in intonation and quieter talk,
were added, as well as the extralinguistic dimensions of gaze direction and applause. More recent formulations (e.g., Jefferson 2002) have introduced a handful
of other codes to mark slower speech or suppressed laughter. A helpful discussion
of the issues involved in employing this system may be found in Psathas and Anderson (1990).
4.2.
Ochs
Following the lead of the seminal paper of Sacks et al. (1974), Elinor Ochs (1979)
proposed a transcription system for verbal and nonverbal features. For verbal features, she proposed the coding of eight dimensions: utterance boundary, latching,
pause length, overlap, self-interruption, intonation or prosodic quality, audible
breathing, and metatranscription. Four additional dimensions were proposed for
nonverbal features: changes in gross motor activity, eye gaze, gestures, and body
orientation. In total, the system uses about 35 codes (for this and other systems, an
exact number is difficult to report, since some codes can be used iteratively in
Ochs system, for example, lengthened syllables are indicated with a colon, and
additional colons can be used to indicate additional beats in time).
664
4.3.
London-Lund Corpus transcription
The London-Lund Corpus of Spoken English (or LLC) (Svartvik and Quirk 1980,
Svartvik 1990) is of considerable historical importance. The project began in London in 1959 and Lund, Sweden in 1975, and represents one of the first attempts to
gather a comprehensive corpus of spoken English. It was derived from the Survey
of English Usage (or SEU), a corpus which consists of a million words. Half of this
total was drawn from spoken English.
The LLC uses a transcription system that principally denotes prosodic features.
It stems from the British School of intonation analysis (see Kingdon 1958,
OConnor and Arnold 1961). The nucleus, or main stressed syllable that has a
clearly perceptible movement of pitch, is divided into seven dimensions (e.g., fall,
level, fall-rise), and there are four codes for the booster, or range of pitch (e.g.,
higher than preceding syllable, very high). Stress is denoted as normal or heavy.
Pauses of varying durations, as well as simultaneous talk and laughter are coded as
well. Although this system is not as comprehensive as some others, it has been
widely employed, and may be sufficient for researchers whose primary interest is
prosody.
4.4.
Tannen
Deborah Tannens (1984/2005) work is well known in sociolinguistics, and her

system has frequently been employed by later researchers. Her system codes for
pauses, stress, pitch, intonation, vowel lengthening, and overlapping speech. Amplitude is described using six codes drawn from musical notation (e.g., piano, fortissimo), and appears under the transcription line. Brackets are used to demarcate
paralinguistic or extralinguistic information (e.g., [laughter]). The system uses
about 30 codes altogether.
4.5.
Discourse Transcription (Du Bois)
The Discourse Transcription system proposed by John Du Bois (Du Bois 1991; Du
Bois et al. 1992, 1993), and often referred to simply as DT, has been adopted by
many researchers and takes into account Du Bois design principles outlined
above. The system includes codes for pauses of various kinds, overlapping sequences, vocal quality, and utterance boundaries. In addition, it includes several
codes for prosody (primarily accent and pitch), transcribers comments, and even
smile quality. The Discourse Transcription system uses about 40 codes.
4.6.
665
HIAT (Ehlich)
HIAT, or Halbinterpretative Arbeitstranskriptionen (Heuristic Interpretative Auditory Transcription, Ehlich 1993) represents a movement away from standard orthography, which may lead to a loss of important information. A system of literary
transcription is proposed instead. This system makes use of symbols from the International Phonetic Alphabet, and uses vertical space on the page to represent
simultaneous events, much like a musical score. The coding of intonation is represented in a similar way. The system excels at tracking multiparty conversations and
overlapping speech, although such transcriptions require a great deal of space.
Computer programs designed to facilitate transcription into the HIAT system are
available.
4.7.
Gumperz and Berenz
Gumperz and Berenz (1993) approach transcription from a sociolinguistic perspective, emphasizing the situated interpretations of the conversational participants. As they put it, our main goal is to reveal the functioning of communicative
signs in the turn-by-turn interpretation of talk, not to record everything that can be
heard or to provide exact measures of duration and pitch (Gumperz and Berenz
1993: 119). The system uses about 22 codes, and may be attractive to researchers
who desire a broader approach to transcription.
4.8.
GAT (Selting et al.)
The Gesprchsanalytisches Transkriptionssystem (GAT) (Selting, Auer, Barden,

Bergmann, Couper-Kuhlen, Gnthner, Meier, Quasthoff, Schlobinski and Uhmann
1998) was created to help reduce the hitherto often unmotivated variation in transcripts (Selting et al. 1998: 91), and includes 14 dimensions and about 50 codes.
Like Tannens, this system provides many gradations for amplitude, using nomenclature derived from musical notation (e.g., piano, forte, crescendo, and diminuendo).
4.9.
CHILDES and CHAT (MacWhinney)
The Child Language Database Exchange System (CHILDES) was begun in 1981
in an attempt to gather together transcripts of child language (MacWhinney 2000).
The project, established by Brian MacWhinney, has grown and evolved over time
and now includes adult interactions in the TalkBank Project. Developments in
computer technology have greatly enhanced the utility of this resource for language researchers. The transcripts themselves are freely available on the Internet
(http://talkbank.org/). In addition, tools for coding and analyzing these corpora
666
have been devised. Codes for the Human Analysis of Transcripts, or CHAT, is the
transcription system and coding format, and CLAN (for Computerized Language
Analysis) is the software tool developed to create and analyze CHILDES transcripts. The current version of CHAT (MacWhinney 2000) provides researchers
with extensive sets of codes for use in transcription, and even accommodates other
notational schemes, such as the Jeffersonian Transcription System described in
section 4.1. However, the sophistication of this system may also be its principal
weakness because researchers may need to devote a considerable amount of time
and effort to mastering its intricacies.
4.10.
Dressler and Kreuz
Dressler and Kreuz (2000) reviewed 24 papers employing transcription that

appeared in one journal (Discourse Processes) over a five-year period. They
found that 21 codes could accommodate the majority of the researchers dimensions, and grouped these dimensions into five classes: intonation, temporal features, intensity, breathing, and transcribers comments. Although this system employs some higher-level categories, such as backchannel communication and
paralinguistic behavior, no attempt was made to include finer distinctions.
4.11.
Powers
Powers (2005), an anthropologist, produced a transcription handbook to be used by

ethnographers. Not surprisingly, therefore, the focus is somewhat different than for
the other systems described here. Specifically, Powers system codes for a smaller
number of dimensions (about 18), and does not include notations for intonation or
breathing. On the other hand, this system explicitly accommodates a number of dimensions of paralinguistic and extralinguistic features, such as weeping, reported
speech, and irony.
5.
Transcription of extralinguistic features
Communication in face-to-face dialogue is not limited to spoken words. Visible actions such as facial expressions and gestures can serve to reinforce words and decrease ambiguity in interpretation.
Bavelas and Chovil (2000) argue that visible actions are only important when
they are part of a communication; for example, scratching ones knee to emphasize
a conversation about a rash one had two days ago would be a communicative gesture, but the same action during a conversation about the price of milk would not.
The authors call these communicative actions visible acts of meaning (Bavelas
and Chovil 2000: 165), and include among them facial displays such as eyebrow
667
raises, hand gestures such as circular motions to depict a circle, and communicative body movements such as shrugging ones shoulders.
Bavelas and Chovil (2000) outline four characteristics that define a visible act
of meaning: (1) the action must occur in face-to-face dialogue and be reduced
when the receiver of the action cannot see the action, (2) the action must stand as a
symbol for something that is not physically present at the moment, (3) the meaning
of the action must be expressed either in words or by a demonstration that the receiver uses the information, and (4) the action must be integrated with the spoken
dialogue. The research questions at hand dictate how these visible actions are transcribed. Facial expressions can be transcribed either as physical actions or as
meaning-based actions (Bavelas, Kenwood, and Phillips 2002).
Several researchers have used Ekman and Friesens (1978) Facial Action
Coding System (FACS), a transcription system based on physical actions that utilizes 44 action units such as head turn right and lip stretcher, several of which
are coded to varying degrees of intensity.3 Chovil (1989) developed a meaningbased system to contrast with Ekman and Friesens (1978) physical transcription
system. Her system uses descriptions of the facial expression as a whole, such as
sadness and skepticism. Bavelas, Kenwood, and Phillips (2002) argue that this
meaning-based approach may not only be more useful for discourse research, but
also less time-consuming to researchers, and indeed Chovil (1989) demonstrated a
higher interrater reliability than FACS. Some might argue, however, that a more
subjective system reduces validity.
These two extremes musculature analysis and subjective ratings may not be
helpful for a variety of researchers. For those looking for a middle ground, Louwerse et al. (2007) devised an attractive alternative. Louwerse et al. (2007) used a
subset of Ekman, Friesen, and Hagers (2002) Facial Action Coding Scheme standard, coding just 20 facial movements that were of interest for their research questions. Other researchers may wish to employ this system or a different subset from
Ekman et al. (2002), based on their own particular research interests.
5.1.
Coding of gesture
The transcription of gestures presents additional challenges to the discourse researcher, because gestures occur simultaneously with talk and some means of mapping the two in time must be considered (see Goldin-Meadow 2003). The following is a review of a subset of gestural coding systems.
5.1.1.
Ochs
Elinor Ochs (1979) system, described in 4.2, includes five codes for gestures like
pointing, holding up, and offering.
668
5.1.2.
Schegloff
In Emanuel Schegloffs (1984) analysis of deictic gestures, he proposed indexing

hand and limb movements on a line above the transcription of the words being uttered. The system utilizes eight codes, denoting, for example, the onset of movement, maximum extension, and pointing, as well as temporal elements.
5.1.3.
Bull
Peter Bull (1987, 1989) proposed a Body Movement Scoring System, in which
body contact and object contact are described in terms of (1) the body part making
the motion, (2) the type of motion, and (3) the body part or object with which contact is made. One attractive aspect of this system is that its practitioners have been
able to achieve high interrater reliability (Bull and Connelly 1985).
5.1.4.
Ehlich
The HIAT system (Ehlich 1993), described in 4.6 above, includes 25 codes for referring to parts of the head, hands, arms, legs, and body.
5.1.5.
CoGesT (Gut et al.)
CoGesT, or Conversational Gesture Transcription system (Gut, Looks, Thies, Trippel and Gibbon 2002), is an attempt to create a notational system based on distinctions between categories of gestural form and function. The system makes distinctions on a variety of dimensions, such as form, phase, location, and directionality.
Specific examples include hand shapes, repetitions, and speed (Gut et al., 2002).
5.1.6.
McNeill
Susan Duncan has developed a coding manual that has been employed by David
McNeill and his collaborators (McNeill 2005). She suggests making eight passes
through the interaction to be analyzed, and in addition to acoustic and prosodic dimensions described earlier, adds the categories of handedness, hand orientation,
hand position, and phases (i.e., points in the gesture process).
5.2.
Eye gaze
The eye gaze of interlocutors can signal important pragmatic features during an interaction. Speakers, for example, tend to establish eye contact with their partners at
the end of a turn (Levelt 1993). A number of the systems described in section 4
contain at least some codes for eye gaze. An example would be the system pro-
669
posed by Ochs (1979), which provides six codes for looking up, down, left, right,
and towards and away from the camera. These codes can be paired with the person
or object being looked to. Damico and Simmons-Mackie (2002) have proposed a
system in which a layer of gaze and gesture information can be mapped onto a base
layer of broader transcription. This proposal is attractive because it allows extralinguistic features to be represented separately from the prosodic and paralinguistic features of discourse.
5.3.
Body posture and orientation
Speakers and listeners rarely remain static during an interaction, although deciding
which body movements are pragmatically salient may be difficult to determine.
Most of the systems described earlier could accommodate such meaningful movements as part of the transcribers comments. Ochs (1979) suggests using a U-shape
to indicate the direction of a speakers pelvis.
6.
Child language transcription
Interpreting the language of adults is difficult enough, yet child language researchers must deal with all these issues and more. A good overview is provided by
Bloom (1993), who proposes a model system for the computer-aided transcription
of the speech of children. She highlights two issues in particular: the biases and
distortions that may be introduced by the observer, and the massive amount of data
reduction, from the recording of the interaction to the transcription process itself.
The conversion of child language into forms that can be accessed electronically
has also been an issue in the transcription literature. Edwards (1992) proposed four
principles for the use of such archived data, which are similar to the design principles discussed in section 3. However, one of her suggestions, the consistent
coding of the data, has been somewhat contentious. Edwards (1993a) noted that the
use of novel variations such as falld and falled, might cause one or the other to
be overlooked in an electronic search for such instances. She argues that this is an
important issue because many forms used by children are rather rare. Her concerns
were expressed with regard to the early forms of CHAT, described above (a discussion of these issues may be found in MacWhinney and Snow 1992).
7.
Signed language transcription
The transcription of sign languages, such as American Sign Language, presents its
own set of issues and complexities. Hoiting and Slobin (2002), inspired in part by
the CHILDES system (see section 4.10), have proposed a scheme to address these
670
issues. This scheme is referred to as the Berkeley Transcript System (BTS), and
BTS transcripts are available via CHILDES (MacWhinney 2000).
8.
Transcribing cognitively impaired individuals
The difficulties involved in transcribing the interactions of adults and children may
pale in comparison to reproducing the productions of those with cognitive impairments. Ball and Rahilly (2002) make some suggestions for transcribing the prosodic features of disordered speech, and propose a scheme that is similar to the
HIAT system reviewed in section 4.6 (Ehlich 1993). TalkBank (see 4.9) also includes a section called Aphasia Bank for the dissemination of transcriptions of aphasic speech.
Haravon, Obler, and Sarno (1994) present a system for analyzing the discourse
of those with brain injury. They suggest that their approach has utility for studying
the productions of aphasics and those suffering from Alzheimers disease. Their
approach is notable in that it explicitly takes into consideration pragmatic issues
(in addition to morphology and syntax).
Mller and Guendouzi (2002) propose a multilayered approach in their system
for transcribing the discourse of Alzheimers patients. Specifically, they recommend employing a baseline or orthographic layer, a layer addressing prosody and
voice issues, and a discourse layer. The codes they use are similar to those in other
transcription systems described above, but the multiple layers provide more clarity
and allow the reader a better chance of making sense of the disordered speech. This
approach is taken even further by Mller and Damico (2002), who propose six
layers: in addition to the levels already described, they add gaze and gesture, speech
(phonetic transcription), and clinical analysis (analysis of specific behaviors).
9.
Critiques of transcription systems
Clearly, the most important attribute of a transcription system is the capability to

recreate an interaction with a high degree of fidelity. The degree to which this fidelity is achieved will depend on many factors that exist outside of the system
being employed. The experience level and the care taken by the transcriber are crucial, since even small errors can completely change the perceived meaning of an
utterance (Easton, McComish, and Greenberg 2000). In addition, once an interaction has been transcribed according to one system, it may be difficult to transfer it
into a different system (Allwood, Henrichsen, Grnqvist, Ahlsn, and Gunnarsson
2005).
In a series of empirical papers, Daniel OConnell and Sabine Kowal have explored a number of issues related to the validity, generalizability, and objectivity of
671
transcription systems in current use. Their pessimistic conclusion is that transcription itself is a limited and defective device (OConnell and Kowal 2008: 93,
emphasis in the original). Although this gloomy assessment may seem overstated,
it is a conclusion that they have come to as the result of their research, which is
summarized below as a cautionary tale for the enterprising transcriber.
As a starting point, OConnell and Kowal (1999) addressed the issue of standardization in transcription notation. In a review of three widely used transcription
systems, they found that a majority of the dimensions were used to transcribe prosodic features, whereas codes for extralinguistic features made up between zero and
22 % of the total for each system. Their conclusion, however, is that standardization is not practical, or even warranted, given the diversity of behaviors that researchers are interested in.
If standardization of transcription systems is not a realistic goal, then surely at
least the reproducibility of transcripts is achievable. However, OConnell and
Kowal (2000) found that reproductions of transcripts in textbooks had, on average,
an error rate of one change per 6.6 syllables. They attribute this high error rate to
the density and relative unfamiliarity of transcription systems, which overload the
scholars and typesetters who reproduce the examples.
The idea of conceptual overload was further explored in a study by Romero,
OConnell, and Kowal (2002). They asked undergraduate participants to reproduce
a 21-syllable question asked by a news reporter. Participants were assigned to a
variety of conditions in which they were provided with only the audio recording,
with an ordinary transcription (verbal features only), or with a transcript that had
been generated using one of three widely employed transcription systems in which
dimensions of prosodic features were explicitly coded. The participants task was
to reproduce the news reporters prosody as closely as possible. When the participants productions were compared to the original, it was found that only one of the
three transcriptions yielded reproductions were better than for those participants
who heard the original recording. In general, the participants found the prosodic
codes difficult to interpret.
Finally, OConnell and Kowal (1995b), in their review of five of the transcription systems mentioned above, conclude that all of these notational schemes violate, to some degree, the seven design principles proposed in OConnell and Kowal
(1994; see section 3).
It is also worth noting that the type of discourse can present considerable problems for transcribers. Lindsay and OConnell (1995) have shown that the fragmentary nature of spontaneous speech filled with incomplete sentences, hesitations,
and overlapping speech can be particularly troublesome to transcribe because of
its complexity (see also Bucholtz 2007b).
Given the tedium of transcribing long stretches of video or audiotape, it should
come as no surprise that such tasks are frequently assigned to graduate or even
undergraduate students with little background in theories of discourse or training
672
in transcription. Some of the issues surrounding the use of such transcribers, such
as issues of training, have been described by Tilley (2003).
The fidelity of a given transcript to a particular notation system can be assessed
by comparing the work of two (or more) transcribers who have independently
applied the system to the same stretch of discourse. The measurements can range
from simple measures of agreement to more sophisticated approaches, such as
Cohens kappa, which controls for chance performance (Cohen 1960). A tradeoff
exists between the number of dimensions employed by a particular transcription
system and a measure of interrater reliability.4 It is worth noting that some researchers have been critical of the quest to achieve high reliability, because putative errors may in fact provide important information (Pye, Wilcox, and Siren
1988).
10.
Conclusions
The range of issues and choices that confront the would-be discourse transcriber
may seem overwhelming. In reality, however, any research project involves a variety of choices and trade-offs, and viewed from this perspective, the selection of a
transcription system is no different from the choice of a statistical test. In both
cases, the ultimate goal is to illuminate the underlying systematicity that exists
within the data, and there may be a variety of legitimate ways to achieve this end.
Furthermore, even though transcription can be very labor intensive, it is possible to
find the process enjoyable (Bird 2005). It is our hope that the information we have
presented can provide guidance for those who wish to explore these issues in
greater depth.
Notes
1. Partial support for the preparation of this chapter was provided by a Center of Excellence
grant from the state of Tennessee to the Department of Psychology at the University of
Memphis, and by a grant from the Human and Social Dynamics program of the National
Science Foundation (BCS 0826825). Max Louwerse, Gina Caucci, and Wolfram Bublitz
provided helpful comments on earlier drafts of this chapter. Comments and questions
concerning this chapter may be directed to Roger Kreuz (rkreuz@memphis.edu).
2. Please refer to Bucholtz (2000, 2007a, 2007b); Green, Franquiz, and Dixon (1997); Jaffe
(2007); Mishler (1991); Preston (1982); Roberts (1997); and Tilley (2003) for further discussion and analysis.
3. For an evaluation of the system, see Sayette, Cohen, Wertz, Perrott and Parrot (2001).
4. For a more extended discussion, see Roberts and Robinson (2004) and Stelma and Cameron (2007).
673
References
Allwood, Jens, Peter J. Henrichsen, Leif Grnqvist, Elisabeth Ahlsn and Magnus Gunnarsson
2005
Transliteration between spoken language corpora. Nordic Journal of Linguistics 28: 536.
Atkinson, J. Maxwell and John Heritage
1984
Transcription notation. In: J. Maxwell Atkinson and John Heritage (eds.),
Structures of Social Action: Studies in Conversation Analysis, ix-xvi. Cambridge: Cambridge University Press.
Ball, Martin J. and Joan Rahilly
2002
Transcribing disordered speech: The segmental and prosodic layers. Clinical
Linguistics and Phonetics 16: 329344.
Bavelas, Janet B. and Nicole Chovil
2000 Visible acts of meaning: An integrated message model of language in face-to-face
dialogue. Journal of Language and Social Psychology 19: 163194.
Bavelas, Janet B., Christine Kenwood and Bruce Phillips
2002
Discourse analysis. In: Mark L. Knapp and John A. Daly (eds.), Handbook of
Interpersonal Communication. 3rd ed., 102129. Thousand Oaks, CA: Sage
Publications.
Bird, Cindy M.
2005
How I stopped dreading and learned to love transcription. Qualitative Inquiry
11: 226248.
Bloom, Lois
1993
Transcription and coding for child language research: The parts are more than
the whole. In: Jane A. Edwards and Martin D. Lampert (eds.), Talking Data:
Transcription and Coding in Discourse Research, 149166. Hillsdale, NJ:
Lawrence Erlbaum.
Bucholtz, Mary
2000
The politics of transcription. Journal of Pragmatics 32: 14391465.
Bucholtz, Mary
2007a
Variation in transcription. Discourse Studies 9: 784808.
Bucholtz, Mary
2007b
Reply: Variability in transcribers. Discourse Studies 9: 837842.
Bull, Peter E.
1987
Posture and gesture. Oxford: Pergamon.
Bull, Peter E.
1989
Psychological approaches to transcription. In: Derek B. Roger and Peter E.
Bull (eds.), Conversation: An Interdisciplinary Approach, 150165. Bristol:
Bull, Peter E. and Gerry Connelly
1985
Body movement and emphasis in speech. Journal of Nonverbal Behavior 9:
169187.
Chovil, Nicole
1989
Communicative functions of facial displays in conversation. Unpublished doctoral dissertation, Department of Psychology, University of Victoria, Victoria,
British Columbia, Canada.
674
Cohen, Jacob
1960
A coefficient of agreement for nominal scales. Educational and Psychological
Measurement 20: 3746.
Clark, Herbert H.
1996
Clark, Herbert H. and Jean E. Fox Tree
2002
Using uh and um in spontaneous speaking. Cognition 84: 73111.
Cook, Guy
1995
Theoretical issues: Transcribing the untranscribable. In: Geoffrey A. Leech,
Greg Myers and Jenny Thomas (eds.), Spoken English on Computer: Transcription, Mark-up and Application, 3553. New York: Longman.
Damico, Jack S. and Nina N. Simmons-Mackie
2002
The base layer and the gaze/gesture layer of transcription. Clinical Linguistics
& Phonetics 16: 317327.
Dressler, Richard A. and Roger J. Kreuz
2000
Transcribing oral discourse: A survey and a model system. Discourse Processes 29: 2536.
Du Bois, John W.
1991
Transcription design principles for spoken discourse research. Pragmatics 1:
71106.
Du Bois, John W.
2006
VoiceWalker: A discourse transcription utility. University of California Regents.
Du Bois, John W., Susanna Cumming, Stephan Schuetze-Coburn and Danae Paolino
1992
Discourse transcription. Santa Barbara Papers in Linguistics 4: 1225.
Du Bois, John W., Stephan Schuetze-Coburn, Susanna Cumming and Danae Paolino
1993
Outline of discourse transcription. In: Jane A. Edwards and Martin D. Lampert
(eds.), Talking Data: Transcription and Coding in Discourse Research, 4589.
Hillsdale, NJ: Lawrence Erlbaum.
Easton, Kristen L., Judith Fry McComish and Rivka Greenberg
2000
Avoiding common pitfalls in qualitative data collection and transcription.
Qualitative Health Research 10: 703707.
Edwards, Jane A.
1992
Computerized methods in child language research: four principles for the use
of archived data. Journal of Child Language 19: 435458.
Edwards, Jane A.
1993a
Perfecting research techniques in an imperfect world: Response to MacWhinney and Snow. Journal of Child Language, 20: 209216.
Edwards, Jane A.
1993b
Principles and contrasting systems of discourse transcription. In: Jane A. Edwards and Martin D. Lampert (eds.), Talking Data: Transcription and Coding
in Discourse Research, 331. Hillsdale, NJ: Lawrence Erlbaum.
Edwards, Jane A.
1995
Principles and alternative systems in the transcription, coding, and mark-up of
spoken discourse. In: Geoffrey A. Leech, Greg Myers and Jenny Thomas
(eds.), Spoken English on Computer: Transcription, Mark-up and Application,
1934. New York: Longman.
675
Ehlich, Konrad
1993
HIAT: A transcription system for discourse data. In: Jane A. Edwards and Martin D. Lampert (eds.), Talking Data: Transcription and Coding in Discourse
Research 8, 123148. Hillsdale, NJ: Lawrence Erlbaum.
Ekman, Paul and Wallace V. Friesen
1978
Facial Action Coding System. Palo Alto: Consulting Psychologist Press.
Ekman, Paul, Wallace V. Friesen and Joseph C. Hager
2002
Facial Action Coding System. CD-ROM. Salt Lake City, UT: A Human Face.
Goldin-Meadow, Susan
2003
Hearing Gesture: How our Hands Help us Think. Cambridge, MA: Harvard
University Press.
Green, Judith, Maria Franquiz and Carol Dixon
1997
The myth of the objective transcript: Transcribing as a situated act. TESOL
Quarterly 31: 172176.
Gumperz, John J. and Norine Berenz
1993
Transcribing conversational exchanges. In: Jane A. Edwards and Martin D.
Lampert (eds.), Talking Data: Transcription and Coding in Discourse Research, 91121. Hillsdale, NJ: Lawrence Erlbaum.
Gut, Ulrike, Karin Looks, Alexandra Thies, Thorsten Trippel and Dafydd Gibbon
2002
CoGesT: Conversational gesture transcription system version 1.0. Universitt
Bielefeld: Technical report 1.
Haravon, Anita, Loraine K. Obler and Martha T. Sarno
1994
A method for microanalysis of discourse in brain-damaged patients. In: Ronald
L. Bloom, Loraine K. Obler, Susan De Santi and Jonathan S. Ehrlich (eds.),
Discourse Analysis and Applications: Studies in Adult Clinical Populations,
4780. Hillsdale, NJ: Lawrence Erlbaum.
Hoiting, Nina and Dan I. Slobin
2002
Transcription as a tool for understanding: The Berkeley Transcription System
for sign language research (BTS). In: Gary Morgan and Bencie Wool (eds.),
Directions in Sign Language Acquisition, 5575. Amsterdam: Benjamins.
Jaffe, Alexandra
2007
Variability in transcription and the complexities of representation, authority
and voice. Discourse Studies 9: 831836.
Jefferson, Gail
2002
Is no an acknowledgement token? Comparing American and British uses of
(+)/(-) tokens. Journal of Pragmatics 34: 13451383.
Jefferson, Gail
2004
Glossary of transcript symbols with an introduction. In: Gene H. Lerner (ed.),
Conversation Analysis: Studies from the First Generation, 1331. Amsterdam:
Benjamins.
Kingdon, Roger
1958
The Groundwork of English Stress. London: Longman.
Kipp, Michael
2003
Anvil 4.0: Annotation of video and spoken language. Saarbrcken: University
of the Saarland.
Kowal, Sabine and Daniel C. OConnell
2004
The transcription of conversations. In: Uwe Flick, Ernst von Kardoff, Ines
Steinke and Bryan Jenner (eds.), A Companion to Qualitative Research: Para-
676
digms, Theories, Methods, Practice and Context, 248252. Thousand Oaks,

CA: Sage Publications.
Kyratzis, Amy
2001
Emotion talk in preschool same-sex friendship groups: Fluidity over time and
context. Early Education & Development 12: 359392.
1995
Spoken English on Computer: Transcription, Mark-up and Application. New
York: Longman.
Levelt, Willem J. M.
1993
Speaking: From Articulation to Intention. Cambridge, MA: MIT Press.
Lindsay, Jean and Daniel C. OConnell
1995
How do transcribers deal with audio recordings of spoken discourse? Journal
of Psycholinguistic Research 24: 101115.
Louwerse, Max M., Nick Benesh, Mohammed E. Hoque, Patrick Jeuniaux, Gwyneth Lewis,
Jie Wu and Megan Zirnstein
2007
Multimodal communication in face-to-face computer-mediated conversations.
Proceedings of the 28th Annual Conference of the Cognitive Science Society,
12351240. Mahwah, NJ: Lawrence Erlbaum.
MacWhinney, Brian
n.d.
TalkBank. Retrieved December 2008 from http://talkbank.org/.
MacWhinney, Brian
2000
The CHILDES Project: Tools for Analyzing Talk, Third Edition. Volume I:
Transcription Format and Programs. Mahwah, NJ: Lawrence Erlbaum.
MacWhinney, Brian
2001
From CHILDES to TalkBank. In: Margareta Almgren, Andoni Barrea, MaraJose Ezeizaberrena, Itziar Idiazabal, and Brian MacWhinney (eds.), Research
on Child Language Acquisition, 1734. Somerville, MA: Cascadilla.
MacWhinney, Brian and Catherine Snow
1992
The wheat and the chaff: Or four confusions regarding CHILDES. Journal of
Child Language 19: 459471.
McNeill, David
2005
Appendix. In: David McNeill, Gesture & Thought, 259288. Chicago: University of Chicago Press.
Mishler, Elliot G.
1991
Representing discourse: The rhetoric of transcription. Journal of Narrative
and Life History 1: 255280.
Mller, Nicole and Jack S. Damico
2002
A transcription toolkit: Theoretical and clinical considerations. Clinical Linguistics and Phonetics 16: 299316.
Mller, Nicole and Jacqueline A. Guendouzi
2002
Transcribing discourse: Interactions with Alzheimers disease. Clinical Linguistics and Phonetics 16: 345359.
Neidle, Carol, Stan Sclaroff and Vassilis Athitsos
2001
SignStream: A tool for linguistic and computer vision research on visual-gestural language data. Behavior Research Methods, Instruments, & Computers
33: 311320.
677
New York Times

1974
The White House Transcripts: Submission of Recorded Presidential Conversations to the Committee on the Judiciary of the House of Representatives by
President Richard Nixon. New York: Viking.
Norris, Sigrid
2004
Analyzing Multimodal Interaction: A Methodological Framework. New York:
Routledge.
Novick, David G., Lisa Walton and Karen Ward
1996
Contribution graphs in multiparty conversations. In: Proceedings of the International Symposium on Spoken Dialogue, 5356. (ISSD-96.) Philadelphia, PA.
Ochs, Elinor
1979
Transcription as theory. In: Elinor Ochs and Bambi B. Schieffelin (eds.), Developmental Pragmatics, 4372. New York: Academic Press.
OConnell, Daniel C. and Sabine Kowal
1994
Some current transcription systems for spoken discourse: A critical analysis.
1995a
Basic principles of transcription. In: Jonathan A. Smith, Rom Harre, and Luk
Van Langenhove (eds.), Rethinking Methods in Psychology, 93105. Thousand
Oaks, CA: Sage.
1995b
Transcription systems for spoken discourse. In Jef Verschueren, Jan-Ola stman, and Jan Blommaert (eds.), Handbook of Pragmatics, 646656. Amsterdam: Benjamins.
OConnell, Daniel C., and Sabine Kowal
1999
Transcription and the issue of standardization. Journal of Psycholinguistic Research 28: 103120.
OConnell, Daniel and Sabine Kowal
2000
Are transcripts reproducible? Pragmatics 10: 247269.
2008
Communicating with One Another: Toward a Psychology of Spontaneous
Spoken Discourse. New York: Springer.
OConnor, Joseph D. and Gordon F. Arnold
1961
Intonation of Colloquial English. London: Longman.
Powers, Willow R.
2005
Transcription Techniques for the Spoken Word. Oxford: AltaMira Press.
Preston, Dennis R.
1982
Ritin fowklower daun rong: Folklorists failures in phonology. Journal of
American Folklore 95: 304326.
Psathas, George and Timothy Anderson
1990
The practices of transcription in conversation analysis. Semiotica 78: 7599.
Pullum, Geoffrey K. and William A. Ladusaw
1996
Phonetic Symbol Guide, 2nd ed. Chicago: University of Chicago Press.
Pye, Clifton A., Kim A. Wilcox, and Kathleen A. Siren
1988
Refining transcription: The significance of transcriber errors. Journal of
678
Roberts, Celia
1997
Transcribing talk: Issues of representation. TESOL Quarterly, 31: 167172.
Roberts, Felicia and Jeffrey D. Robinson
2004
Interobserver agreement on first-stage conversation analytic transcription.
Human Communication Research 30: 376410.
Romero, Catherine, Daniel C. OConnell and Sabine Kowal
2002
Notation systems for transcription: An empirical investigation. Journal of Psycholinguistic Research 31: 619631.
1974
Sayette, Michael A., Jeffrey F. Cohen, Joan M. Wertz, Michael A. Perrott and Dominic J.
Parrot
2001
A psychometric evaluation of the Facial Action Coding System for assessing
spontaneous facial expression. Journal of Nonverbal Behavior 25: 167185.
1984
On some gestures relation to talk. In: J. Maxwell Atkinson and John Heritage
Schenkein, Jim
1978
Explanation of transcript notation. In: Jim Schenkein (ed.), Studies in the Organization of Conversational Interaction, xi-xvi. New York: Academic Press.
Schiffrin, Deborah
1987
Searle, John
1975
Semantics. Volume 3: Speech Acts, 5982. New York: Academic Press.
Selting, Margret, Peter Auer, Birgit Barden, Jrg Bergmann, Elizabeth Couper-Kuhlen,
Susanne Gnthner, Christoph Meier, Uta Quasthoff, Peter Schlobinski and
Susanne Uhmann
1998
Gesprchsanalytisches Transkriptionssytem (GAT). Linguistische Berichte
173: 91122.
Shriberg, Lawrence D. and Gregory L. Lof
1991
Reliability studies in broad and narrow phonetic transcription. Clinical Linguistics and Phonetics 5: 225279.
Stelma, Juurd H. and Lynne J. Cameron
2007
Intonation units in spoken interaction: Developing transcription skills. Text &
Talk 27: 361393.
Svartvik, Jan (ed.).
1990
The London Corpus of Spoken English: Description and Research. Lund: Lund
University Press.
Svartvik, Jan and Randolph Quirk
1980
A Corpus of English Conversation. Lund: Lund University Press.
Tannen, Deborah
1984/2005 Conversational Style: Analyzing Talk Among Friends. Oxford: Oxford University Press.
679
Tilley, Susan A.
2003
Challenging research practices: Turning a critical lens on the work of transcription. Qualitative Inquiry 9: 750773.
Woods, David K.
2007
Transana. Retrieved December 2008 from http://www.transana.org.
About the authors
Gisle Andersen is Associate Professor in the Department of Professional and

Intercultural Communication (Norwegian School of Economics and Business Administration). He is the author of Pragmatic Markers and Sociolinguistic Variation a Relevance-theoretic Approach to the Language of Adolescents (Benjamins, 2001) and has co-authored Trends in Teenage Talk Corpus Compilation,
Analysis and Findings (Benjamins, 2002), Pragmatic Markers and Propositional
Attitude (Benjamins, 2000) and is currently co-editing Pragmatics of Society,
which is the 5th volume in the current handbook series. His research mainly
focuses on different aspects of spoken interaction, and his work deals especially
with the use of corpora for studies in pragmatics and sociolinguistics, covering
such topics as discourse markers, vague language and listenership. He has also
worked on written communication, lexicography and terminology, and the influence
of English on the Norwegian language. He has been deeply involved in various
corpus compilation projects, including COLT (The Bergen Corpus of London Teenage Language) and the Norwegian Newspaper Corpus, and has coordinated and
participated in projects within language technology and language resources. He is
participating in various projects funded by the European Commission and the Norwegian Research Council. He is also a board member of the ICAME organisation
(International Computer Archive of Modern and Medieval English).
Monika Bednarek is Lecturer in Linguistics at the University of Sydney, Australia. Her research interests include news and television discourse, corpus-based discourse analysis, ecolinguistics, and the linguistic expression of opinion and emotion. She is the author of Evaluation in Media Discourse (Continuum, 2006),
Emotion Talk Across Corpora (Palgrave Macmillan, 2008) and The Language of
Fictional Television: Drama and Identity (Continuum, 2010). Other publications
on these and related research areas include journal articles in Discourse & Communication, Functions of Language, Text and Talk and the Journal of Pragmatics.
Richard Breheny is Reader in Linguistics in the Division of Psychology and Language Sciences, University College London, UK. His research is in semantics,
pragmatics, and experimental pragmatics, including online processing and pragmatic development. Recent publications include Tian, Y., Breheny, R. & H.J. Ferguson Why we simulate negated information: a dynamic pragmatic account,
Quarterly Journal of Experimental Psychology 63, 2010; Breheny, R. A new look
at the semantics and pragmatics of numerically quantified noun phrases, Journal
682
About the authors
of Semantics 25, 2008; Breheny, R., Katsos, N. & J. Williams Are scalar implicatures generated on-line by default?, Cognition 100, 2006; Breheny, R. Communication and folk psychology, Mind and Language 21, 2006.
Wolfram Bublitz is Full Professor of English Linguistics at the University of
Augsburg, Germany. He received his doctorate from the University of Hamburg
and his habilitation from the University of Trier, Germany. He has also been Full
Professor at the University of Braunschweig, Exchange Professor at Northern Illinois University and Visiting Professor at the Universities of Pittsburgh and Hamburg. His research interests concern all areas of text analysis, (constructivist) pragmatics and recently computer-mediated communication, with an emphasis on
modal and discourse particles, modality and negation, discourse topic and topical
acts, repetition and other forms of parallelism, coherence and cohesion, collocation
and semantic prosody, politeness and evaluation, context and common ground,
systems of knowledge and memory, hypertext and quoting. His monographs and
edited books include Englische Pragmatik (2nd ed. Schmidt, 2009), Metapragmatics in Use (ed. with A. Hbler, Benjamins, 2007), Coherence in Spoken and
Written Discourse: How to Create It and How to Describe It (ed. with U. Lenk and
E. Ventola, Benjamins, 1999).
Piotr Cap is Professor of Linguistics in the Institute of English, University of
dz, Poland. His research interests are in pragmatics, (critical) discourse analysis,
political linguistics, business communication, and methodology of linguistic
analysis. His monographic and (co-)edited publications include Explorations in
Political Discourse (Lang, 2002), Pragmatics Today (Lang, 2005), Legitimisation
in Political Discourse (Cambridge Scholars Publishing, 2006/2008), New Approaches to Discourse and Business Communication (Palgrave Macmillan, 2009),
and Perspectives in Politics and Discourse (Benjamins, 2010). He is Managing
Editor of International Review of Pragmatics (Brill) and member of editorial and
advisory boards of several journals in pragmatics and discourse analysis.
Elena Collavin is a Lecturer in the Department of Communication at the University
of California San Diego, USA. Her work has focused on Italian political debate and
on discourse surrounding food biotechnologies in Italy. More recently she has
studied everyday family conversations among Friulian speakers in San Giorgio di
Nogaro, in north eastern Italy. Her recent publications include Participation in
Italy: The public and genetically modified crops, Democratic Transgressions of
Law, ed. by A. Bora and H. Hausendorf (Brill, 2010); Marketing concern in food
products, IF, Journal of Italo-Finnish Studies, 2008; Tavano, A. and E. Collavin,
Pragmatiche furlane, Manul di Lenghistiche Furlane, ed. by F. Fabbro (Udine,
Forum Edizioni, 2007); Food Biotechnologies in Italy: A Social Psychological
Study (Social Psychology Department, Helsinki University, PhD dissertation, 2007).
About the authors
683
Manfred Consten is Lecturer for discourse linguistics and linguistic pragmatics at

the Institute for Germanic Linguistics, University of Jena, Germany. His main research areas are anaphoric reference, cognitive approaches to text coherence (esp.
the discourse function of demonstratives), and meta-linguistics. His doctoral thesis
(2004) suggests a unified model of anaphora and deixis as instances of domain
bound reference. He is co-editor of the volume Anaphors in Text. Cognitive, Formal and Applied Approaches to Anaphoric Reference (Benjamins, 2007). Fuller information is provided at www.textlinguistik.net.
Maeve Cooke is Professor and Head of the School of Philosophy at University
College Dublin, Ireland, and a member of the Royal Irish Academy. Her current
research focuses on re-conceptualising truth, freedom and power and on issues
in current debates on religion and politics. Her principal book publications are
Language and Reason: A Study of Habermass Pragmatics (MIT Press, 1994) and
Re-Presenting the Good Society (MIT Press, 2006). She is editor and translator of
Habermas: On the Pragmatics of Communication (MIT Press, 1998) and has published numerous articles in scholarly journals and books, mainly in the areas of
social and political philosophy.
Elizabeth Couper-Kuhlen is Finland Distinguished Professor at the University
of Helsinki, Finland. Her main research areas are situated in the field of language
use in interaction, including prosody, grammar and most recently, affect and emotion. Among her main publications are the monographs An Introduction to English Prosody (Niemeyer and Edward Arnold, 1986), English Speech Rhythm (Benjamins, 1993), Language in Time (with P. Auer & F. Mueller, Oxford University
Press, 1999) and the edited volumes Prosody in Conversation (with M. Selting,
Cambridge University Press, 1996), Cause, Condition, Contrast, Concession:
Cognitive and Discourse Perspectives (with B. Kortmann, de Gruyter Mouton.
2000), Studies in Interactional Linguistics (with M. Selting, Benjamins, 2001) and
Sound Patterns in Interaction (with C.E. Ford, Benjamins, 2004).
Saskia Daalder teaches Dutch linguistics at the Vrije University Amsterdam, The
Netherlands. Her main research areas are the pragmatics of verbal communication (Relevance Theory), the theory of grammar in the light of pragmatic theory,
and the history of twentieth-century linguistics (in particular, the Prague School
of functional linguistics and its reflections in the Netherlands). Selected publications are Conditional constructions: The special case of modern Dutch mits,
Journal of Germanic Linguistics 21, 2009; H.J. Pos (18981955): Studies over zijn
filosofie van taal en taalwetenschap (2 vols, Mnster: Nodus-Publikationen,
1999).
684
About the authors
Arnulf Deppermann is head of the Pragmatics Department at the Institut fr

Deutsche Sprache and Professor for German Linguistics at Mannheim University,
Germany. His areas of research cover conversation analysis, cognition and understanding in interaction, multimodal interaction, positioning in narratives, medical
and therapeutic interaction, construction grammar and semantics in talk-in-interaction. Main publications are Gesprche analysieren (Wiesbaden: VS, 42008);
Grammatik und Semantik aus gesprchsanalytischer Sicht (de Gruyter, 2007); Verstehen in professionellen Handlungsfeldern (with U. Reitemeier, R. Schmitt and
Th. Spranz-Fogasy, Narr, 2010); Data and transcription, (together with W.
Schtte), Handbook of Interpersonal Communication, ed. by G. Antos et al (de
Gruyter, 2008); Constructions vs. lexical items as sources of complex meanings,
Emergent constructions, ed. by P. Auer and S. Pfnder (de Gruyter, 2010); How
does cognition matter to the analysis of talk-in-interaction?, Language Sciences
33, 2011.
Christine Domke is a research assistant at the department of German studies at the
Technical University Chemnitz, Germany. She holds a doctorate in linguistics
from the University of Bielefeld with a thesis on the process of decision-making in
business meetings. Currently she is working on her habilitation about the characteristics of place bound communication in railway stations, airports and inner
cities. Her major research interests are in combining organisation theory and organisational communication based on qualitative approaches like conversational
analysis (CA), and also in the semiotics of mass media texts and recent forms of
advertising. Her publications include Besprechungen als organisationale Entscheidungskommunikation (de Gruyter, 2006), Name: Deutsch, Alter: 1200, Befund: Gesund, (with J. Kilian), Aisthesis 2009; Wissen in (Inter)Aktion (with
U. Dausendschn-Gay and Sren Ohlhus (de Gruyter, 2010).
Anita Fetzer is Full Professor of English linguistics at Wrzburg University, Germany. She received her doctorate from Stuttgart University in 1993 and her habilitation in 2003, and is currently engaged in research projects on evidentiality, the
strategic use of pronouns and attitudinal markers in political discourse, and on the
form and function of the theme zone in present-day English. She has had a series of
articles published on rejections, context, political interviews and intercultural
communication. Her most recent publications are Pragmatic and discourse-analytic approaches to present-day English, (with Karin Aijmer), Journal of Pragmatics 2008, Political Discourse in the Media (with Gerda Lauerbach, Benjamins,
2007), Lexical Markers of Common Grounds (with Kerstin Fischer, Elsevier,
2007), Context and Appropriateness (Benjamins, 2007), and Recontextualizing
Context: Grammaticality Meets Appropriateness (Benjamins, 2004).
About the authors
685
William Hanks earned the Joint PhD in Anthropology and Linguistics at The University of Chicago, 1983. Since 2000, he has held the Berkeley Distinguished
Chair in Linguistic Anthropology at University of California, Berkeley, USA. His
research bears on semantics and pragmatics, with a special focus on indexicality in
interaction, ordinary referring and ritual speech in Yucatec Maya. His relevant
books include Referential Practice, Language and Lived space among the Maya
(University of Chicago Press, 1990), Language and Communicative Practices
(Westview Press, 1995), and Converting Words, Maya in the Age of the Cross (University of California Press, 2010, winner of the Edward Sapir Book Prize, Society
for Linguistic Anthropology).
Werner Holly is Full Professor of Germanistische Sprachwissenschaft at the
Technical University Chemnitz, Germany. His main research areas are: pragmatics, linguistics of text and conversation, language in politics, language and the
media, audiovisuality. Major publications include Imagearbeit in Gesprchen
(Niemeyer, 1979); Politikersprache (de Gruyter, 1990); Politische Fernsehdiskussionen (with Peter Khn and Ulrich Pschel, Niemeyer, 1986); Der sprechende
Zuschauer (ed. with Ulrich Pschel and Jrg Bergmann, Westdeutscher Verlag,
2001); Einfhrung in die Pragmalinguistik (Langenscheidt, 2001); Fernsehen
(Niemeyer, 2004); ber Geld spricht man (with Stephan Habscheid et al.,Verlag
fr Sozialwissenschaft, 2006); Linguistische Hermeneutik (ed. with Fritz Hermanns, Niemeyer, 2007).
Yan Huang (PhD Cambridge, DPhil Oxford) is Professor of Linguistics at the University of Auckland, New Zealand. He has previously taught linguistics at the
universities of Cambridge, Oxford, and Reading, UK, where he was Professor of
Theoretical Linguistics. His main research interests are in pragmatics, especially
the pragmatics-semantics interface and the pragmatics-syntax interface including
anaphora. His books include The Syntax and Pragmatics of Anaphora (Cambridge
University Press, 1994, re-issued in 2007), Anaphora: A Cross-Linguistic Study
(Oxford University Press, 2000), and Pragmatics (Oxford University Press, 2007).
His Pragmatics is being translated into a number of languages, and the Chinese
edition and the Korean translation have already been published. He has also published a number of articles and reviews in leading international journals of linguistics. He has been invited to lecture in around 90 universities and research institutes in many countries in Europe, North America, Asia, Australasia, and North
Africa.
Axel Hbler is Professor of English Linguistics; before his retirement in 2009, he
held the chair of English Linguistics at the University of Jena, Germany. His main
research areas are Semantics, (Historical) Pragmatics and Nonverbal Communication. Main publications include Understatements and Hedges in English (Benja-
686
About the authors
mins, 1983), The Expressivity of Grammar: Grammatical Devices Expressing

Emotion across Time (de Gruyter, 1998), Das Konzept Krper in den Sprachund Kommunikationswissenschaften (A. Francke, 2001), The Nonverbal Shift in
Early Modern English Conversation (Benjamins, 2007), Introducing Metapragmatics in Use (with W. Bublitz), Metapragmatics in Use, ed. by W. Bublitz and
A. Hbler (Benjamins, 2007).
Nikola Kompa is Assistant Professor of Theoretical Philosophy at the University
of Bern in Switzerland where she teaches philosophy. She works in philosophy of
language, epistemology, metaphysics, logic and the history of philosophy. Within
philosophy of language, she is particularly interested in theories of meaning, the semantics/pragmatics-interface, the connection between meaning and categorization,
vagueness, and context-sensitivity. She is author of Wissen und Kontext (mentis,
2001) and co-editor of The Apriori and its Role in Philosophy (mentis, 2009).
Wataru Koyama is Professor of Pragmatics and Linguistics at the Graduate
School of Intercultural Communication, Rikkyo (St. Pauls) University, Tokyo,
Japan. He received his doctorate in linguistics from The University of Chicago in
2003. His particular research interests focus on language in culture, social semiotics, and socio-historical pragmatics. Some of his publications have appeared
in Journal of Pragmatics, Studies in Language, Rask, Language and Communication, and the Encyclopedia of Language and Linguistics (2nd edition).
Roger Kreuz is Professor of Psychology at the University of Memphis, USA, a
position that he has held since 1988. He is also the director of the departments
masters program. He earned his masters and doctoral degrees in cognitive psychology at Princeton University, and undergraduate degrees in psychology and linguistics at the University of Toledo. Before joining the University of Memphis faculty, he held a post-doctoral position at Duke University. His primary research
areas are discourse processing and pragmatics, nonliteral language production and
comprehension, and computer-mediated communication. With Susan Fussell, he
edited Social and Cognitive Approaches to Interpersonal Communication (Erlbaum, 1998).
Sophia Marmaridou is Professor of Linguistics at the Faculty of English Language and Literature of the University of Athens, Greece. She holds a BA in English and Greek Language and Literature from the University of Athens and a PhD
in Linguistics from the University of Cambridge. Her main research areas are Semantics, Pragmatics, Cognitive Linguistics, Construction Grammar, and Lexicography. Some of her recent publications are: Pragmatic Meaning and Cognition
(Benjamins, 2000), On the conceptual, cultural and discursive motivation of
Greek pain lexicalizations, Cognitive Linguistics 17, 2006; Cognitive, cultural,
About the authors
687
and constructional motivations of polysemy and semantic change. The case of the
Greek , Pragmatics and Cognition 18, 2010. She is one of the editors of
Pragmatics, a member of the Editorial Board of the Pragmatics & Beyond New
Series (Benjamins), and a Board Member of the International Pragmatics Association. She is also a member of the International Cognitive Linguistics Association.
Georg Meggle is Professor Emeritus of Philosophy, Logic, Anthropology, Cognitice Sciences and Ethics at the universities of Saarbrcken, Mnster and Leipzig,
Germany. His main research topics include Theory of Action, Communication and
Meaning, Collective Intentionality, Ethics and Applied Ethics, in particular of War
and Terrorism. His publications include: Grundbegriffe der Kommunikation (2nd
ed. de Gruyter, 1997); Sprache und Ethik (with G. Grewendorf, Suhrkamp, 1974);
Analytische Handlungstheorie (Suhrkamp, 1977, re-issued 1985), Handlung,
Kommunikation, Bedeutung (Suhrkamp, 1979; Sprachphilosophie Philosophy of
Language La philosophie du langage (ed. with M. Dascal, D. Gerhardus, K. Lorenz, de Gruyter, vol 1: 1992, vol. 2: 1995); Analyomen 1 Proceedings of the 1st
Conference on Perspectives in Analytical Philosophy (ed. with U. Wessels, de
Gruyter, 1994); Zum moralischen Denken (with C. Fehige, Suhrkamp, 1994, 2
vols); ANALYOMEN 2 Proceedings of the 2nd Conference on Perspectives in
Analytical Philosophy (de Gruyter, 1997, 3 vols); Actions, Norms, Values: Discussions with Georg Henrik von Wright (de Gruyter, 1998); Der Sinn des Lebens
(ed. with C. Fehlige and U. Wessels, dtv, 2000); Speech Acts, Mind, and Social
Reality (ed. with G. Grewendorf, Kluwer, 2002); Social Facts & Collective Intentionality (German Library of Sciences, 2002); Saying, Meaning, Implicating (ed.
with C. Plunze, Leipzig Universittsverlag, 2003); Terror und Der Krieg gegen ihn
(mentis, 2003); Humanitre Interventionsethik (mentis, 2004); Ethics of Terrorism
and Counter-Terrorism (ontos, 2005); Deutschland Israel Palstina. Streitschriften (Europische Verlagsanstalt, 2007).
Jacob L. Mey is Professor Emeritus of Linguistics at the University of Southern
Denmark. Previous appointments include the University of Oslo, the University of
Texas at Austin, Georgetown University, Tsukuba University, The National Language Research Institute, Tokyo, City University of Hong Kong, University of
Frankfurt, University of Campinas, University of Braslia, University of Haifa,
Shanghai International Studies University, University of Graz, University of
Science and Technology, Kuwait, as well as numerous other institutions of research and higher learning. His research interests concern all areas of pragmatics,
with an emphasis on the social aspects of language use, the pragmatic impact of
computer technologies, and the pragmatic use of literary devices. Among his most
recent publications in these areas are Pragmatics: An Introduction (Blackwell,
2001); When Voices Clash: A Study in Literary Pragmatics (de Gruyter, 2000); As
Vozes da Sociedade (The Voices of Society; Campinas, S.P.: Mercado de Letras,
688
About the authors
2004); Cognition and Technology (with B. Gorayska, Benjamins, 2004); Intention,

Common Ground and the Egocentric Speaker-Hearer (with I. Kecskes, de Gruyter,
2008). He edited the 1200-page Concise Encyclopedia of Pragmatics (Elsevier,
2009). In 1977, he founded (with Hartmut Haberland) the Journal of Pragmatics,
of which he was Editor-in-Chief until Elsevier terminated him in 2009. That same
year, he founded the new journal Pragmatics and Society (Benjamins), the first two
issues of which have appeared in 2010. He holds honorary Dr. Phil. degrees from
the Universities of Zaragoza, Spain (1993) and Bucharest, Romania (2006). In
2008, he was presented with a Life Time Award from the University of Southern
Denmark for his work in pragmatics.
Andreas Musolff is Professor of Intercultural Communication at the University of
East Anglia in Norwich, UK. He has published on the History of Political Discourse, Metaphor Theory and the History of Pragmatics. His monographs include
Metaphor, Nation and the Holocaust (Routledge, 2010), Metaphor and Political
Discourse (Palgrave Macmillan, 2004), Mirror Images of Europe (Iudicium,
2000), and Kommunikative Kreativitt. Karl Bhlers Zweifelderlehre als Ansatz zu
einer Theorie innovativen Sprachgebrauchs (1990). He is a member of the Executive Committee of the Societas Linguistica Europaea and of the editorial boards
of Aptum, Zeitschrift fr Sprachkritik und Sprachkultur, Journal of Germanic
Studies, German as a Foreign Language, Metaphor and the Social World.
Winfried Nth is Emeritus Professor of Linguistics and Semiotics at the University of Kassel, Germany, and Visiting Professor at the Catholic University of So
Paulo. His 250 articles and 27 authored or edited books are on topics of English linguistics (esp. semantic opposites), semiotic aspects of language, literature, the
image, maps, the media, systems theory, culture, and evolution. His Handbook of
Semiotics (Bloomington, 1990, transl. into Bahasa, and in a rev. transl. into German, Metzler, 2000, and Croatian) was awarded the Choice Outstanding Academic
Book prize. Among his books are: Literatursemiotische Analysen zu Lewis Carrolls Alice-Bchern (Narr, 1980), Origins of Semiosis (ed., de Gruyter, 1994), Semiotics of the Media (ed., de Gruyter, 1997), Crisis of Representation (ed. with C.
Ljungberg, de Gruyter, 2003), Imagen: Comunicacin, semitica y medios (Reichenberger, 2003), Comunicao e semitica (Hacker, 2004), Self-Reference in the
Media (ed. with N. Bishara, de Gruyter, 2007), Mediale Selbstreferenz: Grundlagen und Fallstudien zu Werbung, Computerspiel und Comics (with B. Bishara
and B. Neitzel, Halem, 2008), and Estratgias semiticas da publicidade (with L.
Santaella, Cengage, 2010).
Neal R. Norrick holds the chair of English Linguistics at Saarland University in
Saarbrcken, Germany. His research specializations include conversation, narrative, verbal humor and formulaicity. His authored publications include Conver-
About the authors
689
sational Joking: Humor in Everyday Talk (Indiana University Press, 1993) and
How Proverbs Mean (Mouton, 1985); his recent edited volumes are Phraseology:
An International Handbook of Contemporary Research, HSK vol. 28, 12 (ed. with
H. Burger, D. Dobrovolskij, and P. Khn, de Gruyter, 2007) and Humor in Interaction (ed. with Delia Chiaro, Benjamins, 2009). His monograph Conversational
Narrative: Storytelling in Everyday Talk (Benjamins, 2000) was reprinted in a paperback edition in 2010. He acts as Co-Editor in Chief of the Journal of Pragmatics, and serves on the advisory boards of the journals Discourse Processes,
Humor, International Review of Pragmatics and Text & Talk.
Monica A. Riordan earned her PhD in Experimental Psychology at the University of Memphis, USA. She earned her masters degree at the University of Memphis and her undergraduate degree at the University of Missouri-Columbia. Her
primary research interest is the expression of emotion in computer-mediated communication. Representative publications include: A survey of emotion expression
in CMC: Reasons for channel choice, (with R.J. Kreuz), Computers in Human Behavior 2010, Cues in online conversation: A corpus analysis, (with R.J. Kreuz),
Computers in Human Behavior.
Christoph Rhlemann is a full-time secondary school teacher and a part-time
Lecturer in modern linguistics at Ludwig-Maximilians-University, Mnchen, Germany. He is the author of Conversation in Context: A Corpus-driven Approach
(Continuum, 2007) and Narrative in English Conversation (Cambridge University
Press, forthcoming). He has published on different topics relating to conversational
English in edited collections and journals such as Applied Linguistics, the ICAME
Journal, the International Journal of Corpus Linguistics, and the Journal of English Linguistics. His main interests are in corpus linguistics, pragmatics and sociolinguistics. He is currently co-editing (with Karin Aijmer) a research volume on
Corpus Pragmatics, covering the fast-growing field of pragmatic studies based on
corpus methods. Further, he is involved in the construction, annotation and analysis of the Narrative Corpus, a corpus of conversational narratives extracted from
the British National Corpus.
John Saeed is a Fellow of Trinity College Dublin, Ireland, and Associate Professor of Linguistics in its School of Linguistic, Speech and Communication
Sciences. His PhD at the School of Oriental and African Studies, London was on
information structure in Somali and his subsequent research has concerned the
grammar and pragmatics interface in a number of languages, most recently in Irish
Sign Language. His books include Somali Reference Grammar (2nd ed. Dunwoody
Press, 1993), Somali (Benjamins, 1999) and Semantics (3rd ed. Wiley-Blackwell,
2009).
690
About the authors
Monika Schwarz-Friesel is Full Professor of General Linguistics at the Technical University Berlin, Germany. Her research focuses on cognitive semantics/pragmatics, the interaction of language, cognition and emotion and discourse comprehension processes, especially the problem of anaphor resolution and coherence.
She is the author and editor of several books on semantics (Kognitive Semantiktheorie und neuropsychologische Realitt, Niemeyer, 1992; Semantik. Ein Arbeitsbuch, Narr, 5th ed. 2007, with J. Chur; Metapher, Winter, 2007, with H. Skirl), cognitive linguistics (ed., Cognitive Semantics, Narr, 1994; Einfhrung in die kognitive
Linguistik, Francke, 3rd ed. 2008) and discourse linguistics (Indirekte Anaphern in
Texten, Niemeyer, 2000; Sprache und Emotion, Francke, 2007; Anaphors in Text,
ed. with M. Consten and M. Knees, Benjamins, 2007; Aktueller Antisemitismus
ein Phnomen der Mitte, ed. with E. Friesel and J. Reinharz, de Gruyter, 2010).
Name index
691
Name index
A
Aarsleff, H. 152
Aarts, J. 592
Abbott, C. 333
Abelson, R.P. 480
Abraham, W. 599
Abse, D.W. 246
Adolphs, S. 539, 542543, 547, 549, 609,
648650
Adorno, T.W. 303
del, A. 115, 589, 609
Agha, A. 143, 390
Agirre, E. 464
Ahlsn, E. 670
Aijmer, K. 39, 592594, 597602, 651
Aitchison, J. 443
Akman, V. 31, 3435
Algeo, J. 591
Allan, W. S. 602
Allwood, J. 194, 670
Alston, W.P. 69, 221, 379
Altenberg, B. 591595
Altmann, G.T.M. 568, 574
Andersen, C. 194
Andersen, G. 17, 26, 364, 547, 590,
593, 595, 597599, 605606, 610,
636637
Anderson, A. 572
Anderson, R.C. 544
Anderson, S.R. 319320, 326, 328332,
334335, 463
Anderson, T. 663
Andrade, M.J. 329, 332333
Anton, C. 108, 117
Anward, J. 496
Apel, K.-O. 147, 154, 156, 290, 307
Apperly, I.A. 571
Archer, D. 10, 87, 97, 100
Ariel, M. 39, 41, 43, 353, 433, 445, 447
Armstrong, L.E. 491492
Arnold, G.F. 491492, 664
Arnovick, L.K. 95, 124
Athitsos, V. 661
Atkinson, J.M. 267, 663
Atlas, J.D. 69, 400, 402, 405, 416, 562
Attardo, S. 61
Ashby, W.J. 431
Asher, N. 358, 363
Aston, G. 592
Auer, P. 26, 116, 156, 271, 273274,
429430, 435, 438, 446, 496497,
499500, 504, 665
Austen, J. 512, 518, 521522, 524, 531
Austin, J.L. 23, 5, 7, 11, 24, 31, 3940,
128, 143149, 153, 156159, 177, 190,
213, 220221, 223224, 272, 294296,
298, 373384, 390392, 461, 539, 561,
687
Averintseva-Klisch, M. 354
Aya, R. 266, 272
Ayers, G.M. 504
B
Bach, K. 28, 31, 4243, 69, 209210, 221,
386, 411413, 464
Baker, P. 549553, 609
Bakhtin, M.M. 16, 113, 517, 531532
Bal, M. 514, 521
Bald, W.D. 593
Baldry, A. 650651
Ball, M.J. 670
Ballmer, T. 120, 130
Bally, C. 113, 118119, 141, 158, 240
Bamberg, M. 544
Bamford, J. 130
Bang, J.C. 81
Banjo, A. 592
Bar-Hillel, Y. 13, 25, 141142, 148,
167
Barclay, R.R. 349
Barden, B. 665
Bardovi-Harlig, K. 9091, 552
Barker, C. 465, 552
Barlow, M. 429
692
Name index
Barr, D.J. 571574

Barth-Weingarten, D. 450
Barwise, J. 213
Baszanger, I. 545546
Bates, E. 81, 429
Bateson, G. 36, 53, 130, 156, 271
Bauer, L. 593
Bauman, R. 56, 69, 144, 157
Bavelas, J.B. 661, 666667
Bazzanella, C. 52
Beaugrande, R. de 35, 364, 588
Beaver, D. 59
Beavin, J.H. 176
Beck, G. 247
Beckman, M.E. 504
Bednarek, M. 16, 26, 538539, 544,
548551, 608, 632633
Beebe, L.M. 89
Behaghel, O. 444
Behne, T. 579
Benhabib, S. 306
Bennett, J. 11, 217, 220
Benseler, F. 131
Benveniste, . 141142, 147, 158, 245,
319320, 334
Benwell, B. 540
Benz, A. 412
Berens, F.-J. 274
Berenz, N. 665
Berger, P. 149, 492
Bergman, M. 182, 188, 194
Bergmann, J.R. 265, 268274, 276277,
665
Bergsland, K. 333
Berkum, J.J.A. van 580
Bernicot, J. 125
Bertolet, R. 63
Beule, J. de 450
Bezuidenhout, A. 209, 212
Biber, D. 56, 9596, 443, 588589,
592593, 609, 640641, 651
Bickel, B. 334
Biondi, M. 130
Bird, C.M. 672
Birdwhistell, R.L. 650
Birner, B. 27, 39, 434
Bishop, D. 576577
Bjrge, A.-K. 598, 606
Blackburn, P. 34
Blakemore, D. 413, 472, 475
Blommaert, J. 153
Bloom, L. 669
Bloom, P. 212
Bloomfield, L. 317
Blum-Kulka, S. 89, 542
Blutner, R. 412
Boas, F. 142, 144, 150, 326, 328329,
332333, 545
Bod, R. 429
Bodman, J. 9091
Bolinger, D. 147, 153, 491, 502
Bonvillain, N. 333
Borg, E. 205, 208211, 466, 574
Bornkessel-Schlesewsky, I. 450
Bosch, P. 364
Bott, L. 564, 577, 580
Bou-Franch, P. 87, 9293
Bouquet, P. 31, 34, 36, 43
Bourdieu, P. 149
Brandom, R.B. 27, 36, 204, 223
Bransford, J.D. 349
Brazil, D. 276, 591, 607
Breheny, R. 6, 16, 26, 544, 564565,
567570, 574, 579
Brennan, S.E. 446, 572573
Brennenstuhl, W. 120, 130
Brinton, L.J. 95, 599, 610
Brock, J. 190, 192, 195
Broek, P. van den 364
Brodda, B. 592
Brown, P. 25, 2728, 31, 40, 89, 144, 157,
277, 386, 538
Brown G. 247, 544
Brown J.D. 94
Brown-Schmidt, S. 571, 573
Brugmann, K. 158, 234, 328, 334
Brumfit, C. 85
Bruner, J. 448
Buba, M. 325, 330, 334
Bublitz, W. 23, 25, 3334, 45, 5253,
109, 111, 147, 275277, 443, 539,
547, 591, 593, 598, 604605,
631633
Bucholtz, M. 660, 671672
Buchstaller, I. 636637
Bhler, C. 247
Name index
Bhler, K. 2, 7, 1112, 68, 141142, 146,
149, 156, 158159, 170, 229230,
232239, 241249, 277, 298, 309, 315,
318, 328331, 334, 428, 440, 442, 525,
531
Bring, D. 433434, 464
Bull, P.E. 668
Bultinck, B. 411
Burenhult, N. 334
Burkhardt, A. 194
Burks, A.W. 244, 248
Burnard, L. 592
Burton-Roberts, N. 411
Busse, W. 247
Butler, J. 146, 549
Butterfill, S.A. 571
Buttrick, S. 322, 334, 348
Buyssens, E. 175
Bybee, J. 429, 445, 447
Byron, D.K. 364
C
Caffi, C. 107, 120, 127129, 131, 155156
Calvo Perez, J. 324, 333
Cameron, D. 110, 146
Cameron, L.J. 672
Cameron-Faulkner, T. 448
Camhy, D.G. 247
Candlin, C.N. 85, 146
Cap, P. 9, 61
Caple, H. 550
Cappelen, H. 15, 412, 466
Caramazza, A. 349
Carnap, R. 1, 7, 11, 142, 156, 159, 167,
206207, 561
Carston, R. 6, 15, 27, 69, 211, 214,
410412, 464, 467, 470, 473474,
476478, 574
Carter, D.M. 364
Carter, R. 594, 610, 634, 650
Cassirer, E.A. 149
Caton, S.C. 142, 157
Cermk, F. 537
Chafe, W.L. 364, 431433, 553, 587588,
592, 596
Channell, J. 631
Chapman, S(iobhan) 142, 154, 158
Chapman, S(teve) 537, 545
693
Chemla, E. 580
Chen, L. 594
Chien, Y.-C. 577
Chierchia, G. 400, 411, 565
Chilton, P. 67
Chomsky, N. 81, 188, 292293, 387, 426,
448, 466467, 484, 596
Chovil, N. 666667
Cicourel, A. 319
Clark C. 118
Clark E.V. 212, 448449
Clark, H.H. 16, 34, 36, 322, 334, 348349,
359, 364, 382, 432, 446, 561, 567, 572,
660
Clarke D.D. 140141, 147148, 150151,
158159, 236, 246249
Claudi, U. 444
Clayman, S. 272
Clift, R. 496
Coates, J. 592
Coffa, J.A. 148
Coffin, C. 549
Cohen, J. 672
Cohen, J.F. 672
Cohen, J.L. 213
Cohen, P.R. 28, 32, 36, 43, 387388
Colapietro, V. 177, 183184, 194
Collins, P. 588
Comrie, B. 147
Comte, A. 289
Connelly, G. 668
Conroy, A. 577578
Consten, M. 13, 354, 358359, 361364,
637
Conte, M.E. 247
Cook, G. 598, 630, 660
Cook, H.M. 94
Cooke, M. 12, 31, 246, 298299, 302, 304,
306309
Cooper, R.M. 565
Cooren, F. 192195
Cornish, F. 359, 364
Corriveau, J.-P. 463
Coulmas, F. 277, 644
Coulthard, M. 273276, 388, 598, 606608
Couper-Kuhlen, E. 15, 278, 430, 435436,
450, 491, 494, 496500, 503504,
597598, 602, 665
694
Name index
Coupland, N. 110, 113, 146, 159, 247

Courts, J. 192
Cousins, W.E. 334
Crain, S. 400, 575
Cresti, E. 593
Croft, W. 211, 214, 425, 428429, 603
Crowdy, S. 592
Cruse, A.D. 211, 214
Cruttenden, A. 603
Crystal, D. 450, 491492, 591
Csibra, G. 579
Culpeper, J. 10, 87, 9697, 100
Cummings, L. 23, 36, 44, 52, 416
Curl, T.S. 494, 499
Cutler, A. 499
D
Dahl, M. 551
Dahl, . 357
Damasio, H. 349
Damico, J.S. 662, 669670
Danes, F. 248
Danon-Boileau, L. 334
Dariusz, G. 159
Dascal, M. 7778, 8182
Dausendschn-Gay, U. 272
Davidson, B.D. 591
Davidson, D. 192, 309, 461
Davies, M. 595, 631
Davis, P.W. 333
Davis, S. 142, 381
Davis, W.A. 69, 223
Deacon, T. 178
Delaney, S.M. 125
Deledalle, G. 194
Dell, H. 12, 245, 275, 546
Denny, J.P. 326, 328329, 332333
Deppermann, A. 14, 278, 437, 450
Derrida, J. 177
Derry, S.J. 350
Deutschmann, M. 609
Dez, J. 334
Diessel, H. 324, 327332, 334335
Diewald, G. 438, 445
Dijk, T.A. van 2526, 40, 65, 67, 6970,
275, 364
Dik, S.C. 427
Dilthey, W. 289
Dilts, P. 651
Dingwall, R. 118
Dixon, C. 672
Dixon, R.M.W. 329, 334
Dodier, N. 545546
Drnyei, Z. 537, 542544, 546, 551, 553
Dr, J. 81
Domke, C. 12, 269, 430, 547
Donnellan, K.S. 142, 348349
Downing, A. 593
Dowty, D.R. 348
Dressler R.A. 662, 666
Dressler, W.U. 35, 364
Drew, P. 269270, 272
Driel, H. 194
Du Bois, J.W. 144, 159, 431, 504, 661, 664
Ducrot, O. 156
Dunn, J.A. 333
Duranti, A. 26, 34, 52, 143144, 156, 159,
244
E
Easton, K.L. 670
Eco, U. 334
Edmonds, P. 464
Edwards, D. 435
Edwards, J.A. 597, 608, 659661, 669
Eelen, G. 89
Eggins, S. 428, 608
Ehlers, K.-H. 248
Ehlich, K. 31, 247, 275, 278, 355, 428, 659,
665, 668, 670
Eisele, C. 183
Eisenstein, M. 9091
Ekman, P. 650, 667
Elffers, E. 246247
Emirbayer, M. 263
Enfield, N.J. 319, 323, 334335, 503
Enkvist, N.E. 591
Erk, F. 359
Erman, B. 591, 601
Eschbach, A. 247
Evans, G. 320
Evans, N. 334
F
Fabbri, P. 187, 192, 195
Faber, D. 592
Name index
Faerch, C. 93, 591
Fairclough, N. 9, 26, 67, 110, 275276, 552
Fang, A.C. 592
Fanshel, D. 275276
Farncescotti, R.M. 413
Farroni, T. 579
Fauconnier, G. 15, 352, 398, 479483
Federmeier, K.D. 450
Feilke, H. 437, 446
Feng, G. 412, 414
Fetzer, A. 8, 2527, 3435, 40, 45, 5253,
56, 68, 99
Feyerabend, P. 224
Fiehler, R. 278
Fill, A. 593
Fillmore, C.J. 57, 68, 141, 316, 324, 326,
328329, 334335, 428, 450, 480, 596
Finegan, E. 9596, 592, 651
Firth, J.R. 2, 7, 144, 491, 606
Fisch, M.H. 181, 194
Fischer, K. 438, 599, 604, 610
Fitzmaurice, S.M. 98
Flader, D. 276
Fletcher, J. 602
Foolen, A. 595
Ford, C.E. 438, 440, 447, 450, 495496,
598
Fortescue, M. 328329, 333
Foucault, M. 151, 275, 308
Fox, B.A. 437438, 440, 450, 495496
Fox Tree, J.E. 660
Francis, G. 364
Francis, W.N. 591, 596
Franozo, E. 7778, 8182
Franks, J.J. 349
Franquiz, M. 672
Fraser, B. 277, 364
Fraurud, K. 357, 359
Freese, J. 499
Frege, G. 11, 68, 143, 148149, 206, 220,
309, 347, 379, 401402, 412, 461
French, P. 495
Freitas, S. 364
Fretheim, T. 364, 434, 599
Freud, S. 82
Friederici, A.D. 450
Friedman, M. 148, 159
Friedrich, P. 334
695
Fries, N. 437
Fries, U. 95
Friesen, W.V. 667
Fritz, G. 95, 275
Fronek, J. 248
Furiasso, C. 595
G
Gabelentz, G. von der 229, 240
Galasinski, D. 541, 552
Gale, R.M. 320
Garcs-Conejos, P. 87, 9293
Gardiner, A.H. 2, 7, 232, 234, 246247
Gardner, R. 438, 606
Garfinkel, H. 2, 7, 12, 26, 34, 141, 145,
149, 261269, 272274, 319, 323, 334
Garnham, A. 364, 591592
Garret, M. 411
Garrod, S.C. 359, 364, 446, 572
Garvey, C. 329, 334
Garvin, P.L. 237, 247
Gazdar, G. 68, 143, 406407, 411, 562,
566
Gee, J.P. 120
Geeraerts, D. 123
Geissner, H. 130
Geluykens, R. 592
Gergely, G. 579
Gernsbacher, M.A. 350, 355, 364
Geurts, B. 400, 411, 563, 567
Giannakidou, A. 400
Gibbon, D. 668
Gibbs, R.W. 16, 125
Giddens, A. 149
Gilliom, L.A. 364
Givn, T. 25, 28, 32, 38, 352353, 355,
364, 446
Glassner, B. 540, 542
Gloning, T. 275
Glucksberg, S. 16, 580
Goffman, E. 2, 7, 1213, 28, 34, 36, 53,
122, 144145, 153, 157, 269271, 273,
277, 390391, 499
Golato, A. 552
Goldberg, A. 428, 450
Goldberg, J. 498
Goldin-Meadow, S. 667
Goodman, N. 213214
696
Name index
Goodwin, C. 6, 34, 52, 143, 156, 268, 272,

319321, 323, 335, 391392, 438, 442,
499500, 502503
Goodwin, M.H. 438, 499500
Goossens, L. 121
Gordon, P.C. 364
Gotti, M. 124125
Goudge, T.A. 186
Graczyk, R. 326, 329, 333
Grady, J. 122, 124
Graesser, A.C. 561
Graf, C. 112
Graffi, G. 246, 248
Granger, S. 592593
Graumann, C.F. 247
Greatbatch, D. 118
Green, G.M. 69, 274
Green, J. 672
Greenbaum, S. 592, 596
Greenberg, R. 670
Grefenstette, G. 595
Greimas, A.J. 190, 192195
Grice, H.P. 2, 57, 11, 24, 29, 31, 3637,
4143, 61, 69, 89, 109, 114, 129130,
142, 145, 147, 153154, 157159, 170,
175176, 178180, 187188, 208,
217219, 223, 273, 319, 381383, 387,
407410, 412414, 416, 440, 448, 461,
467469, 471472, 474, 539, 561562,
566567, 578, 580
Gries, S.T. 449
Grimm-Vogel, I.H. 246
Grodner, D.J. 411, 565566, 569570
Grodzinsky, Y. 577
Grnqvist, L. 670
Grondelaers, S. 123
Grosz, B. J. 36, 364
Gruber, H. 547
Grnbaum, A.A. 236237, 239, 247
Grundy, P. 6869, 274
Guasti, M.T. 576
Guendouzi, J.A. 670
Gnthner, S. 278, 439, 450, 665
Gundel, J.K. 359, 364, 433434
Gunnarsson, M. 670
Gumperz, J.J. 2627, 34, 56, 117118, 120,
122, 131, 144, 156, 274275, 319, 334,
462, 492, 546, 665
Gustafsson, M. 591
Gut, U. 538, 551, 668
Guy, G. 602
H
Haan, F. de 593,
Haastrup, K. 542
Habel, C. 364
Haberland, H. 158
Habermas, J. 2, 7, 12, 31, 111, 116, 148,
171, 178, 246, 289310
Haegeman, L. 591
Hagemann, J. 107, 130
Hagen, K. 598
Hager, J.C. 667
Haiman, J. 444
Hajicov, E. 248
Halliday, M.A.K. 12, 25, 38, 52, 55, 114,
117, 130, 141, 147, 156, 190, 229, 232,
246, 317, 357, 427, 433434, 440,
491492, 538, 596, 603, 607, 609
Hancher, M. 155
Hanks, W.F. 6, 13, 26, 28, 36, 58, 68, 143,
316, 320, 323, 325327, 330, 333335,
463, 651
Hanna, J. 571
Haravon, A. 670
Harlow, L.L. 9293
Harnish, R.M. 221, 386, 411
Harris, R.A. 143, 147
Hart, H.L.A. 222
Hartford, B.S. 91, 552
Hasan, R. 52, 55, 114, 117, 130, 147, 156,
317, 357
Haslerud, V. 593
Haspelmath, M. 444
Hasselgrd, H. 364, 594
Hausendorf, H. 269, 275276
Have, P. ten 70, 268
Haverkate, H. 79
Haviland, J.B. 334335
Haviland, S.E. 364
Hawkins, J. 359, 364
Hay, J. 429
Hayashi, M. 436
Heath, C. 268, 272, 548
Heath, J. 325, 328, 330, 332334
Hedberg, N. 433
Name index
Hedstrm, K.E. 591
Heger, K. 247
Heim, I. 359, 363364, 407
Heine, B. 444, 590
Heller, D. 571
Hellermann, J. 496
Hellmann, C. 357
Hemingway, E. 518519
Hengeveld, K. 427
Henrichsen, P.J. 670
Henzler, S. 247
Hepburn, A. 504
Heritage, J. 26, 34, 145, 261, 264,
267269, 272, 278, 335, 526, 663
Herrmann, T. 247
Hervey, S.G.J. 176
Herzog, M.I. 240
Heusinger, K. von 353, 414
Heuvel, T. van den 592
Hill, C. 325, 334
Hilpinen, R. 179, 188, 190, 194195
Himmelmann, N.P. 334335
Hindelang, G. 275
Hintzman, D.L. 355
Hirschberg, J. 410411, 597
Hobbs, J.R. 364
Hrmann, H. 247
Hoey, M. 549, 632
Hoffmann, L. 428
Hoffmann, S. 630
Hofland, K. 591592, 595
Hoiting, N. 596
Holdcroft, D. 63
Holenstein, E. 149, 151, 248249
Holly, W. 12, 277, 430, 547
Holmes, J. 87, 551552, 592593
Holt, E. 496
Honneth, A. 303, 307308
Hood, S. 550
Hopper, P.J. 429430, 439, 444445, 447
Horn, L.R. 6, 2728, 3839, 69, 400, 405,
407408, 410414, 416, 464, 470471,
473, 562, 566
House, J. 93
Houser, N. 183, 194
Huang, Y. 3, 14, 2728, 32, 39, 43, 60,
6869, 140, 381, 397398, 400, 402,
405413, 470474, 580
697
Huang, Y.T. 569570

Hudson, R. 590, 637
Hbler, A. 10, 109, 111, 124, 147, 547
Hlzer, H. 246
Hnnemeyer, F. 444
Hundsnurscher, F. 275
Hundt, M. 695
Hunston, S. 549, 608, 633
Hurewitz, F. 411
Husserl, E. 148149, 158159, 233, 236,
247, 261, 291, 319, 394
Hutchby, I. 267268
Hutchins, E. 393
Hyland, K. 115
Hymes, D.H. 2, 7, 1112, 56, 144,
244245, 274275, 336, 546
I
Ide, S. 31
Imo, W. 278, 446, 450
Innis, R.E. 247
Irvine, J.T. 144
Israel, M. 398
J
Jackendoff, R. 350, 480
Jackson, D.D. 176
Jacobs, A. 9596, 100
Jaffe, A. 672
Jaggar, P. 325, 330, 334
Jakobson, R. 2, 7, 1113, 129130, 134,
141142, 146147, 149152, 156, 158,
170171, 229230, 232, 240, 242246,
248249, 309, 318, 334
Jannedy, S. 429
Janssen, T. 334
Jasperson, R. 499
Jaszczolt, K. 2729, 45
Jaworski, A. 110, 159
Jefferson, G. 70, 145, 265267, 274, 335,
379, 388, 391, 435436, 492, 498, 504,
531, 663
Jeffrey, R.C. 217
Jespersen, O. 13, 130, 141, 240, 244, 318,
334
Jewitt, C. 547
Johannessen, J.B. 598
Johansen, J.D. 181, 188, 195
698
Name index
Johansson, S. 443, 591592, 594, 596

Johnson, K. 85
Johnson, M.H. 121122, 178, 214
Johnson-Laird, P.N. 364, 366
Johnston, T. 549
Jonge, C.C. de 407
Jrgensen, A.M. 593
Joshi, A.K. 364
Joswick, H. 195
Juchem, J.G. 232, 246247
Jucker, A.H. 7879, 9597, 100, 599, 630,
646648, 651
K
Kallmeyer, W. 274
Kamide, Y. 568, 574
Kamio, A. 525
Kamp, H. 363
Kamp, R. 247
Kaplan, D. 11, 142, 207, 209, 223, 462, 466
Kappner, S. 176
Karkkinen, E. 446
Karttunen, L. 68, 143, 402, 406407
Kasper, G. 8890, 9293, 543, 551, 591
Katsos, N. 576577
Kay, P. 429, 450
Keenan, E.L. 89, 319320, 326, 328332,
334335, 463
Kehoe, A. 595
Keller, R. 214, 446
Kelly, J. 493, 495, 498
Kemmer, S. 429
Kendon, A. 321322, 598
Kennedy, C. 465
Kennedy, G. 587588
Kenwood, C. 661, 667
Keysar, B. 571
Kibble, R. 364
Kilgarriff, A. 595, 638639
King, J.C. 412, 462
Kingdon, R. 664
Kintsch, W. 70, 364
Kipp, M. 661
Kirk, J. 608
Kirschbaum, I. 214
Kissmann, U.T. 272
Kita, S. 335
Kitzinger, C. 504
Kjellmer, G. 606
Klein, W. 248
Klewitz, G. 496
Knees, M. 358
Knoblauch, H. 272
Knobloch, C. 246247
Knowles, G. 592593, 608
Koerner, E.F.K. 247
Kohnen, T. 97, 651
Kohrt, M. 275
Koktov, E. 533
Koons, R. 209
Kowal, S. 659, 662, 670671
Koyama, W. 10, 23, 143, 223
Krafft, U. 272
Kress, G. 442443
Kreuz, R.J. 18, 537, 547, 590, 662, 666,
672
Krifka, M. 434
Kripke, S. 142
Kronmuller, E. 573
Kroskrity, P.V. 159
Krzyzanowski, M. 551
Kubo, S. 3133
Khn, P. 277
Kuklick, H. 145
Kulick, D. 146
Kuno, S. 334
Kumpf, L.E. 431
Kurylowicz, J. 328, 335
Kussmaul, P. 591
Kutas, M. 450
Kutschera, F. von 217, 220
Kyburg, A. 465
Kyratzis, A. 658
Kyt, M. 97, 592
L
Labov, W. 240, 275276, 444, 544, 546
Ladd, D.R. 499
Ladusaw, W.A. 399, 658
Lakoff, G. 15, 121122, 143, 178, 214,
428, 478, 480
Lakoff, R.T. 38, 146
Lambrecht, K. 39, 364, 369, 431
Lampert, M.D. 597, 608
Langacker, R.W. 15, 53, 428429, 441,
447, 478479, 483
Name index
Langford, D. 608
Lasater, A. 324
Lasersohn, P.N. 465
Lasnik, H. 426
Lass, R. 444
Lauridsen, K.M. 592
Laval, V. 125
Lawrence, L. 592
Lazarus, M. 246
Laziczius, J. von 247
Leder, D. 108
Lee, D. 550
Leech, G.N. 9, 38, 7778, 80, 8286, 89,
92, 99100, 109, 114, 147, 158, 170171,
186188, 443, 538, 588, 597, 642, 661
Leeuwen, T. van 442443
Lehrer, A. 410
Leicher, S. 630
Lenk, U. 364, 443, 599, 651
Lenz, F. 358
Leonard, R. 334
LePage, R.B. 444
Lepore, E. 15, 412, 461, 466
Lerner, G. 436
Levelt, W.J.M. 668
Levenston, E. 542
Levinson, S.C. 6, 10, 23, 2729, 3132, 36,
38, 4043, 5657, 61, 6869, 79, 83, 89,
141, 143145, 153, 155159, 167170,
175, 186187, 194, 223, 242, 244, 247,
273277, 319, 334335, 383, 386,
388390, 392, 398, 400, 402403, 408,
410413, 416, 465, 469473, 475, 484,
538, 562567, 642
Lewis, D.K. 11, 23, 33, 36, 204, 209, 217,
220, 223, 481, 561
Liberman, A. 242
Liebal, K. 579
Liedtke, F. 436
Lieven, E. 448
Lindquist, H. 590
Lindsay, J. 671
Linell, P. 56, 493
Lipton, P. 223
Liszka, J.J. 194
Lizskowski, U. 578
Local, J.K. 493, 495498, 501, 504
Locke, J. 140, 149150, 154, 159
699
Loehr, D.P. 598

Lof, G.L. 658
Lohmann, J. 247
Loll, A. 354, 364
Loman, B. 591
Looks, K. 668
Lopes, J.G. 364
Louw, B. 539, 549, 608, 632633, 635
Louwerse, M.M. 667, 672
Luckmann, T. 56, 69, 149, 492
Lucy, J.A. 111, 116, 129130, 154, 156
Luff, P. 268, 272
Luhmann, N. 178, 269, 308
Luzio, A. di 26, 156
Lycan, W.G. 638
Lyons, J. 141, 242, 244, 320, 335, 358, 464,
642
M
MacDonald, M.C. 566
Macfarlane, S. 495
Mackenzie, J.L. 427
MacWhinney, B. 349, 429, 660, 665666,
669670
Maes, A. 364
Mahlberg, M. 551
Mahon, B. 349
Mair, C. 590
Makin, V.S. 572
Malinowski, B. 2, 7, 144145, 150, 232,
243, 246, 248, 545
Manelis Klein, H.E. 333
Manes, J. 646647
Mann, W. 55
Manning, P. 57
Manor, R. 205
Markman, A.B. 572
Marmaridou, S.S.A. 9, 25, 38, 45, 54, 57,
69, 79
Marshall, C. 334, 348
Marslen-Wilson, W. 349
Martens, E. 195
Martin, J.R. 548
Martinet, A. 149, 229, 319
Marty, A. 158159, 233, 247
Masaryk, T.G. 240
Mathesius, V. 229230, 239242, 245, 248
Matisoff, J. 329, 334
700
Name index
Matthiessen, C.M.I.M. 550, 552

Maturana, H.R. 127
Mauranen, A. 589
Mauro, T. de 594
Mautner, G. 547, 551552
Maynard, C. 630
Maynard, D.W. 263, 499
Mazeland, H. 496
McCarthy, M.J. 594, 606, 610, 634, 639
McCarthy, T. 289
McComish, J.F. 670
McDonald, J.L. 349
McEnery, T. 18, 587589, 594
McIntyre, D. 642645
McKee, C. 575
McKoon, G. 361
McNamara, T. 10, 9294
McNeill, D. 115, 127, 598, 668
McNeill, P. 537, 540, 545, 551
Meggle, G. 11, 217, 220, 223, 539
Meibauer, J. 274, 450
Meier, C. 665
Meijs, W. 592
Meillet, A. 158, 319
Merin, A. 412
Merlan, F. 334
Merleau-Ponty, M. 148, 152, 319
Metzing, C. 572573
Mey, J.L. 9, 15, 2425, 54, 6869, 7778,
80, 128129, 140141, 143144,
146149, 153, 155156, 158159, 168,
274, 526, 528, 532, 629
Meyer-Hermann, R. 112, 130
Mill, J.S. 407
Miller, J. 540, 542
Milroy, L. 596
Ming, T. 594
Mishler, E.G. 672
Mitterer, H. 503
Moeschler, J. 275
Mondana, L. 334
Moneglia, M. 593
Montague, R. 141142, 167, 207, 484
Monville-Burston, M. 248249
Moore, G.E. 158159, 379, 416, 528
Moravcsik, J.M. 212213
Morel, M.-A. 334
Morgan, J. 28, 32, 36, 43
Morley, J. 631633
Morpurgo Davies, A. 246
Morreau, M. 465
Morris, C.W. 1, 2, 7, 11, 13, 24, 79, 142,
148, 150, 158, 167169, 178, 183, 186,
190, 203204, 248, 316, 334
Mosegaard Hansen, M.-B. 587, 597,
599600, 610
Moxey, L.M. 364
Muck, B. 124
Mller, F. 496, 499500, 504
Mller, N. 662, 670
Mulder, J. 441
Mulligan, K. 150, 159
Murphy, G.L. 364, 477478
Murphy, J.P. 183
Musolff, A. 11, 247, 309, 428
Musolino, J. 575576
Myers, G. 661
N
Nadin, M. 184, 194
Nsslin, S. 591
Neidle, C. 661
Nekula, M. 248
Nerlich, B. 140141, 147148, 150152,
158159, 236, 246249
Nevala, M. 98
Nevalainen, T. 96, 592593
Newman, J. 651
Newmeyer, F.J. 426
Niedzielski, N.A. 120
Nth, W. 11, 168170, 172, 174, 176,
182185, 189190, 194195
Norrick, N. 604, 610, 651
Norris, S. 442, 660
Noveck, I.A. 411, 544, 563564, 575577,
580
Novick, D.G. 657
Nunberg, G. 68, 316
O
OConnell, D.C. 659, 662, 670671
OConnor, J.D. 491492, 664
OConnor, M.C. 450
ODonnell, M. 548
OHalloran, K. 549, 633
OSullivan, M. 650
Name index
Oakhill, J. 364
Obler, L.K. 670
Ochs, E. 278, 335, 450, 600, 660, 663, 667,
669
Oehler, K. 179, 194195
stman, J.-O. 591, 600
Ogden, R. 494495, 497, 503
Olsen, S.E. 80, 100
Olshtain, E. 89
Ono, R. 542543, 552
Ono, T. 450, 504
Orestrm, B. 591, 604
Ortner, H. 247248
P
Palen, L. 393
Paltridge, B. 553
Panizza, D. 411
Pankow, C. 194
Papafragou, A. 575576
Pape, H. 186, 188, 195
Paradis, C. 591
Parks, E. 516517
Parmentier, R.J. 177
Parrot, D.J. 672
Parret, H. 176, 184, 194
Parsons, T. 261, 265
Partee, B. 205, 207
Partington, A. 589, 631633
Paul, I. 108
Peirce, C.S. 1, 7, 11, 13, 23, 141142,
148150, 158159, 168169, 171174,
176177, 179186, 188195, 244, 263,
316317, 334, 382
Pekarek Doehler, S. 439
Penco, C. 36
Pepp, S. 495
Perrault, C.R. 387388
Perrott, M.A. 672
Perry, J. 213, 462
Peter, G. 466
Petrus, K. 566, 580
Pfnder, S. 429430
Phillips, B. 661, 667
Pickering, M.J. 446
Piercy, M. 522523
Pietarinen, A.-V. 179180, 182, 188, 190,
194
701
Pietroski, P.M. 400, 467

Pike, K.L. 144, 491
Pinker, S. 448
Plato 172, 233
Pollak, A. 547
Pollack, M.E. 28, 32, 36, 43
Pomerantz, A. 267
Portner, P.H. 205, 407
Posner, R. 167, 194
Potter, J. 70, 504
Potts, C. 413414
Pouscoulous, N. 563, 567, 576
Power, R. 364
Powers, E. 529
Powers, W.R. 666
Pressman, J.F. 142
Preston, D.R. 120, 672
Preyer, G. 466
Prieto, L.J. 176, 194
Prince, E.F. 364, 433, 439
Prucha, J. 9, 77, 80
Psathas, G. 267, 274, 663
Pschel, U. 277
Pullum, G.K. 658
Putnam, H. 142
Pye, C.A. 672
Q
Quasthoff, U. 665
Quinn, N. 131
Quirk, R. 114, 130, 591, 596, 664
R
Rahilly, J. 670
Ransdell, J. 182
Ratcliff, R. 361
Rauch, E. 277
Raumolin-Brunberg, H. 96
Rawls, J. 124
Raymond, G. 441
Raymond, W.D. 597
Reber, E. 499
Reboul, A. 411
Recanati, F. 27, 29, 3132, 155, 209, 214,
364, 411412, 464, 466, 484
Redder, A. 276, 428
Reddy, M. 29, 121, 124
Rees, M.A. van 390
702
Name index
Rehbein, J. 275, 278

Reichenbach, H. 320, 335
Reinhart, T. 364, 577578
Rellstab, D. 194
Renouf, A. 592, 595
Reppen, R. 589, 609
Richardson, A. 148
Rickheit, G. 364
Ricur, P. 157
Rieber, S. 413
Riley, P. 8486
Rim, B. 650
Rissanen, M. 592, 596
Risselada, R. 364
Roberts, C. 672
Roberts, F. 672
Robinson, J.D. 672
Robinson, M.A. 90
Roever, C. 10, 87, 9294
Romero, J. 603604
Rommetveit, R. 319
Rorty, R.M. 147
Rosaldo, M.Z. 89, 144, 159
Rose, K.R. 88, 90, 9293, 542543, 552
Routarinne, S. 495
Rubio-Fernandez, P. 580
Rudy, S. 243, 248
Rhlemann, C. 17, 26, 587, 590, 599, 634,
651
Ruesch, J. 156
Ruiter, J.P. de 503
Russell, B(enjamin) 400
Russell, B(ertrand) 143, 156, 309, 320, 335,
348, 405406, 461, 638, 640
S
Saba, W.S. 463
Sacks, H. 2, 7, 12, 70, 143, 265269,
272274, 316, 319, 335, 379, 388, 391,
433, 435, 492, 503, 531, 663
Sadock, J.M. 61, 69, 143, 159, 464
Saeed, J.I. 15, 29, 68, 223, 400, 416, 430
Sajavaara, K. 591
Sanders, T. 351, 364
Sandford, D. 588, 609
Sanford, A.J. 16, 359, 364
Sankoff, D. 601
Santaella, L. 169, 189, 194195
Sapir, E. 142, 151, 545

Sarangi, S. 146
Sarno, M.T. 670
Sauerland, U. 400, 411, 566
Saul, J. 416, 566, 580
Saunders, R. 333
Saussure, F. de 5, 126, 152, 154, 168, 174,
234236, 308
Savan, D. 185
Savigny, E. von 221222
Saville-Troike, M. 56
Sayette, M. A. 672
Sbis, M. 28, 3132, 187, 192193, 195,
373, 382384, 387
Schallenberger, S. 247
Schank, G. 277
Schank, R.C. 480
Scheerer, T.M. 247
Schegloff, E.A. 12, 70, 145, 261, 265269,
273274, 278, 319, 323, 330, 334335,
379, 388, 390391, 433, 435436, 440,
450, 492493, 496, 498, 500, 503504,
531, 547, 663, 668
Schelfhout, C. 593
Schenkein, J.N. 267268, 663
Schiaratura, L. 650
Schieffelin, B.B. 159
Schiffer, S. 217
Schiffrin, D. 244, 438, 658
Schlieben-Lange, B. 247
Schlobinski, P. 443, 665
Schmid, H.-J. 35, 130, 357, 439, 609
Schmitt, R. 272
Schmitz, U. 443
Schneider, S. 593
Schneider, W.L. 269
Schnurr, S. 551552
Scholes, R.E. 513, 519, 532
Scholz, O.R. 223
Schreuder, R. 322, 334, 348
Schubiger, M. 492
Schtz, A. 261, 264, 273
Schtze, F. 274
Schunn, C.D. 350
Schutz, A. 148, 319321, 335
Schwarz, F. 580
Schwarz, M. 350352, 354, 357359,
362364
Name index
Schwarz-Friesel, M. 13, 354, 358, 363, 637
Schwitalla, J. 274, 277278
Sclaroff, S. 661
Searle, J.R. 2, 5, 7, 11, 24, 2728, 3133,
3537, 3941, 69, 93, 107, 127128,
143145, 147148, 154, 157, 171,
175176, 180, 184192, 212, 220, 223,
236, 272, 294296, 316, 334, 349, 373,
381388, 390, 392, 432, 436, 461, 466,
539, 641, 646, 648, 658
Sedivy, J.C. 565567, 569570, 577
Sefi, S. 526,
Seidenberg, M.S. 566
Selinker, L. 88
Selting, M. 278, 435, 450, 495, 497, 504,
598, 602, 665
Semino, E. 121122, 548
Sengul, C.J. 349
Senft, G. 335
Shannon, C.E. 170, 242
Shastri, S.V. 592
Sherzer, J. 144, 334
Short, M. 548
Short, M.H. 642
Short, T.L. 179
Shriberg, L.D. 658
Sidnell, J. 334335
Sidner, C.L. 36
Sifianou, M. 87
Silverman, D. 537
Silverstein, M. 116, 142, 147, 153154,
156157, 159, 318, 323, 333334
Singer, M. 364
Simmons-Mackie, N.N. 669
Simon-Vandenbergen, A.-M. 122123,
594, 599
Simpson, G.B. 464
Sinclair, J.M. 274275, 388, 550551,
558589, 592, 598, 606609, 631,
639
Siren, K.A. 672
Skaffari, J. 549
Slobin, D.I. 669
Smith, B.C. 461
Smith, B.L. 123
Smith, H.L.J. 491
Smith, S. 599
Smits, R. 335
703
Snedeker, J. 569570
Snow, C. 669
Soames, S. 69
Sontag, S. 523524, 533
Sorjonen, M.-L. 438
Southgate, V. 579
Sperber, D. 6, 15, 27, 3234, 36, 4142, 68,
112, 129, 157, 168, 173, 175, 179, 351,
382, 410412, 473475, 477, 544, 564,
567, 580, 594, 599, 636
Spinks, C.W. 194
Spooren, W. 351, 364
Spranz-Fogasy, T. 278
Stalnaker, R.C. 11, 3133, 35, 6869, 223,
406407, 484
Stanley, J. 206, 210, 412, 462, 484
Steels, L. 450
Stefanowitsch, A. 449
Stenning, K.D. 349, 364
Stenstrm, A.-B. 588, 591593, 597598,
603, 607609, 644, 651
Steinbach, M. 450
Stelma, J.H. 672
Stetter, C. 174
Stiles, W.B. 630
Stivers, T. 335
Stocking, G.W. Jr. 151
Storm, T. 514
Strawson, P.F. 2, 5, 7, 11, 68, 143, 147, 153,
187, 309, 348, 382, 403, 406, 416, 461,
466, 638, 656
Streeck, J. 272, 278, 443
Stckl, H. 443
Stokoe, E. 540
Stubbe, M. 548
Stubbs, M. 549, 607, 609, 631, 633634
Sudnow, D. 262, 265, 267
Svartvik, J. 591592, 596, 664
Swanton, J. 326, 333
Sweetser, E.E. 444
Swiggers, P. 247
Szab, Z.G. 467, 484
Szczepek Reed, B. 495, 497, 500
T
Taavitsainen, I. 95, 100
Tabouret-Keller, A. 444
Tagliamonte, S. 590, 637
704
Name index
Tajfel, H. 118
Takahashi, S. 90
Takahashi, T. 89
Talmy, L. 478
Tanaka, H. 495
Tanenhaus, M.K. 565
Tannen, D. 144, 277, 544545, 664665
Tao, H. 325, 334
Tarplee, C. 494
Taylor, L. 592
Taylor, S. 537, 548, 551
Tenchini, M.P. 246
Tesnire, L. 194
Thavenius, C. 591
Thibaud, P. 195
Thibault, P. 56, 69, 549, 650651
Thies, A. 668
Thomas, J. 9, 69, 7778, 80, 82, 8587, 99,
155, 592, 661
Thomason R. H. 33
Thompson, G. 549
Thompson, S.A. 55, 278, 335, 435436,
438441, 446447, 450, 495
Thornton, R. 577
Thrane, T. 348
Tilley, S.A. 672
Titscher, S. 537, 545546, 548, 553
Tognini-Bonelli, E. 538, 588
Toman, J. 242243, 248
Tomasello, M. 212, 448450, 567,
578579
Toobin, J. 377
Tottie, G. 591, 593, 598, 601, 605
Tracy, K. 277
Trager, G.L. 491
Traugott, E.C. 39, 100, 124, 444445,
590
Travis, C. 212
Trippel, T. 668
Trnka, B. 248
Trosborg, A. 89
Trotha, T. von 276
Tsohatzidis, S. 69
Turner, K. 414
Turner, M. 479, 483
Turner, R.M. 262265, 268
Tynjanov, J. 243
Tyson, S. 593
U
Uhmann, S. 278, 434, 496, 665
Ungeheuer, G. 112, 232, 247
Urban, G. 153
V
Vachek, J. 248
Vanderveken, D. 3133, 436
Vanparys, J. 121
Venditti, J.J. 597
Ventola, E. 190, 275, 443
Verschueren, J. 25, 52, 108, 116, 120,
127128, 130, 147, 156, 158159, 588
Vicher, A. 601
Virtanen, T. 450, 587588
Volosinov, V.N. 157
Vonk, F. 247
Vonwiller, J. 602
Voormann, H. 538, 551
W
Walker, G. 496498, 504, 597
Walsh, S. 589590
Waltereit, R. 594
Walton, L. 657
Ward, G. 27, 39, 434
Ward I.C. 491492
Ward, K. 657
Wardhaugh, R. 546, 592
Warner, R. 566
Warnock, G.J. 147, 374, 381, 383
Wasow, T. 444
Watts, R. 31
Watzlawick, P. 130, 176
Waugh, L.R. 243, 248249
Weaver, W. 170, 242
Webber, B.L. 358
Weber, M. 150, 154, 223, 273, 289, 300
Wegener, P. 2, 7, 158, 230232, 234,
240243, 245247
Weil, H. 241
Weinback, L. 89
Weinreich, U. 240
Weissenborn, J. 247
Welker, K. 410
Wells, B. 495
Wells, W.H.G 495
Welte, W. 130
Name index
Werlen, I. 277
Wertz, J.M. 672
Westney, P. 592
Wetherell, M. 541
Wettersten, J. 247
Wexler, K. 577
White, P.R.R. 548
Whitsitt, S. 632
Wichmann, A. 496, 593, 597598,
602603
Wichmann, S. 333
Widdowson, H.G. 35, 85
Widmer, J. 269
Wierzbicka, A. 89, 100, 599
Wilcox, K.A. 672
Wilkes-Gibbs, D. 446
Wilkinson, S. 504
Williams, G. 146
Wilson, A. 587588, 592
Wilson, D. 6, 15, 27, 3234, 36, 4142, 68,
112, 129130, 157, 168, 173, 175, 179,
351, 382, 410412, 473475, 477479,
564, 567, 594, 599, 636
Wittgenstein, L. 2, 7, 11, 23, 100, 144, 147,
156, 172, 216, 222, 224, 248, 273,
293296, 374, 389, 461
Wodak, R. 67, 100, 275, 537, 547, 551
Wolf, K. 247
Wolfson, N. 89, 646647
Wood, J.L. 97
Woodfield, H.P. 542543, 551
Woods, D.K. 661
Woods, J. 415
Wooffitt, R. 118, 267
Woolard, K. 159
Woolf, V. 531
Wootton, A. 269270
Wright, G.H. von 154
Wright, M. 498
Wunderlich, D. 247
Wundt, W. 150, 233, 247
Wynne, M. 538, 551
X
Xiao, R. 594
Y
Yaeger-Droh, M. 500
Yang, Y. 325
Yngve, V. 604
Yuan, Y. 543, 546, 551552
Yule, G. 24, 100, 247, 364, 544
Z
Zacharsky, R. 433
Zelinsky-Wibbelt, C. 364
Zide, N. 325326, 329, 333334
Ziv, Y. 599, 651
705
706
Subject index
Subject index
A
anaphora 4, 7, 13, 51, 55, 156, 205, 238,
317318, 325, 327, 330333, 347,
353358, 363, 412, 428, 433, 440, 462,
599600, 603, 607, 640641, 644
indirect ~ 359360, 364, 640641
anaphoric see anaphora
appropriateness 910, 25, 27, 29, 77, 92,
94, 99100, 143144, 146, 268, 414
C
cataphora 111, 238, 317, 330, 640641,
644
cataphoric see cataphora
cognition 14, 185186, 291, 305, 307,
349350, 391, 393, 430, 474, 480, 566
cognitive
~ models 121, 123, 480
~ pragmatics 26, 34, 68, 566, 574
~ process 15, 43, 82, 347, 352, 355356,
358, 360361, 363, 431, 450,
479481, 483, 542, 566
~ processing see ~ process
~ science 4, 2334, 38, 44, 127, 350,
578, 588
~ semantics 15, 363, 478479
coherence 116, 351, 353, 356357, 363,
438, 651
common ground 8, 31, 3335, 38, 58, 69,
120, 178, 180, 184, 188189, 269, 320,
327, 330332, 348, 407, 415, 432, 589,
599660
compositionality 2728, 45, 206, 325326,
406
Construction Grammar 14, 278, 428, 430,
450
context
cognitive ~ 9, 25, 29, 3536, 39, 52, 54,
62, 68, 70
extralinguistic ~ see non-linguistic ~
linguistic ~ 9, 25, 2829, 3536, 56,
565566
nonlinguistic ~ 55, 59, 66, 206, 317,

324, 398, 408, 642
~ sensitive see ~ sensitivity
~ sensitivity 205, 207211, 215, 414,
449, 465467, 548, 629
social ~ 9, 2529, 32, 3536, 39, 52, 62,
70, 7980, 96, 99, 246, 278, 304, 306,
491, 552, 590, 609, 630, 660
Conversation Analysis 14, 16, 26, 65,
144145, 149, 153, 155, 265270,
272278, 319320, 390391, 430, 492,
531, 537, 547548, 587, 589, 597,
663
cooperative principle 24, 3738, 4143,
61, 83, 382, 387, 407408, 413, 468,
648
corpora see corpus
corpus 56, 1718, 95100, 359, 364, 429,
449, 538, 545552, 587599, 601603,
605610, 629631, 636640, 642, 644,
646651, 664665
~ based 6, 8, 17, 436, 538, 551,
587593, 595598, 600, 602606,
609610, 638, 648, 651
~ driven 538, 588590, 610
Corpus Analysis see Corpus Linguistics
Corpus Linguistics 8, 10, 1618, 56,
7778, 95, 538539, 547549, 552,
587590, 593, 595597, 610, 629,
630631, 633, 638, 646647, 649651
Corpus linguistic see Corpus Linguistics
Critical Discourse Analysis 9, 26, 6770,
9697, 100, 144, 146, 275, 548, 552
Critical Social Theory 12, 148, 290, 300,
303, 305, 307308
D
deictic center 142143, 642
deictic field 141142, 147, 237238, 247,
315316, 322323, 328
deictics 13, 58, 117, 315318, 320,
324333, 428, 435, 444, 642, 644645
Subject index
deixis 45, 7, 9, 13, 15, 18, 28, 31, 51,
5559, 6164, 68, 79, 95, 98, 141,
184, 186, 204205, 235, 238, 244,
246247, 316324, 326, 329, 331332,
354, 440, 462, 481, 638, 642, 644645,
651
discourse ~ 28, 36, 58, 315, 327,
358
participant ~ see person ~
person ~ 36, 315, 644
referential ~ 315, 319, 644
social ~ 36, 98, 315, 392
spatial ~ 186, 238, 315, 317, 327330
temporal ~ 36, 186, 315, 324, 327
Discourse Analysis 9, 2627, 120, 153,
155156, 274275, 373, 390, 547548,
551552, 587, 597, 607, 609
discourse marker 57, 28, 95, 119, 350,
438439, 443, 446, 448, 476, 526,
587, 590591, 594595, 598599,
601602, 604, 609610, 637, 651,
658660
E
entailment 4, 7, 14, 397401, 405, 408,
410, 413416, 470, 566
ethnography 8, 12, 16, 27, 55, 120, 144,
229, 263, 274275, 319, 322, 388,
390392, 537, 545546, 596
ethnographic(al) see ethnography
Ethnomethodology 2, 7, 12, 34, 145, 149,
261265, 269, 272, 274, 319
explicature 1415, 42, 397, 410412,
430, 475, 477
F
face 28, 31, 40, 94, 144, 270, 277, 543, 597,
603
falsity see truth
felicity 23, 357, 379
~ condition 8, 3233, 35, 4041, 62, 69,
143144, 172, 178, 383385, 390,
403, 576
frame 4, 14, 34, 36, 53, 9697, 111,
144, 153, 178, 271, 277, 322, 360,
362, 441442, 449, 480, 514, 544,
640
Frankfurt School 146, 290, 303, 307
707
function of language
conative ~ 146, 152153, 157, 171, 243,
248
emotive ~ 146, 152153, 157, 170171,
243
expressive ~ 170, 241, 243
metalinguistic ~ 146, 152153, 159,
244
phatic ~ 146, 152, 171, 244, 248
poetic ~ 146, 152153, 156, 159, 171,
243
referential ~ 146147, 152153, 155,
157158, 170, 243
I
illocution 9, 15, 32, 40, 51, 54, 57, 62,
77, 8385, 8889, 109, 148, 171, 177,
190194, 220222, 224, 296298, 349,
380381, 383387, 389390, 643, 648,
658
illocutionary act see illocution
illocutionary force see illocution
illocutionary force indicating devices
383385, 436, 646
implicature 4, 67, 9, 1416, 28, 3133,
37, 39, 4244, 51, 5564, 6769, 79,
94, 98, 139, 142, 145, 153, 157158,
179180, 188, 203205, 208, 319, 323,
382, 387, 397, 400, 404, 407416,
429430, 445, 467470, 473, 475476,
484, 561570, 574578, 580, 630
conventional ~ 14, 28, 37, 4243, 69,
397, 412415, 469
conversational ~ 14, 28, 37, 4243, 69,
179, 208, 382, 387, 397, 404,
407416, 445, 467470, 473, 580, 630
default ~ 42, 563564, 566567
generalized ~ 562, 564, 566, 569
scalar ~ 39, 400, 411, 562564, 567,
570
impliciture 28, 397, 411412, 430
indexical see indexicality
indexicality 45, 7, 13, 3133, 35, 5556,
116117, 119, 139, 141143, 146147,
159, 167, 183186, 207, 209210,
223, 238, 244245, 248249, 264265,
276, 315319, 320321, 323, 326328,
331333, 462, 466467, 600
708
Subject index
inference 45, 7, 1415, 2728, 3233,

35, 4144, 56, 59, 61, 63, 98, 223, 231,
266, 278, 319, 351352, 359, 361363,
397398, 401, 403, 411, 415, 425, 445,
448, 462, 465, 470, 474, 477, 484, 500,
565, 576, 579, 633, 638, 640, 646,
648
pragmatic ~ 51, 203, 205, 208215, 387,
412, 416, 469, 473, 475476, 563,
566567, 569570, 573575, 603, 636
intention 911, 13, 28, 3133, 3639, 41,
43, 44, 51, 54, 62, 6467, 69, 7879, 86,
95, 108112, 116, 124125, 144155,
169, 171, 173177, 180182, 187189,
193, 206, 209211, 213, 215223, 236,
271, 289291, 293, 295297, 299, 302,
307309, 322, 347, 349, 377, 379380,
382388, 392, 427, 445446, 448, 462,
467, 475, 526527, 567, 570, 572575,
578580, 594, 596597, 605
intentionality see intention
introspection 5, 8, 1617, 192, 461, 491,
539, 563, 596, 631
K
knowledge
metapragmatic ~ 119120, 122, 125,
432
L
language acquisition 87, 90, 92, 212, 236,
373, 448, 542, 544
lifeworld 291, 295, 298, 300302, 304,
306
locution 62, 296, 298, 380, 383, 390
locutionary see locution
Logical Positivism 159, 167, 374
M
maxim of conversation
exploiting ~ 4243, 208209, 409
flouting ~ 43, 114, 208, 409
manner ~ 37, 109, 114, 408, 468
quality ~ 37, 109, 114, 409410, 468
quantity ~ 37, 109, 114, 440, 468,
471472
relation ~ 37, 109, 114, 468
~ of relevance see relation ~
meaning
context-dependent ~ 24, 26, 34, 462
conventional ~ 41, 186188, 208,
219220, 222, 414, 467, 479, 570
hearer ~ 187188
intended ~ 2425, 33, 39, 176, 178, 187,
571
intersubjective ~ 219, 291292
lexical ~ 27, 352, 357, 477, 479
pragmatic ~ 9, 31, 3738, 77, 96, 187,
426, 437, 549, 552, 630631
sentence ~ 2, 170, 187, 222, 382, 385,
461, 484
speaker ~ 170, 187188, 411, 416, 561,
578, 635
subjective ~ 219, 223, 290291, 293,
309
utterance ~ 3, 27, 29, 31, 81, 183, 187,
322, 385
metacommunication 107109, 111113,
116, 119, 129130, 156157
metalanguage 108, 116, 129130, 140,
156, 171
metaphor 1516, 121124, 131, 213214,
232, 319, 321, 331, 385, 429, 444445,
478480, 483, 580, 609610
metapragmatics 10, 107108, 112, 117,
119, 125, 129131, 139, 147, 156
multimodal 1415, 18, 442, 449450,
502503, 543, 547549, 610, 629,
650651
multimodality see multimodal
N
narration 513514, 517519, 522, 530, 532
O
Ordinary Language Philosophy 27, 39,
153155, 159, 272
Organon model 141, 170, 232233, 235,
237238
origo of deixis see deictic center
P
participation framework 13, 271, 321
performative 25, 31, 3940, 55, 124, 139,
146, 148, 155159, 194, 294, 296,
375380, 390, 413, 436, 651
Subject index
~ verb 112, 120, 297, 377, 384
performativity see performative
perlocution 9, 23, 27, 40, 62, 127, 143, 148,
154, 177, 192193, 220222, 224,
298381, 387, 392
perlocutionary see perlocution
politeness 15, 27, 31, 38, 79, 81, 83, 8689,
9293, 95, 100, 109, 114, 157, 191, 277,
386, 441, 446, 538, 547, 597
polysemy 211214, 464
Positivism see Logical Positivism
pragmatic
~ competence 85, 89, 561, 566567,
574576, 578579
~ development 78, 84, 9091, 95, 99,
578
~ enrichment 27, 464, 475, 477
~ failure 43, 7778, 8488, 9193, 95,
99
~ intrusion 205, 412, 465, 473
~ transfer 7778, 84, 8792, 95, 99
pragmatics
development of ~ 12, 461, 561
empirical ~ 26, 292293, 552, 587588
formal ~ 12, 2627, 141, 292, 294,
299300, 302303, 305308
functional ~ 275, 278, 428, 548
general ~ 8, 25, 31, 38, 44, 8284, 89
historical ~ 17, 26, 7778, 84, 95,
98100
inferential ~ 467, 473
interlanguage ~ 26, 8890
literary ~ 16, 511
micro ~ and macro ~ 7, 9, 51, 5357,
6061, 63, 6668, 70
universal ~ 12, 292
presupposition 45, 79, 1516, 28, 31, 35,
51, 5556, 6065, 69, 79, 117, 119, 130,
139, 143, 153, 174, 192, 204, 297, 299,
304307, 397, 401405, 413416, 427,
465, 470, 481, 544, 561, 565, 580, 609,
638
pragmatic ~ 14, 3233, 57, 62, 397,
406407
~ trigger 32, 59, 402, 406
proposition 29, 3233, 35, 37, 40, 51,
55, 68, 112113, 147148, 150, 158,
172, 179, 188192, 203, 205206, 215,
709
296297, 304305, 349, 351352,

358, 379380, 383385, 397399,
401, 403407, 410, 413, 430431,
461467, 469, 473, 475476, 484,
567, 577, 589, 599, 601, 604, 607,
635636, 642643
propositional see proposition
prosodic see prosody
prosody 6, 15, 116118, 430, 438, 449,
491495, 497503, 539, 548549, 590,
592, 598, 600603, 608609, 629,
631635, 650, 664, 670671
semantic ~ 18, 539, 549, 608609, 629,
631635
R
reference 7, 11, 13, 28, 37, 40, 5859, 98,
109, 112, 117, 142, 148, 158159,
183185, 206, 315, 317318, 320,
322323, 327328, 330331, 333,
347350, 357, 359, 363364, 380,
442444, 462, 465, 481, 483, 501, 512,
519520, 544, 568569, 572, 637638,
640646
co-~ 317, 355356, 432, 447, 577578
~ resolution 355, 475, 646
referential see referentiality
referentiality 155, 127, 155, 157, 269,
350
reflexive (language) see reflexivity
reflexivity 111, 116, 130, 139, 155, 157,
217218, 264265, 267, 307, 320
relevance
conditional ~ 35, 267
principle of ~ 129, 179, 351, 474,
594
Relevance Theory 15, 33, 4142, 129, 157,
410413, 473, 475, 477479, 484, 567,
599, 636
rheme 39, 157, 229, 241, 248, 380,
433434, 603
rule
constitutive ~ 35, 39, 107
regulative ~ 35, 39
S
second language acquisition see language
acquisition
710
Subject index
semantic
~ preference 608609, 631, 635
~ prosody 18, 539, 549, 608609, 629,
631635
shifter 13, 116117, 129130, 141,
244246, 249, 316, 318, 327
sincerity condition 40, 125, 384385
stance 15, 320, 425, 437438, 442, 463,
493, 499500, 503, 517, 600, 651
T
theme 39, 157, 229, 241, 433434, 603
Theory of Communicative Action 178,
246, 290, 295, 298, 302303, 308309
transcription 6, 8, 1718, 391, 503,
590591, 597598, 601602, 629
systems of ~ 537, 547, 657672
truth
~ conditional see ~ conditions
~ conditions 29, 167, 172, 203,
205208, 211, 215, 348, 358, 374,
384, 401, 412, 414415, 466467,
469, 475, 479480, 484, 574
~ value 32, 207, 348, 358, 374376, 378,

398, 406, 466, 575, 638
turn 85, 94, 110, 113, 115, 117, 268, 430,
436, 438, 440442, 446, 494, 498501,
504, 517, 523, 605, 608, 650, 658660,
665, 668
~ constructional see ~ construction
~ construction 15, 435436, 441, 493,
496, 503504
~ continuation 436, 493, 497
~ -taking 15, 35, 66, 122, 266267, 272,
321, 392, 425, 435, 495496, 503,
531, 543, 548, 587, 591592, 597,
602604
U
underspecification 15, 41, 350351, 364,
464465, 477, 484
underspecified see underspecification
Z
Zeigfeld see deictic field
HoPs 1
Handbooks of Pragmatics
Editors
Wolfram Bublitz
Andreas H. Jucker
Klaus P. Schneider
Volume 1
De Gruyter Mouton
Edited by
Wolfram Bublitz
Neal R. Norrick
De Gruyter Mouton
ISBN 978-3-11-021425-3
e-ISBN 978-3-11-021426-0
Library of Congress Cataloging-in-Publication Data
Foundations of pragmatics / edited by Wolfram Bublitz, Neal R.
Norrick.
p. cm. (Handbook of pragmatics; 1)
Includes bibliographical references and index.
ISBN 978-3-11-021425-3 (alk. paper)
1. Pragmatics. I. Bublitz, Wolfram. II. Norrick, Neal R.
P99.4.P72F68 2011
4011.45dc22
2011013980
Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie;
detailed bibliographic data are available in the Internet at http://dnb.d-nb.de.
2011 Walter de Gruyter GmbH & Co. KG, Berlin/Boston
Cover image: Vetta Collection/iStockphoto
Typesetting: Drlemann Satz GmbH & Co. KG, Lemfrde
Printing: Hubert & Co. GmbH & Co. KG, Gttingen
Printed on acid-free paper
Printed in Germany
www.degruyter.com

The series Handbooks of Pragmatics, which comprises nine self-contained volumes, provides a comprehensive overview of the entire field of pragmatics. It is
meant to reflect the substantial and wide-ranging significance of pragmatics as a
genuinely multi- and transdisciplinary field for nearly all areas of language description, and also to account for its remarkable and continuously rising popularity
in linguistics and adjoining disciplines.
All nine handbooks share the same wide understanding of pragmatics as the
scientific study of all aspects of linguistic behaviour. Its purview includes patterns
of linguistic actions, language functions, types of inferences, principles of communication, frames of knowledge, attitude and belief, as well as organisational
principles of text and discourse. Pragmatics deals with meaning-in-context, which
for analytical purposes can be viewed from different perspectives (that of the
speaker, the recipient, the analyst, etc.). It bridges the gap between the system side
of language and the use side, and relates both of them at the same time. Unlike syntax, semantics, sociolinguistics and other linguistic disciplines, pragmatics is defined by its point of view more than by its objects of investigation. The former precedes (actually creates) the latter. Researchers in pragmatics work in all areas of
linguistics (and beyond), but from a distinctive perspective that makes their work
pragmatic and leads to new findings and to reinterpretations of old findings. The
focal point of pragmatics (from the Greek prgma ,act) is linguistic action (and
inter-action): it is the hub around which all accounts in these handbooks revolve.
Despite its roots in philosophy, classical rhetorical tradition and stylistics, pragmatics is a relatively recent discipline within linguistics. C.S. Peirce and C. Morris
introduced pragmatics into semiotics early in the twentieth century. But it was not
until the late 1960s and early 1970s that linguists took note of the term and began
referring to performance phenomena and, subsequently, to ideas developed and advanced by Wittgenstein, Ryle, Austin and other ordinary language philosophers.
Since the ensuing pragmatic turn, pragmatics has developed more rapidly and diversely than any other linguistic discipline.
The series is characertised by two general objectives. Firstly, it sets out to reflect the field by presenting in-depth articles covering the central and multifarious
theories and methodological approaches as well as core concepts and topics characteristic of pragmatics as the analysis of language use in social contexts. All articles are both state of the art reviews and critical evaluations of their topic in the
light of recent developments. Secondly, while we accept its extraordinary complexity and diversity (which we consider a decided asset), we suggest a definite
structure, which gives coherence to the entire field of pragmatics and provides
vi
orientation to the user of these handbooks. The series specifically pursues the following aims:
it operates with a wide conception of pragmatics, dealing with approaches that
are traditional and contemporary, linguistic and philosophical, social and cultural, text- and context-based, as well as diachronic and synchronic;
it views pragmatics from both theoretical and applied perspectives;
it reflects the state of the art in a comprehensive and coherent way, providing a
systematic overview of past, present and possible future developments;
it describes theoretical paradigms, methodological accounts and a large
number and variety of topical areas comprehensively yet concisely;
it is organised in a principled fashion reflecting our understanding of the structure of the field, with entries appearing in conceptually related groups;
it serves as a comprehensive, reliable, authoritative guide to the central issues
in pragmatics;
it is internationally oriented, meeting the needs of the international pragmatic
community;
it is interdisciplinary, including pragmatically relevant entries from adjacent
fields such as philosophy, anthropology and sociology, neuroscience and psychology, semantics, grammar and discourse analysis;
it provides reliable orientational overviews useful both to students and more
advanced scholars and teachers.
The nine volumes are arranged according to the following principles. The first
three volumes are dedicated to the foundations of pragmatics with a focus on micro
and macro units: Foundations must be at the beginning (volume 1), followed by
the core concepts in pragmatics, speech actions (micro level in volume 2) and discourse (macro level in volume 3). The following three volumes provide cognitive
(volume 4), societal (volume 5) and interactional (volume 6) perspectives. The
remaining three volumes discuss variability from a cultural and contrastive (volume 7), a diachronic (volume 8) and a medial perspective (volume 9):
1. Foundations of pragmatics
2. Pragmatics of speech actions
Marina Sbis and Ken Turner
3. Pragmatics of discourse
Klaus P. Schneider and Anne Barron
4. Cognitive pragmatics
Hans-Jrg Schmid
5. Pragmatics of society
Gisle Andersen and Karin Aijmer
6. Interpersonal pragmatics
Miriam A. Locher and Sage L. Graham
7. Pragmatics across languages and cultures
Anna Trosborg
8. Historical pragmatics
Andreas H. Jucker and Irma Taavitsainen
9. Pragmatics of computer-mediated communication
Susan Herring, Dieter Stein and Tuija Virtanen
vii
Acknowledgements
This initial, foundational volume in this series of handbooks represents the work of
a large number of individuals over a fairly long period of time. We must, of course,
thank our contributors for producing high quality articles reflecting scholarly rigor
and a strong sense of what our readers seek in a handbook of this scope. Without
conscientious, willing reviewers, projects like this are simply impossible. Our sincere thanks go out to all our reviewers, though we cannot identify them by name.
We editors are surrounded by competent, trustworthy staff in Augsburg and Saarbrcken, Germany. We own a debt of gratitude to Claudia Enzweiler for proofreading and Sylvia Monzon for her clerical services, to Katharina Rters, Katrin Stuis
and Claudia Rieger for their help with the desk editing and compilation of the indexes. Finally, we are grateful to the De Gruyter team: Barbara Karlson, Wolfgang
Konwitschny and Anke Beck.
Table of contents
Preface to the handbook series . . . . . . . . . . . . . . . . . . . . . . . .

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
v
ix

Wolfram Bublitz and Neal R. Norrick . . . . . . . . . . . . . . . . . . . . .

1.
2.
3.
4.

Anita Fetzer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
23

Piotr Cap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
51

Sophia Marmaridou . . . . . . . . . . . . . . . . . . . . . . . . . .
77
Metapragmatics
Axel Hbler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

5.

Wataru Koyama . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
6.

Winfried Nth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
7.

Nikola Kompa and Georg Meggle . . . . . . . . . . . . . . . . . . . 203
8.

Saskia Daalder and Andreas Musolff . . . . . . . . . . . . . . . . . . 229
9.

Christine Domke and Werner Holly . . . . . . . . . . . . . . . . . . 261
xii
Table of contents
10.

Maeve Cooke . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
289

11.
12.
13.
14.

William F. Hanks . . . . . . . . . . . . . . . . . . . . . . . . . . . .
315

Monika Schwarz-Friesel and Manfred Consten . . . . . . . . . . . .
347
Speech acts
Elena Collavin . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
373

Yan Huang . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
397

15.
16.
17.
18.

Arnulf Deppermann . . . . . . . . . . . . . . . . . . . . . . . . . .
425

John Saeed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
461

Elizabeth Couper-Kuhlen . . . . . . . . . . . . . . . . . . . . . . .
491

Jacob L. Mey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
511

19.
20.

Monika Bednarek . . . . . . . . . . . . . . . . . . . . . . . . . . .
537
Richard Breheny . . . . . . . . . . . . . . . . . . . . . . . . . . . .
561
Table of contents
xiii
21.

Gisle Andersen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
22.

Christoph Rhlemann . . . . . . . . . . . . . . . . . . . . . . . . . . 629
23.

Roger J. Kreuz and Monica A. Riordan . . . . . . . . . . . . . . . . 657
About the authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 681

Name index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 691
Subject index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 706
xiv
Table of contents

1.
The aim of the handbook
The present volume, as the initial handbook in the nine volume series Handbooks
of Pragmatics, provides a comprehensive overview of the foundations of pragmatics. Its in-depth articles cover the roots and evolution of those central theories
and approaches as well as key concepts and topics that are characteristic of twenty-first century pragmatics as an approach to the means and ways of using language
in authentic social contexts. The articles provide reliable orientational overviews
useful to researchers, students, and teachers. They offer both state-of-the-art reviews of their topics and critical evaluations in the light of on-going developments.
Thus, topics are considered not only within their contemporary scholarly context
but are also critically evaluated from one or more current perspectives. As the
opening volume in the series, Foundations of Pragmatics provides historical, conceptual, theoretical and methodological vantage points from which the articles in
the following eight volumes can be related to each other and to the development of
the entire field.
The series as a whole seeks to be reasonably comprehensive to account for the
exceptionally vast, unusually heterogeneous and still rapidly expanding field of
pragmatics. At the same time, it sets out to give structure and coherence to the field
through its organization and choice of topics as well as by revealing traits and contours that have evolved during the past half-century. The nine volumes are meant to
reflect the substantial and wide-ranging significance of this transdisciplinary subject for nearly all areas of language description, to trace its origins and influences,
as well as to account for its remarkable and continuously rising popularity in linguistics and adjoining disciplines.
2.
The history of pragmatics
Even though its roots can be traced back to early classical traditions of rhetoric and
stylistics, to Immanuel Kants conception of pragmatics as empirical and purposive and to William James, who pointed out its practical nature, modern pragmatics
is a fairly recent discipline. Its inauguration as an independent field of study within
semiotics took place early in the 20th century by C. Morris, R. Carnap and ultimately C.S. Peirce. The classic division between syntax, semantics, and pragmatics goes back to Morris, who distinguished three separate dimensions of semiosis within his science of signs:

Syntax studies the relations signs bear to other signs, semantics the relation between signs and objects, and pragmatics the relation between signs and their interpreters. Of course, there were and are differences of opinion on where exactly to
draw the line between semantics and pragmatics, as many of the chapters in this
volume will show.
Some thirty years elapsed before pragmatics finally made its way into modern
linguistics in the late 1960s, when linguists began to explore so-called performance phenomena. To this end, they adopted ideas developed and advanced by
L. Wittgenstein, G. Ryle, P. Strawson, J.L. Austin and other eminent (ordinary or
natural) language philosophers. It seems safe to claim that the ensuing pragmatic
turn was most notably induced by J.L. Austin, J.R. Searle and H.P. Grice, who
were interested in utterance meaning rather than sentence or word meaning, i.e. in
studying unique historical events created by actual speakers to perform linguistic
acts in actual situational contexts in order to accomplish specific goals. Other
scientific movements that nourished pragmatics include anthropology (B. Malinowski, P. Wegener, A. Gardiner), contextualism (J.R. Firth) and functionalism
(K. Bhler, R. Jakobson, D. Hymes), ethnomethodology (H. Garfinkel, E. Goffman, H. Sacks) and European sociology (J. Habermas). This volume addresses all
of these influences.
Since the pragmatic turn, pragmatics has arguably developed more rapidly and
diversely than any other linguistic discipline. Since the 1970s, the early AngloAmerican framework of pragmatic linguistic study has been immensely expanded
and enhanced by research in Continental Europe and elsewhere. With historiographic hindsight, it can be seen that the broadening, i.e. the interdisciplinary expansion, of the field of pragmatics has been a cumulative process; the broader conception of pragmatics chronologically (and causally) followed the narrower one.
We can easily detect a first shift from the binarity of early structuralist concepts
(such as lexical/word meaning versus causal/sentence meaning) to the multiplicity
of speech act related concepts. A further step towards a (conceptual as well as methodological) widening of the field took place in the 1970s and 1980s, when structure and action oriented pragmatics linked up with the emerging interactional paradigms in sociology (in general) and ethnomethodology (in particular). The ensuing
developments have seen inter alia a realignment of context (from a static and au-
tonomous to a dynamic and collaborative concept, which is cognitively, situationally and socio-culturally much more refined), and a shift from the concept of the
unilateral, i.e. the individual speakers (speech) act (as advocated by pioneers like
Austin) to the interactionally expanded concept of the bi- or (in some types of interactive computer-mediated forms of communication) multi-lateral inter-act. As
such, contemporary broad pragmatics takes account of the interactional turn
that can be observed in the most recent development of interactive (Web 2.0-based)
media formats.
Taken all in all, the establishment and development of linguistic pragmatics has
been an authentic success story.
3.
The definition of pragmatics
Despite its scientific acclaim, the notion of pragmatics remains somewhat enigmatic and is still difficult to define. This holds for its readings in everyday discourse as well as in scholarly contexts. Nonetheless, when we refer to attitudes and
modes of behavior as pragmatic, we mean that they have a factual kind of orientation in common. People who act pragmatically or take a pragmatic perspective
generally have a preference for a practical, matter of fact and realistic rather than a
theoretical, speculative and idealistic way of approaching imminent problems and
handling everyday affairs. To put it differently, they share a concrete, situation-dependent approach geared to action and usage rather than an abstract, situation-independent and system-related point of view. To assume a pragmatic stance in
everyday social encounters as well as in political, historical and related kinds of
discourse, means to handle the related affairs in a goal-directed and object-directed, common-sense and down to earth kind of way. Such an understanding of
pragmatics as an attitude in non-scientific discourse has obviously left its traces in
scientific definitions of the term. By and large we can say that in semiotics and
philosophy, pragmatic characterizes those theoretical and methodological approaches that are oriented toward use and context rather than toward some system,
and that they regard use and context as creating a high degree of analytical surplus.
While essentially the same is true for linguistics in general, there is no commonly accepted definition of pragmatics in linguistics which would refer to a
single, unified and homogeneous field of study. In contemporary linguistics, we
can identify a narrow and a broad way of delineating pragmatics (of which the
former is sometimes allocated to an Anglo-American and the latter to a Continental [European] tradition of pragmatics, cf. Huang 2007: xi). According to the
narrow view, pragmatics is understood as the systematic investigation of what and
how people mean when they use language as a vehicle of action in a particular context and with a particular goal in mind. Thus, the context-dependency of utterance
meaning is the central component of more narrowly defined accounts of prag-
matics, which focus on a few key issues that can be juxtaposed with related issues
in other modules of language theory such as grammar and semantics. Those issues
include indexicality/deixis (versus anaphora), presuppositions, implicatures (versus entailments) and speech acts (versus types of sentences), to name only the most
conspicuous topics.
In this volume, as indeed in the whole Handbooks of Pragmatics series, we
adopt a much broader point of view and understand pragmatics as the scientific
study of all aspects of linguistic behavior. In particular,
pragmatics includes patterns of linguistic actions, language functions, types of inferences, principles of communication, frames of knowledge, attitude and belief, as well as
organisational principles of text and discourse. Pragmatics deals with meaning-in-context, which for analytical purposes can be viewed from different perspectives (that of
the speaker, the recipient, the analyst, etc.). It bridges the gap between the system side of
language and the use side, and relates both of them at the same time. Unlike syntax, semantics, sociolinguistics and other linguistic disciplines, pragmatics is defined by its
point of view more than by its objects of investigation. The former precedes (actually
creates) the latter. Researchers in pragmatics work in all areas of linguistics (and
beyond), but from a distinctive perspective that makes their work pragmatic and leads
to new findings and to reinterpretations of old findings. The focal point of pragmatics
(from the Greek prgma act) is linguistic action (and inter-action); it is the hub around
which all accounts in these handbooks revolve. (Preface to the handbook series)
Pragmatics is fundamentally concerned with communicative action in any kind of

context. The multifaceted research paradigm of pragmatics has provided new directions and perspectives in the arts and humanities, philosophy, cognitive science,
computer science and the social sciences. Pragmatic perspectives have been employed in information technology and in the social sciences, particularly in economics, politics and education.
In the pragmatic perspective, language use and language users in interaction
are primary, as opposed to language as a system of signs or a set of rules. The pragmatic perspective scrutinizes neither just individual words nor sentences nor even
isolated texts, but rather whole speech events or language games in real social contexts, considering both the present state of affairs and its connectedness with prior
and succeeding actions. It rejects a localization of language in a limited segment of
the acts of speaking, understanding and responding or within the conscious of the
individual. It supplants a view of language as an abstraction without variation by
speaker, region or time, of language as a non-cultural, non-social, static, depersonalized fact independent of context and discourse. Pragmatics goes beyond the perspective of written texts with their carefully marshalled grammatical sentences to
embrace everyday talk and the messiness of language in real embodied human
contexts, where participants with personalities, feelings and goals interact in complex ways with physical objects and other participants within institutions and communities.
That pragmatics overcomes any narrow focus on language as a system of signs

or a set of rules does not mean, of course, that pragmatics (in contradistinction to
other scientific disciplines) attempts to describe language without recourse to the
systematic level of theoretical abstraction. Ultimately and in order to create and secure a sound scientific foundation, pragmatics, like any other scientific theory, has
to accomplish the transfer from the level of token to the level of type. For the pragmatic scholar, specific concrete linguistic events (and their contexts) are only relevant in as far as their properties and constraints can be integrated under analogous
conditions into a general concept of language and communication. While pragmatics is on a par with other scientific theories in this respect, it transcends them in
various ways, in particular by adopting a constructivist point of view which allows
for the inclusion of new contextual, situational and cognitive variables. Describing
language pragmatically thus means going beyond the description of language as an
autonomous, type-restricted principle and taking into account extra-linguistic phenomena and conditions emanating from the context and concrete situation of language use. Turning contextual and related conditions into prototypical conditions
(and thus theorizing them) constitutes the pragmatic surplus.
4.
The purview of pragmatics
If pragmatics is more a perspective or orientation toward language use than a particular theory or specific research object, then the data practitioners choose to analyze assume great importance. In linguistic research, as Saussure realized, it is the
viewpoint that determines the object rather than the other way around. In the tradition growing out of Natural Language Philosophy the intuitions and casual observations of the individual writer provide the stuff for analysis. Introspection counts
as data. All the early work on speech acts, presupposition and inference was based
on introspection and invented examples in imagined contexts (Austin, Searle,
Strawson and Grice).
Other traditional areas of pragmatic research were based on authentic data from
the start. Indexicality/deixis, discourse markers, particles and the like are ubiquitous in texts, and examples for study are thus easy to collect. As corpora of spoken
language became increasingly available and reliable, the problem of evidence versus introspection began to take care of itself. Scholars no longer needed to guess
about distributions and frequencies of linguistic items. One might begin with a
small pre-selected corpus representing a specific set of items or types of interaction
to identify a range of functions for a particular linguistic feature before embarking
on a general search in a large corpus to determine distributions and frequencies in
the corpus at large only to return to a small corpus representing specific contexts
for careful qualitative analysis. By contrast, investigations of large corpora might
provide the impetus for research, illustrating a range of items or contexts not found
in a small corpus and suggesting hypotheses to be tested. For statistical purposes,

larger was presumably always better, because scholars tended to correlate frequency with typicality (the more frequent the more typical). On the other hand, increasing interest in multi-modal data and description in research on spoken interaction brought researchers back to small, carefully annotated corpora. By its very
nature, research on prosody, gaze, gesture and the details of interaction with objects in a specific physical setting must proceed from a narrow recording and transcript of a particular event: Large corpora are extremely difficult to manage in
these circumstances.
Many pragmaticians maintain that language data must be as authentic as possible, recorded in real life contexts where something is at stake for the participants.
But the whole matter of recording, often with camcorders as well as microphones,
raises the spectre of the Observers Paradox. How natural is the language interaction, when we require technical apparatus to record it? Some pragmatic scholars
ascribe to the notion that language use must be understood in terms of embedded
practices, as part of the social world in which listening and speaking are modalities
of action, in line with the treatment of speech production and reception as embodied social facts (Hanks 1996), and practices as shared ways of doing things, ways
of talking, beliefs and values as a function of engagement in an activity (Wenger
1998; Eckert 2000). Conceptualizing speaking and listening in terms of practices
affords a natural purchase on nonverbal features of participant behavior, including
gaze, posture and gesture in face-to-face interaction (Goodwin 2003). This has
consequences for what counts as data, how much context must be considered, what
must be transcribed and in how much detail.
At the same time, certain questions will always go unanswered on the basis of
corpus data, no matter how careful the transcription and no matter how large the
corpus, since every individual exchange is participant designed and context dependent. With some kinds of structures like prosody, discourse markers and interjections, there is the problem of how to get enough examples to make sensible generalizations. Experimental procedures could complement corpus-based research in
order to generate parallel structures and contexts to test for recurrent speaker and
listener practices. Indeed, Experimental Pragmatics represents a new area of research at the intersection of traditional linguistic pragmatics and psycholinguistics. It seeks to apply experimental techniques from psycholinguistics to solve
problems defined in linguistic pragmatics (see Breheny this volume). So far such
experiments have focused on issues surrounding the conversational maxims and
implicatures of Grice (1975) and his followers (both Neo-Griceans like Horn 1984
and Levinson 2000, and Post-Griceans like Sperber and Wilson 1986 and Carston
2002), but creative experiment design could extend this new research paradigm to
all sorts of questions about spoken language. All the same, investigations of large
corpora provide a body of natural data against which experimental results can be
evaluated.
The discussions between scholars of different persuasions about what constitutes appropriate data are ongoing. The historical development of different data
sources and areas of inquiry as well as the continuing debates about their usability
and validity resonate in this handbook. Indeed, the authors contributing to this
handbook were chosen to represent various positions in these debates.
5.
Organization of this Handbook
To distinguish pragmatics as a linguistic discipline from pragmatics as an attitude or

a mode of behavior (a practical rather than an idealistic approach to problems and
dealings) as used in everyday, in political, in economic and in other kinds of discourse, the present handbook Foundations of Pragmatics opens with a section on
Conceptual Foundations consisting of the description of pragmatics as a linguistic
concept. Fundamental concepts covered are: micropragmatics, i.e. the pragmatics of
utterance based concepts such as speech acts etc. (which are covered in more detail
in vol. 2 of the series) as opposed to macropragmatics, i.e. the pragmatics of discourse or text based concepts such as topics, discourse markers etc. (cf. a more detailed overview in vol. 3 of the series), pragmalinguistics (dealing with forms) as opposed to sociopragmatics (dealing with pragmatic strategies), and metapragmatics.
The second major section is dedicated to the Theoretical Foundations of pragmatics. It begins with a historiographic overview of how pragmatics was established as a linguistic field of study, how it developed and spread, and how it eventually became an international success story. After all, since the so-called
pragmatic turn, linguistic pragmatics has, arguably, developed more rapidly and
diversely than any other linguistic field. The scientific disciplines that nourished
pragmatics in the first place range from semiotics (Peirce, Morris, Carnap) and
(Natural Language) philosophy (Wittgenstein, Strawson, Austin, Searle, Grice) to
anthropology (Malinowski, Wegener, Gardiner), contextualism (Firth) and functionalism (Bhler, Jakobson, Hymes), from ethnomethodology (Garfinkel, Goffman, Sacks) to European sociology (Habermas).
The third major section, Key Topics in Pragmatic Description, investigates
those central concepts of pragmatic description which were the first to make their
way from grammar into pragmatics (indexicality and deixis, followed by reference
and anaphora), on the one hand, and those notions that rapidly became the focal descriptive construct in mainstream pragmatics (speech acts and types of inference
such as entailment, presupposition, implicature), on the other.
In the fourth major section, the Place of Pragmatics in the Description of Discourse, the scope of pragmatics within linguistic description is delimited vis--vis
grammar, semantics, prosody. A final chapter on pragmatics and literature underscores the contention that the social aspect of language use is as prominent in literary contexts as in any non-literary context.
Methods and Tools constitutes the fifth and final major section. It contains an
initial chapter comparing various methodological approaches to pragmatics such
as intuition, introspection, ethnographic field work, elicitation and corpus analysis,
before it then proceeds with articles on experimental pragmatics, corpus-based
pragmatics and the transcription of face-to-face interaction.
No handbook can include every individual perspective and specific topic even
regarding the foundations of an area of study. Our decisions regarding which fields
and issues to include and how they should be addressed reflect both our own views
of pragmatics and the expertise and interests of our authors. As authors were recruited for the handbook project, joined us, and began to think and write about their
areas, topics were redefined, expanded and narrowed, titles were adjusted, some
boundaries were sharpened, others redrawn. Recent developments in pragmatics,
the general availability of new kinds of multi-media data and analytic procedures
along with dynamically evolving attitudes toward topics and research methods
helped shape the contents of this volume. Inevitably, the exigencies of scheduling
and the special interests of the contributing authors conspire to yield a final set of
articles representing individual views at a specific point in time. We embrace this
particular character of the snapshot offered by the articles in this handbook.
6.
In the lead article, Pragmatics as a linguistic concept, Anita Fetzer seeks to limit
the concept of linguistic pragmatics. The chapter compares pragmatics as a linguistic concept with pragmatics as used in everyday, in political, in economic and
other kinds of discourse. It discusses what is special about linguistic pragmatics,
covering General Pragmatics as well. Pragmatics as a common-sense notion is
functionally synonymous with practical or just right at that stage, describing a
language-bound and/or action-based phenomenon within a particular situation, as
is reflected in its usage in everyday, in political, in economic and other kinds of discourse. In linguistics, semiotics or philosophy, pragmatics refers to a theoretical
concept, comprising contextual, generally usage-related constraints and requirements, for instance presuppositions, indexical expressions, felicity conditions,
common ground and background assumptions. These are necessary conditions for
communication to be felicitous in context. Thus, the perspective is shifted from the
analysis of the language system and its constitutive parts to that of its rule-governed instantiation and embeddedness in context. In General Pragmatics the focus
of investigation is on universals of communicative action, such as directness versus indirectness. These are refined in Linguistic Pragmatics, where they undergo
language- and culture-specific modification. That change in perspective has important consequences for methodology. To account for the parts-whole connectedness, a relational frame of reference is required which accommodates the relation
between the language system and language use, and interlocutors and what they do
with language in context, comprising linguistic context (or co-text), social context
and cognitive context.
In the second article, Micropragmatics and macropragmatics, Piotr Cap addresses the central distinction between micropragmatics and macropragmatics.
This chapter distinguishes micropragmatics as referring to speech acts/actions etc.
from macropragmatics as directed at the discourse/text level. Defining pragmatics
by its general perspective, rather than by its specific object of investigation, entails
looking at the analytic scope of pragmatics in terms of micro- and macro-level concepts. These have been traditionally (cf. Mey 1993) associated with the opposition
between the analysis of speech act force at an utterance level and the analysis of
global intentionality at the level of a discourse/text. Following this division, as
well as its later refinements, the article, first, reviews the locutionary, illocutionary
and perlocutionary constituents of force of an utterance, relating them to pragmatic
concepts which characterize the process of encoding and decoding its message
(deixis, presupposition, implicature etc.). Second, it demonstrates how individual
utterances comprising different sets of constituents and markers of force, including
syntactic markers of explicitness and inferable carriers of implicitness, can form
sequences to shape the global intentionality of a discourse/text. Invoking such notions as speech event and macro speech act, the article shows which pragmatic concepts utilized in (micro-) analysis of individual utterances are essentially complementary, i.e. which markers of force denoted by these concepts can collectively
generate complex macro illocutions responsible for accomplishing global
discourse/text goals.
The important distinction between Pragmalinguistics and sociopragmatics
forms the topic of chapter 3 by Sophia Marmaridou. In Principles of Pragmatics
(1983), Leech introduced a distinction between pragmalinguistics (roughly, what
form to use to achieve an intended pragmatic effect) and sociopragmatics (roughly,
when to use a particular pragmatic strategy). This chapter describes past and current research on this distinction in pragmatics. Within a broad definition of pragmatics as the scientific study of all aspects of linguistic behavior, the distinction
between pragmalinguistics and sociopragmatics is intended to focus on two methodological approaches. Pragmalinguistics concerns the study of the particular resources that a given language provides for conveying pragmatic meaning (illocutionary and interpersonal), whereas sociopragmatics, as the sociological interface
of pragmatics (Leech 1983: 10), relates pragmatic meaning to an assessment of
participants social distance, the language communitys social rules and appropriateness norms, discourse practices, and accepted behaviors (Thomas 1995). The
above methodological distinction to some extent relates to earlier work in Marxist
pragmatics (Mey 1979; Prucha 1983), foreshadows work in Critical Discourse
Analysis (Fairclough 1995), and intersects with studies in sociolinguistics. A critical assessment of the two concepts and the respective methodologies emerges from
10
current research practices. Even though the borderline between the two concepts
often seems fuzzy, their relevance to language learning, teaching and testing has
been enhanced (McNamara and Roever 2006) and so has its significance in historical corpus linguistics (Archer and Culpeper 2003).
In the fourth chapter, Axel Hbler addresses the subject of Metapragmatics.
He focuses on metapragmatics concerning the actual management of discourse.
Key tasks in metapragmatics are the investigation of that area of speakers competence which reflects the judgments of appropriateness of their own and other
peoples communicative behavior and the exploration of the know how regarding the control and planning of, as well as feedback on, the ongoing interaction.
The human capacity for reflecting about language and language use/communication is a feature of language-related competence that no animal is endowed with.
Part of this capacity extends to areas which linguists have become accustomed to
referring to as metapragmatics. The term is fairly new, and its usage not yet fully
regulated; it accommodates various systematic perspectives and pertains to a wide
range of language-reflexive phenomena. The article provides a comprehensive survey along two lines: (1) what linguists (and experts from other disciplines such as
philosophy or social-psychology) have observed about how people act and interact
and what they express about their practice; and (2) what these experts have observed about their own descriptive, analytical and theorizing practice. Line (1) includes theories of pragmatics as prefiguring observation and resulting from it; and
it takes account of what could be categorized as (implicit or explicit) meta-practice
by members of a speech community (e.g., monitoring, remedying, evaluating by,
e.g., quoting, topicalizing, metaphorizing). Line (2) comprises meta-theoretical
considerations, which (critically) reflect upon and compare varying theoretical
stances or elaborate on the difference to be drawn between a scientific view on
pragmatics and folk theories.
7.
The opening chapter 5 in this section, The rise of pragmatics: a historiographic

overview, by Wataru Koyama traces the history of the still young field of linguistic pragmatics. This chapter stresses the success story as well as the newness of
the field of pragmatics. Especially after the artificial impoverishment of linguistics
due to structuralism and generative grammar, there was clearly a need for pragmatics as a field to amalgamate insights from rhetoric, anthropology, functionalism
etc. When and by whom was linguistic pragmatics invented? Who were its
(extra-linguistic) predecessors? What was the nature of the scientific and, particularly, the linguistic environment, in which pragmatics could arise (the 1950s; logical positivism etc.). Then the author considers further developments: how and why
pragmatics developed, slowly at first, abruptly later in the 1980s (Levinson 1983,
11
International Pragmatics Association, Journal of Pragmatics etc.). Finally, he

looks at how todays pragmatics differs from earlier stages of pragmatics.
In the sixth article, Semiotic foundations of pragmatics, Winfried Nth relates pragmatics to its foundations in semiotics. The chapter shows how semiotics
influenced the development of the theoretical foundations of pragmatics, focusing
on the contributions of Peirce, Morris and Carnap. Semiotics, the general study of
signs and sign processes, is fundamental to pragmatics, the study of language use,
speech acts, and communicative processes, and it extends the framework of language use to include nonverbal, visual, and other signs in the context of verbal behavior. According to the general framework outlined by Carnap and Morris, pragmatics is one of the three branches of semiotics besides syntax and semantics, but
twentieth century structuralist semiotics tended to neglect the pragmatic dimension
in its study of language and discourse. Central issues of pragmatics, such as the
questions concerning intention and purpose, meaning and reference, word and object, interpretation and communication, or the relationship between language and
action, find fundamental answers in the writings of the founder of general semiotics and philosophical pragmatism (or pragmaticism), Charles S. Peirce.
Chapter 7, Pragmatics in modern philosophy of language by Nikola Kompa
and Georg Meggle, traces the origins of linguistic pragmatics in the Philosophy of
Language, particularly in Natural Language Philosophy in the work of Wittgenstein, Strawson, Austin, Searle, Grice. There are two paradigms of Pragmatics in
modern studies in the Philosophy of Language. In the first, semantics is taken as
given and the aim is to incorporate as many context-sensitive aspects of language
as possible into that semantics; in the second, research is trying to spell out the semantics on a pragmatical (action-theoretical) basis itself. Accordingly, the difference between the two paradigms could be considered that between semantic pragmatics on the one side and pragmatic semantics on the other side. Characteristic
contributions to the first approach are those by Frege, Carnap, Kaplan and Stalnaker; characteristic contributions to the second are by Wittgenstein, Austin,
Grice, Lewis and Bennett. This article considers the basic problems as well as the
present state of the art in both paradigms of pragmatics and attempts to achieve
some clarity about their deeper (dis-) connection in particular
In chapter 8, Foundations of pragmatics in functional linguistics, Saskia
Daalder and Andreas Musolff trace the roots of pragmatics in functional approaches to linguistics. Pragmatics addresses the functions of language as found in
the work of Bhler, Jakobson, Hymes. This chapter traces the historical connections from these sources to contemporary thinking in pragmatics. The study of linguistic pragmatics received a major impetus from the systematic analysis of language functions as started by Karl Bhler, who distinguished three fundamental
dimensions of the linguistic sign: it functions at the same time as symbol, symptom
and signal. The model has proved its enduring value in the development of functionalist language theories, via its modification and augmentation into a six-func-
12
tions-model by Roman Jakobson. His model in turn informed subsequent functionally oriented theories of language and communication, e.g. Dell Hymes
Ethnography of Speaking (1962) and Michael A.K. Hallidays concept of Language as Social Semiotic (1978). In retrospect, Bhlers model (and, to a lesser extent, Jakobsons modified model) can be judged to suffer from its forced integration with the structuralist concept of the linguistic sign, which initially
hampered a fully fledged pragmatic interpretation.
In their chapter 9, Foundations: Ethnomethodology and Erving Goffman,
Christine Domke and Werner Holly discuss two related questions: What is Ethnomethodology or rather what are the crucial questions and assumptions of this
qualitative research approach? And how did Ethnomethodology impact linguistic
pragmatics? At first they overview the studies of Harold Garfinkel which contain
the groundbreaking thoughts about the most significant question: How is social
reality constructed and organized by activities of the societys members? Accordingly, it is shown how conversation analysts (e.g. Sacks, Schegloff) are applying
Garfinkels fundamental assumption of the social reality accomplished through
peoples everyday life activities in analyzing talk-in-interaction. Subsequently, the
main focus lies on the interface between social order, which Ethnomethodology
considers as produced by people interacting in their daily life, and the fundamental
pragmatic question what people are doing using speech and talking. How the idea
of (re-)produced social order is in line with and can be linked to the pragmatic
focus on the role of language/verbal interaction in different forms of daily interaction.
The work of Habermas has been central in the development of pragmatics in
Europe. In her contribution in chapter 10, Pragmatics in Habermas Critical Social Theory, Maeve Cooke considers the influence of Habermas on linguistic
pragmatics. She deals with the pragmatic turn that Jrgen Habermas introduced to
critical social theory with his program of universal pragmatics in the 1970s. With
this program, subsequently referred to as formal pragmatics, Habermas investigated everyday linguistic interaction from the point of view of its orientation towards validity. His investigations were guided by the critical intentions of his social theory. His principal intentions were to provide a normative basis for a conception of practical rationality without reference to otherworldly authorities or
entities and to develop a corresponding normative account of social action. However, his formal pragmatics has profound implications for social theory in general
and, indeed, for anyone concerned with questions of meaning and validity. This article sets out the key elements of Habermas formal pragmatics, shows its implications for social theory and beyond, and discusses some important objections that
can be raised against him.
8.
13
Pragmatics first made inroads into linguistics through the pioneering work of Yehoshua Bar-Hillel. In this chapter 11, William Hanks describes past and current research on Deixis and indexicality from a pragmatic perspective. He starts from a
historical sketch of indexicality in the semiotics of Peirce and Morris, and traces
the development of the concept by philosophers, linguists, linguistic anthropologists, sociolinguists (via style, register, contextualization) and conversation
analysts. Called shifters by Jespersen and Jakobson, deictics encode a relational
structure projecting from an indexical ground (a context of utterance) to a referential figure (the object denoted) via a relation (proximity, perceptibility, anaphora,
etc.). Deictic systems differ cross-linguistically in ways just beginning to emerge.
Pronouns are a related type of shifter and are closely tied into participation frameworks as treated by Goffman, conversation analysts and linguistic anthropologists. Recent debate is summarized regarding the role of spatial proximity vs.
cognitive-social accessibility as a basis for deixis. The final section outlines empirical approaches to pragmatic research on deixis.
Reference and anaphora are inherently pragmatic in nature and their treatment
in linguistic description helped bring pragmatic notions into prominence.
Chapter 12 on Reference and anaphora by Monika Schwarz-Friesel and Manfred
Consten concerns reference, the area in scientific research which deals with the
ancient problem of how language relates to the world. In the tradition of philosophical logics and formal semantics, reference has been defined as a static relationship between expressions and the things they denote. In the course of pragmatic change in linguistics, models of reference have arisen that (to different
extents) take cognitive properties of language users into account. In the framework
of the latter notion, anaphora is a specific kind of reference that has to be resolved
mainly through the activation of discourse based knowledge established by the
preceding text. Thus, anaphora resolution is analysed as a process of reactivating
pre-mentioned referents. The traditional view of anaphora as a means of referential
continuity and the availability of given information has to be revised. From a procedural perspective, anaphora may add new information to the mental files of the
referents in the text-world model.
Elena Collavin addresses the key topic of Speech acts in chapter 13. The
chapter provides an overview of the notions of act/action (versus behavior) and
speech act: how they developed, how they made their way into linguistics, how
they became the core notions of various (speech act) theories, how the latter developed etc. Collavin presents an account of the notion of speech act, which includes
an illustration of the main formulations of the theory, its limitations, and its current
relevance and application in the broad field of pragmatic research. She traces the
origins of this notion and describes the main formulations of Speech Act Theory,
focusing on conventionality, intentionality, and action as the key aspects of speech
14
acts around which similarities and incompatibilities in different versions of the theory are established. The discussion then turns to the criticisms moved to various
notions of speech act. The relevance of speech acts for the empirical analysis of
language use in interaction has been strongly debated. Collavin discusses how
some notion of speech act can relate to conceptualizations nowadays fundamental
in the study of language usage such as frame, voicing, indexicality, multimodality,
shared cognition and embodiment.
Chapter 14 provides an overview of the varieties of inference recognized in
pragmatics, how they developed, how they made their way into linguistics, how
they became core notions of various theories, how the latter developed etc. In this
chapter, Types of inference: entailment, presupposition, and implicature, Yan
Huang provides a critical survey of three fundamental types of inference recognized in semantics and pragmatics, namely entailment, presupposition and implicature, covering their origin, development, and the roles they play in various semantic/pragmatic theories in linguistics and the philosophy of language. Of the three
categories of inference, entailment is essentially semantic in nature, though the notion of pragmatic entailment is discussed. Presupposition straddles the semanticspragmatics boundary, resulting in semantic and pragmatic presupposition. Finally,
with regard to the two subtypes of implicature, while conversational implicature is
pragmatic in nature, conventional implicature can be considered either as semantic
or pragmatic, depending on how the distinction between semantics and pragmatics
is drawn. Notions parallel to conversational implicature (such as explicature) and
the argument that conversational implicature should not be treated as inference are
also considered. Finally, comparisons between entailment and presupposition on
the one hand, and between presupposition and implicature on the other, are presented.
9.
Part IV opens with chapter 15 Pragmatics and grammar by Arnulf Deppermann.

This chapter details the delineation of pragmatics and grammar and the integration
of pragmatics and grammar into linguistic theory in relation to each other, focusing
on issues, methods and areas covered. Deppermann discusses theoretical approaches, linguistic phenomena and empirical methods concerning the relationship
between grammar and pragmatics. He sketches functional approaches to grammar,
which claim that grammar is shaped and constrained by cognition, text and/or social interaction (functional grammar, systemic-functional linguistics, construction
grammar, cognitive grammar, interactional linguistics). Historical studies on the
paths of grammaticalization, psycholinguistic insights from usage-based approaches, and conversation analysis are introduced as methods for the study of
the relationship between grammar and pragmatics. The relevance of grammar for
15
pragmatics is discussed concerning a broad range of issues, such as cognitive processes (implicature, inference, information structure), the expression of cognitive
states (epistemic stance), conversational organization (repair, turn-construction
and turn-taking, collaborative action, intersubjective grounding), action structure
(speech acts, genres), bodily action (gesture) and socio-structural concerns (politeness, power, social styles).
In chapter 16 on Pragmatics and semantics, John Saeed discusses some contrasting views of the relationship between semantics and pragmatics. His chapter
details the delineation of pragmatics and semantics in linguistic theory, focusing
on issues, methods, areas covered. Within the philosophical and formal traditions,
he considers the growth of contextualism, that is the increasing emphasis on the
underspecification of meaning in linguistic form, illustrated by phenomena beyond
traditional deixis (indexicals) such as gradable and comparative adjectives, quantifiers, possessives etc. The article also looks at the defense against this from semantic minimalism (e.g. Cappelen and Lepore 2005). As an example of contextualism, Saeed discusses aspects of Relevance Theory (Sperber and Wilson 1986,
Carston 2002), with its identification of pragmatic principles reaching deep into
what earlier accounts viewed as semantic content. In particular, he considers the
theorys notions of explicatures and higher level explicatures, and looks at its view
of lexical pragmatics. Finally, he brings into the discussion the quite different view
from cognitive linguistics, where the very distinction between semantics and pragmatics is questioned. So in theories like Fauconniers, mechanisms of mental
spaces and conceptual blending seek to model phenomena from across the traditional divide, including e.g. tense, modality, presupposition and metaphor. An interesting question is whether a pragmatic theory like Relevance Theory and the
cognitive semantics of writers like Lakoff, Fauconnier and Langacker are in fact
compatible.
Elisabeth Couper-Kuhlen describes the role of prosody in pragmatics, and the
integration of prosody into linguistic theory. In chapter 17 Pragmatics and prosody, she contemplates what has been achieved so far in the field of prosody and
interaction and what still remains to be done. Instead of terms such as suprasegmentals, sentence types and illocutions, she speaks of prosodic resources for
single- and multi-turn construction, action formation and floor management. Comparing the use of prosodic resources across turns-at-talk, she acknowledges the
contribution of prosody to marking the continuation of same-speaker turns, to organizing sequences of turns, to cueing problems, negotiating agreement and signaling affiliation in talk. But recent developments also present challenges to researchers in the field of prosody and interaction: (i) What counts as prosodic? and
(ii) What is prosodys place in the analysis of multimodal interaction? The chapter
discusses these issues and ways of dealing with them in the future.
In chapter 18, Jacob Mey discusses Pragmatics and literature. According to
Mey (2000), pragmatics is concerned with the way humans use language in a social
16
context: When people use language in a literary context, the social aspect is no less
prominent. Literary pragmatics, therefore, concerns itself with the relation between the language users in the context of literature, that is, the producers (the
authors) and the consumers (the readers). In the context of literary production,
the processes of producing and consuming are paramount. The product (the written
text) can only be evaluated properly in terms of its social conditions and the concomitant roles of (authorial) production and (readerly) consumption. Here, the key
concept of voice (originally due to Bakhtin) plays a major role. The author assumes a narrative voice, while retaining a distinct personal voice. Conversely, the
readers have a voice, inasmuch as they identify with, and co-construct, the narrative personae; in this way, they take part in the building of the literary universe
and making it come alive. Some implications of this pragmatic perspective are followed up through analysis of snippets of literary texts.
10.
The initial chapter 19 in this final part by Monika Bednarek, Approaching the data
of pragmatics, provides an overview of how various aspects of language use
are studied using different methodological approaches: intuition, introspection,
ethnographic field work, elicitation (and experimentation), and corpus analysis.
It compares critically the various methodologies that are employed in areas such
as conversation analysis, politeness theory, and speech act theory as well as in related functional approaches. The chapter covers both quantitative and qualitative
methodologies ranging from intuition and introspection to interviews, discourse
completion tasks, ethnographic field work, elicited data and corpus analysis. It
takes a broad approach to the data of pragmatics in that it not only includes research in key pragmatic areas but also disciplines related to pragmatics such as ethnography and sociolinguistics. It describes different methodologies and compares
the insights gained from them as well as the different perspectives they provide on
linguistic data.
In chapter 20, Experimental pragmatics, Richard Breheny surveys recent research in this emerging field. Experimental pragmatics aims to use the methods of
experimental psychology to test claims found in the more theoretical linguistics literature. The strategy of testing theoretical ideas experimentally has been around
for a while and is notably exemplified in the work of Clark, Sanford, Gibbs and
Glucksberg. More recently, there has been an upturn in the interest in experimental
methods, coinciding with renewed interest in pragmatic questions such as the nature of generalized implicatures, presupposition, metaphor and the like. Breheny
reviews studies involving children, adolescents and adults from clinical populations as well as typically developing groups using the full range of methods to be
found in psycholinguistics and developmental psychology (including both beha-
17
vioral and neuro-scientific methodologies). The contribution also contains a discussion of the difficulty of drawing empirically testable predictions from claims
made in the theoretical literature.
The final three chapters in the volume all address the role and forms of corpus
data in linguistic pragmatics. The first two consider corpus-based research in
pragmatics, first qualitative then quantitative studies, and the third discusses
methods of transcription. In chapter 21, Corpus-based pragmatics I: qualitative
studies, Gisle Andersen describes the practice and potentials of corpus-based research in pragmatics, focusing on qualitative approaches to language data. He
considers the development of spoken and written corpora like the Brown Corpus,
the London-Lund Corpus, the Helsinki Corpus, the British National Corpus, the
Santa Barbara Corpus of Spoken American English etc., and their growing influence on research in pragmatics over the last thirty years, paying attention to
special corpora, large general corpora, continually growing monitor corpora and
web-based corpora. Corpora allow for in-depth studies of patterns of frequency,
distribution and collocation, thus facilitating statistically based observations of
innovative language use and variation. Dynamic and ever-growing corpora allow
for the observation of emerging repetitive patterns such as the routinization of
forms with discourse marking functions, ongoing grammaticalization and the development of new abstract and interactional meanings from old forms. In contrast
to quantitative, statistical approaches to corpus data, qualitative studies see corpora primarily as a source of natural data and a way of overcoming the fallible introspection of armchair linguistics. In the area of qualitative pragmatic studies,
corpora have provided a constantly growing pool of data, allowing researchers to
test their hypotheses about individual items and constructions, as well as practices
and usage within particular groups, but also suggesting new avenues of research.
Diachronically oriented corpora are responsible for the rapid progress made
in historical pragmatics. At the same time, corpora representing specific groups,
genres, registers or stages of a language have led to research of various kinds, including the language of adolescents, foreign language learners and other user
groups.
In the companion chapter 22, Corpus-based pragmatics II: quantitative
studies, Christoph Rhlemann takes up quantitative corpus-based research. Due
to the massive dependence of pragmatic phenomena on context, corpora, as a
unfit for use in pragmatic research. Nonetheless, corpus linguistic analyses, both
into key pragmatic phenomena. The aim of this article is to illustrate key quantitative corpus studies into phenomena of pragmatic interest. The article is divided
into four sections. Following a brief introduction that addresses general issues in
corpus linguistic research on pragmatic phenomena, such as the question of how
much or little context-sensitive corpora are, the second section introduces quanti-
18
tative studies carried out on the small number of corpora that are pragmatically annotated. The third, and largest, section discusses key corpus linguistic research into
pragmatic phenomena such as speech acts, deixis, evaluation, discourse marking,
backchannelling and vagueness. Further, as a borderline case between pragmatics
and semantics, research into semantic prosody will be reviewed. The concluding
section looks to the future, outlining inter alia recent attempts at building multimodal corpora that will potentially shed light on the interplay between utterance
and kinesics.
The final contribution 23 deals with various methods of transcribing face-toface interaction and their significance for pragmatic theorizing. In this chapter,
The transcription of face-to-face interaction, Roger J. Kreuz and Monica A.
Riordan consider issues in the representation of linguistic data. Researchers in
pragmatics frequently transcribe audio- or videotaped recordings of conversational
participants. A comprehensive transcription coding system would help to unify diverse lines of research and make collaboration and interpretation of results easier.
However, a comprehensive system currently does not exist. Kreuz and Riordan
review coding schemes that have been previously proposed and assess their
strengths and weaknesses. Most systems have focused solely on the acoustic (i.e.,
lexical and intonational) aspects of discourse between adult speakers, but they also
review systems that have been developed to transcribe the speech of other populations (e.g., children, aphasics and cognitively impaired individuals). In addition,
they make recommendations about how non-acoustic information can be coded.
Specifically, the challenges and utility of coding for facial and gestural information
are addressed. It is hoped that a more comprehensive coding system will make it
easier for researchers to describe and study the multimodal aspects of conversational interaction.
Taken together, these chapters constitute an outline of the fundamental issues
and trends in the study of pragmatics at the beginning of the twenty-first century.
They can be taken as a kind of extended definition of the term pragmatics as
practiced today, along with critical attempts to set down the history of the discipline and identify the sorts of questions taken to be important and the sorts of
answers being proposed. We hope in this way to contribute to the definition and coherence of the field, and thereby to foster research in the generation to come.
References
2003
Frankfurt am Main: Peter Lang.
19

2005
Carston, Robyn
2002
Eckert, Penelope
2000
Linguistic Variation as Social Practice. Malden, MA: Blackwell.
Fairclough, Norman
1995
Critical Discourse Analysis. London: Longman.
Goodwin, Charles
2003
Pointing as situated practice. In: Sotaro Kita (ed.), Pointing: Where Language,
Culture, and Cognition Meet, 217242. Mahwah, NJ: Erlbaum.
Grice, Herbert Paul
1975
Logic and Conversation. In: Peter Cole and Jerry L. Morgan (eds.). Syntax and
Semantics. Vol. 3: Speech Acts, 4158. New York: Academic Press.
1978
Language as Social Semiotic. London: Arnold.
Hanks, William
1996
Language and Communicative Practices. Boulder, CO: Westview Press.
Horn, Lawrence R.
1984
Toward a new taxonomy for pragmatic inference: Q-based and R-based implicature. In: Deborah Schiffrin (ed.), Meaning, Form, and Use in Context: Linguistic Applications, 1142. Washington DC.: Georgetown University Press.
Huang, Yan
2007
Pragmatics. Oxford: University Press.
Hymes, Dell
1962
The ethnography of speaking. In: Thomas Gladwin and William C. Sturtevant
(eds.), Anthropology and Human Behavior, 1353. Washington: Anthropological Society.
Leech, Geoffrey
1983
Levinson, Stephen
1983
Levinson, Stephen
2000
Presumptive Meanings. The Theory of Generalized Conversational Implicature. Cambridge, MA: MIT Press.
2006
56/2: 4379.
Mey, Jacob L.
1979
Mey, Jacob L.
1993
Mey, Jacob L.
2000
When Voices Clash: A Study in Literary Pragmatics. Berlin: Mouton de
Gruyter.
20
Morris, Charles
1938
Charles Morris (eds.). International Encyclopaedia of Unified Science,
77138. Chicago: University of Chicago Press.
Prucha, Jan
1983
1986. Relevance: Communication and Cognition. Oxford: Blackwell.
Thomas, Jenny
1995
Wenger, Etienne
1998
Communities of Practice: Learning, Meaning, and Identity. Cambridge: Cambridge University Press.
Part I
Conceptual foundations
1.
Pragmatics as a linguistic concept1

Anita Fetzer
1.
Introduction
Pragmatics is fundamentally concerned with communicative action and its felicity

in context, investigating action with respect to the questions of what action is, what
may count as action, what action is composed of, what conditions need to be satisfied for action to be felicitous, and how action is related to context. These research questions and the object of research require action in general and communicative action in particular to be conceived of as relational concepts, relating action
and context, relating action and communicative action, relating communicative action and interlocutors, and relating interlocutors with the things they do with words
in context. The inherent parts-whole perspective requires the explicit accommodation of a theory of context, capturing the dynamic processes from the connectedness between parts and wholes thus transcending clearly delimited frames of investigation.
The multifaceted research paradigm of pragmatics has provided new directions
and perspectives in arts and humanities, philosophy, cognitive science, computer
science and social science. Pragmatic perspectives have been employed in information technology and in the social sciences, particularly in economics, politics
and education. In line with Wittgensteins conception of language games and Levinsons conception of activity types (Levinson 1979; Wittgenstein 1958), game
theory (Lewis 1979) has been adapted to their particular contextual constraints and
requirements. The change of perspective from objects to interlocutors (or users),
and the accommodation of their wants and needs is reflected in the paradigm-anchored collocations of pragmatic thinking and learning, pragmatic software, pragmatic design, pragmatic modelling and pragmatic technology on the one hand, and
pragmatic foreign policy, pragmatic politicians, pragmatic sanction and pragmatic
nationalism on the other. Hence, it is no longer solely individual actions and their
perlocutionary effects, which are at the heart of analysis, but rather whole games,
considering both the present state of affairs and its connectedness with prior and
succeeding actions. In philosophical contexts, pragmatics is related closely to William James, who pointed out its practical nature, and to Charles Peirce, who
referred back to Immanuel Kants conception of pragmatics as experimental, empirical and purposive thought (e.g., The New Encyclopaedia Britannica 1976; cf.
also Koyama this volume). Pragmatics is also frequently connected with Greek
pragma, meaning action and instructive or useful (e.g., Bublitz 2009; Cummings 2005).
24
Anita Fetzer
The connectedness between action and purpose is also reflected in its ordinary-language meaning learning by doing (cf. Longman DCE 2003; Collins
Cobuild 1990). As a non-scholarly, ordinary-language notion, pragmatics is functionally synonymous with practical or just right at that stage, describing a surface-bound and action-based phenomenon within a particular situation. It occurs
primarily in its adjectival form, foregrounding practical and unexpected, rather
than theoretical and dogmatic considerations. For instance, politicians and economists may argue for a case on wholly pragmatic grounds, brushing aside traditionally established borders, which tend to result from presupposed and taken-forgranted ideology. Whenever an action, decision or solution is assigned the status
of a pragmatic action, a pragmatic decision or a pragmatic solution, it is contrasted implicitly with its theory-based and abstract-principles-driven counterparts. Against this background, the adjective pragmatic in its function as a modifier anchors the head of a noun phrase to the here and now of discourse. To employ
Gricean terminology from his cooperative principle, it signals that the ongoing discourse is such as is required, at the current stage at which it occurs, by the accepted purpose or direction of the talk exchange (Grice 1975: 45). Pragmatics in
its non-scholarly ordinary-language meaning thus meets with one of the most fundamental premises of scholarly pragmatics, the Gricean cooperative principle.
There is a further bridging point between non-scholarly pragmatics and scholarly
pragmatics: the differentiation between brute facts and institutional facts. According to Searle (1995), facts, that is brute facts, mental facts, social facts and institutional facts, constitute and regulate the world. To have a good knowledge of the
present situation, interlocutors need to have a good knowledge of what regulates
the world and what constitutes the world. Only then may they construct alternative
visions of a present situation or alternative social realities.
2.
Pragmatics in context
Pragmatics is frequently conceptualized as the science of language use, the study

of context-dependent meaning and the study of speaker-intended meaning, presupposing the existence of language, language user and context on the one hand, and
context-independent meaning on the other. To capture its multifaceted nature, definitions tend not to concentrate on the questions of what pragmatics is and what it
does, but rather on what it is not and what it does not do. The majority of conceptualizations pay tribute to Charles Morriss definition as the study of the relation
of signs to interpreters (Morris 1971: 6) and to Austins differentiation between
constative and performative (Austin 1971), foregrounding the performance of
communicative action and the effects the uttering of words may have. Against this
background, pragmatics is considered to be the study of invisible meaning (Yule
1996: 127), the science of the unsaid (Mey 2001: 194), the study of meaning as it
25
emerges in language use (Marmaridou 2000: 1), and the study of linguistic acts
and communicative action and their appropriateness (Bublitz 2009; van Dijk 2008;
Fetzer 2004, 2007).
There have been various attempts to systematize the multifaceted and heterogeneous field of pragmatics, or to bring some order into the pragmatic wastebasket
(Bar-Hillel 1971), as is reflected in the differentiation between the broad fields of
scholarly pragmatics: general (or non-linguistic) pragmatics and linguistic pragmatics. Implicit in the philosophy-anchored general-action and linguistic-actionbased paradigms is the premise that pragmatics is conceptualized best as a perspective, which comprises a general pragmatic perspective, a social perspective, a compositional perspective and a relational perspective. These are introduced below.
2.1.
The pragmatic-perspective paradigm does not represent a clearly delimited field of

research but rather offers a general perspective towards the object to be examined.
It provides a point of view for analysis, in which not the object in isolation is examined but rather the nature of its connectedness with users and other objects, or with
the theoretical framework to which it is anchored. The perspective is thus shifted
from an analysis of the language system and its constitutive parts to that of its rulegoverned instantiation and embeddedness in context, considering its generalized
and particularized conditions of use. That change in perspective has important consequences for methodology. To account for the parts-whole connectedness, a relational frame of reference is adopted, which accommodates the relation between
the language system, language use and language users, and what they do with
words in context, comprising linguistic context (or co-text), social context and
cognitive context.
The pragmatic-perspective paradigm provides a general cognitive, social and
cultural perspective on linguistic phenomena in relation to their usage in forms of
behaviour, accounting for the dynamics of language and language use, as is reflected in the premises that meaning is not a product and given but rather dynamic,
multifaceted and negotiated in context. Language use is dependent on variability,
and language users adapt to contextual constraints accommodating them not
only in the formulation of utterances but also in their interpretation (e.g. Mey 2001;
Verschueren 1999). As in functional grammar and systemic functional grammar
(Givn 1993; Halliday 1994), the dynamic construct of meaning is a necessary
condition for making linguistic choices. Following Verschueren (1999: 5558), the
choices are made at every possible level of structure as regards form and strategy.
In line with the sociolinguistic premise of linguistic variation, which claims that
speaker-intended meaning can be formulated with different linguistic surfaces
(Brown 1995), these choices are not equivalent in status. They carry along their alternatives and are always seen and interpreted against the background of other,
26
Anita Fetzer
more or less marked possibilities. While unmarked variants are seen as more frequent in distribution, more conventional in formulation and easier to process,
marked variants are less frequent in distribution, less conventional in formulation
and harder to process (Fetzer 2002, 2007).
The pragmatic perspective is not only reflected in the more general approach
examined above, considering linguistic communication as well as linguistic behaviour. It is also found in the particularized, hyphenated fields of pragmatics, as in
cognitive pragmatics (cf. vol. 4 Cognitive pragmatics of this series Handbooks of
Pragmatics), empirical pragmatics (cf. Bednarek this volume, Andersen this volume, Rhlemann this volume), experimental pragmatics (cf. Breheny this volume), formal pragmatics, historical pragmatics (cf. vol. 8 Historical pragmatics of
this series Handbooks of Pragmatics), intercultural pragmatics (cf. vol. 7 Pragmatics across languages and cultures of this series Handbooks of Pragmatics), interlanguage pragmatics, sociopragmatics (cf. vol. 5 Pragmatics of society of this
series Handbooks of Pragmatics) and theoretical pragmatics.
2.2.
The social perspective
An analysis of the rule-governed and strategic use of language in context anchored

in a social-perspective goes beyond traditional linguistic analyses. It also considers
social and sociocultural aspects of communication in an explicit manner, assigning
them the status of a constitutive part of communicative action, thus overlapping
with the research paradigms of anthropological linguistics (Duranti 1997; Hanks
1996), interactional sociolinguistics (Auer and Di Luzio 1992; Gumperz 1992,
1996, 2003), (critical) discourse analysis (Fairclough 1992; van Dijk 2008) and
ethnomethodological conversation analysis (Garfinkel 1994; Heritage 1984).
Societal pragmatics examines social parameters which affect the production
and interpretation of utterances, placing language use in an external relation to language users. It is based on the premise that language use and social structure are
connected dialectically, (re)constructing social and sociocultural context by confirming (or disconfirming) social values in interaction, for instance gender, power
and social status. Consequently, the language user and her/his social roles and
identities are at the heart of societal pragmatics, analysing social and sociocultural
context on the micro level of face-to-face interaction as well as on the macro level
of institution. In interaction, language users interactionally organize their social
roles and identities while at the same time constituting social context. To use Heritages (1984: 242) terminology, the production of talk is doubly contextual. An utterance relies upon the existing context for its production and interpretation, and it
is, in its own right, an event that shapes a new context for the action that will follow.
The social perspective to pragmatics has shifted the focus of investigating context-dependent meaning from semantics-based methodologies and their premise of
27
semantic well-formedness and propositional truth to social context-based approaches and their premise of negotiated communicative meaning and sociocultural appropriateness (Fetzer 2007), thus adding a further layer of meaning: social
and sociocultural meaning. In this perspective, meaning is not given per se and carried as such into the utterance, metaphorically speaking. Rather, meaning, viz.
lexical meaning and utterance meaning, is brought into the utterance and its diverse layers of meaning are brought out in the utterance, foregrounding salient
meanings while backgrounding non-salient ones (to employ interactional-sociolinguistic terminology, Gumperz 1992, 1996). Language is assigned the status of a
sociocultural construct which is used strategically by rational language users in
context, considering possible perlocutionary effects their utterances may trigger as
regards negative and positive politeness (Brown and Levinson 1987), for instance.
The social-meaning perspective also prevails in ethnographic studies, sociolinguistics and interactional sociolinguistics, as well as in ordinary language philosophy (Brandom 1994; Recanati 2004; Searle 1969, 1975, 1995). It is in disaccordance with frameworks conceptualizing language and language use as mutually
exclusive concepts, as is the case with the dichotomies of linguistic competence
and communicative performance, I(nternalized)-language and E(xternalized)-language, or langue and parole.
In the Anglo-American research tradition, the social perspective and the pragmatic perspective are referred to as European-Continental pragmatics. This field
defines pragmatics in a far broader way, encompassing much that goes under the
rubric of sociolinguistics, psycholinguistics, and discourse analysis (Huang 2007:
4), thus not falling strictly within formal linguistic theory (Horn and Ward 2004:
xi). The Anglo-American tradition tends to favour the compositional perspective.
2.3.
The compositional perspective
In the pragmatics-as-component perspective pragmatics is assigned the status of

a necessary component (or module) of grammar, thus anchoring pragmatics firmly
to linguistics (Birner and Ward 2006; Carston 2002; Huang 2007; Horn 1984;
Jaszczolt 2005; Levinson 2000; Sperber and Wilson 1995). Pragmatics is conceived of as an individual component of grammar, interacting with the components
of semantics, morphology, syntax and phonology, as is the case in its most prominent representatives: theoretical pragmatics and formal pragmatics. Radical pragmatics departs from the premise of strict modularity or discrete compositionality
by merging semantics and pragmatics. It is based on the premise of sense generality and contextualism, presuming that the output of semantic meaning is underdetermined and assigned determinate meaning in context through inference and pragmatic enrichment.
The compositional perspective is based on the premise that language is a theoretical construct focussing on language-internal rules and regularities, and on in-
28
Anita Fetzer
ferencing processes, thus considering language use and context in an explicit

manner. In contrast to societal pragmatics, however, its conception of context tends
to be restricted to linguistic context. In the case of deixis and its realization as deictic expressions, e.g., temporal and local deixis, participant deixis and discourse
deixis (Hanks 1996 and this volume; Levinson 1983), the compositional perspective needs to accommodate physical components anchored in the social context,
e.g., speaker and hearer, and time and place. Participants tend to be conceptualized
as the prototypical constructs of Model Speaker, Model Addressee or Model Person, viz. ideally competent interlocutors in the domains of linguistic competence,
communicative competence and rationality (Brown and Levinson 1987; Jaszczolt
2005), abstracting from their social, discursive and interactional roles. The same
holds for the conceptualization of time and place.
At the heart of compositional-perspective investigations are pragmatic universals, viz. deixis, reference, non-literal meaning, indefiniteness, presupposition, information structure, discourse marker, speech act and implicature. To account for
their function, pragmatics differentiates between an abstract construct and its linguistic realization (or representation). For instance, the pragmatic universal of deixis is examined with respect to deictic expression, and the universal of speech
act is analysed with respect to its constitutive acts and their direct, indirect, and
conventionalized indirect realizations (Sbis 2002; Searle 1995; Brown and Levinson 1987). Closely connected with indirect speech acts are speaker- and heareroriented cognitive operations: the former are conventional implicature, generalized and particularized conversational implicature (Levinson 2000; Horn 1989;
Huang this volume), and impliciture (Bach 2006), and the latter are inference and
abduction (Cohen, Morgan and Pollack 1992; Givn 1989, 2005). Inferencing is
also indispensable for the definition of intentionality of communicative action and
non-literal meaning, for indefiniteness and discourse markers, and for presupposition and information structure. Another pragmatic universal is face (Brown and
Levinson 1987; Goffman 1981, 1986), which is also of key importance to the social perspective.
More recently, the compositional perspective has undergone a shift in focus
and methodology, calling for the accommodation of methodological compositionality to bridge the gap between two competing paradigms (Jaszczolt 2005).
Only then is it possible to account for speech act and uptake, speech act type and
cognitive reality, and speech act type and situation (mis)match. The change in direction from monological, speaker-centred methodologies to dialogical, co-participant-centred interaction requires the explicit consideration of the connectedness
amongst world knowledge, linguistic knowledge, discourse knowledge, the properties of the human inferential system and stereotypes and presumptions about society and culture, as is accommodated in the relational perspective.
2.4
29
The relational perspective
The three perspectives adopted towards the heterogeneous field of pragmatics all
share a particular perspective towards their object of investigation, accounting for
its connectedness with context as regards utterance meaning and conditions of use.
For this reason, they adopt a more or less explicit relational frame of reference.
They go separate ways, however, in their accommodation of social context. While
the social perspective considers social context to be a key concept, the compositional perspective tends to disregard it, concentrating on linguistic context and
cognitive context. All of them are based on the premise that language is as a rulegoverned system, and all of them intend to account for form, function and meaning. The types of meaning under examination and their conceptualizations are different, though.
In a relational frame of reference, language may be examined from a parts perspective, analysing the form, function and meaning of individual parts, viz.
sounds, morphemes and constituents. It may be investigated from a holistic viewpoint, viz. phonological words and lexical words, and propositions and sentences,
and it may be analysed from a parts-whole perspective considering the nature of
their connectedness. The latter perspective is intrinsically dynamic, focussing not
on the object as such, but rather on its embeddedness in context, its conditions of
use and its appropriateness. The examination of linguistic form does not generally
cause delimitation problems, while the analyses of linguistic meaning and function
have become increasingly messy, necessitating the explicit accommodation of a
pragmatic wastebasket for non-fits and other-fits.
Pragmatics is generally delimited from semantics by shifting the focus of investigation from what is said, viz. semantic, literal or truth-conditional meaning, to
what is meant, viz. what is communicated implicitly, what is implicated, utterance
meaning or meaningnn (e.g., Grice 1975; Jaszczolt 1999, 2005; Levinson 1983,
2000; Recanati 2004; Saeed this volume). While the object of research is meaning
in semantics as well as in pragmatics, accounting for world-to-language, languageto-world and language-to-world-to-mind relations, the delimitation of the object,
the focus of investigation and the methodologies employed differ. In semantics the
meaning of a lexeme, proposition or discourse unit is defined from a unit-of-investigation-internal perspective, and in pragmatics, the meaning of an utterance,
speech act or conversational contribution is defined from unit-of-investigation-external and unit-of-investigation-internal perspectives. Metaphorically speaking,
semantics examines what is in the container, and pragmatics examines what is
brought out of the container (Reddy 1979).
The following figure attempts to bring some order in the pragmatic wastebasket
by systematising the two closely related but yet distinct fields of research, which
intersect and overlap in their analysis of meaning in context.2 While meaning is
conceptualised as conventional, truth-conditional meaning with discrete bound-
30
Anita Fetzer
Figure 1. Meaning in context
31
aries in semantics, pragmatic meaning is seen as speaker-intended, non-truthconditional meaning with fuzzy boundaries, as is indicated by the dotted lines. Furthermore, meaning is autonomous and represented by well-formed context-independent sentences in semantics, and it is inferred and represented by appropriate
context-dependent utterances in pragmatics. Semantics and pragmatics meet at the
interpretation interface, where the calculation of utterance meaning is informed by
the linguistic code and at the same time by pragmatic inferencing. While the
former is a unidirectional operation as is signified by single arrows, the latter is dialectical and interactive in nature as is indicated by double arrows, capturing the dynamics of pragmatic meaning:
The performative and its constitutive parts, viz. meaningnn, non-truth-conditional meaning and presupposition, are among the most important pillars of pragmatics. They are also relevant to the definition of speech acts and their realization
in context as utterance meaning, which is a linguistic-pragmatics phenomenon par
excellence. It is at the interpretation interface where context-dependent implicatures are calculated and where pragmatics interacts with semantics, syntax, morphology and phonology.
In the following, the research domain of general pragmatics is introduced, concentrating on its bridging points with linguistic pragmatics.
3.
General pragmatics
General pragmatics is set firmly in the research paradigm of philosophy, considering cooperation, action theory, intentionality, rationality and context, thus providing relevant bridging points for language-anchored linguistic pragmatics. Both research paradigms analyse pragmatic universals and pragmatic mechanisms, such
as speech act and implicature (e.g. Austin 1971; Bach 1992, 2006; Grice 1975;
Sbis 2002; Searle 1975; Vanderveken and Kubo 2002), indexical and deixis (Recanati 2004; Levinson 1983), common ground and presupposition (Akman et al.
2001; Bouquet et al. 1999), focussing on the connectedness between parts and
wholes, and between content and context (Stalnaker 1999). While general pragmatics adopts an action-based frame of reference, linguistic pragmatics examines
the form and function of speech acts in various languages.
In general pragmatics, action and cooperation are analysed in the framework of
an ideal pragmatic situation in which communication is neither distorted by social
factors, e.g. institutional constraints, unequal social hierarchies and other forms of
oppression (e.g., Habermas 1998, cf. Cooke this volume), nor by interpersonal
considerations, such as face-wants and face needs (Brown and Levinson 1987),
politeness or etiquette (Watts, Ide and Ehlich 1992). The high degree of idealization inherent in this research paradigm is also reflected in its conception of language and language use, which does not refer to natural language as spoken by or-
32
Anita Fetzer
dinary language users in context but rather an ideal language built in accordance
with the language-system specific constraints and requirements. In an ideal pragmatic situation, all necessary felicity conditions obtain, such as normal input and
output conditions, propositional content and preparatory conditions, sincerity and
essential conditions (Searle 1969). Consequently, rational agents have ideal linguistic, communicative and cognitive competence as well as an ideal world knowledge (cf. Cohen, Morgan and Pollack 1992).
Pragmatic universals can be assigned the status of basic building blocks of
communication. Among the most prominent ones are the speech act and its constitutive acts, viz. propositional act composed of further constitutive acts and illocutionary act (Searle 1969; Vanderveken and Kubo 2002), indexicals, for instance
pure and impure indexicals, intentionals and perspectivals (Recanati 2004), and
presuppositions and presupposition triggers (Huang 2007; Levinson 1983; Stalnaker 1999). Another basic building block is the rational agent or model user,
who employs building blocks in a strategic manner to speaker-intend communicative meaning, thus achieving particular goals in context. The model user can be a
natural user, that is speaker and hearer of a natural language, or an artificial user,
such as a robot or a dialogue system. In accordance with the fundamental pragmatic premises of rationality and intentionality, the model user is equipped with a
reasoning device, accommodating pragmatic inferencing and default inferencing,
top-down reasoning and bottom-up reasoning, and abduction. This is a necessary
condition for a model user to calculate and formulate speaker-intended implicatures and to perform the hearer-intended logical operations of inference and abduction (Cohen, Morgan and Pollack 1992; Givn 1989, 2005; Huang this volume;
Sperber and Wilson 1995).
3.1.
Presupposition and common ground
Presupposition refers to a proposition or an inference whose validity is taken for

granted for a sentence to be true or for a speech act to be felicitous. Philosophy and
linguistics distinguish between pragmatic presupposition and semantic presupposition. The satisfaction of semantic presupposition is a necessary condition for
the truth value of a sentence, and the satisfaction of pragmatic presupposition is
necessary for a speech act to be appropriate in context. Pragmatic presuppositions
are accommodated in speech act theorys felicity conditions, which are considered
as linguistic and social context categories and their satisfaction is assigned the
status of a default configuration (Sbis 2002). Presuppositions are generated by
presupposition triggers, for instance definite descriptions, factive verbs, aspectual
verbs, change of state verbs, iteratives and clefts (cf. Huang 2007, this volume; Levinson 1983).
Some contexts allow presupposition neutralisation, and more recently presupposition triggers have been classified as hard presupposition triggers and soft pre-
33
supposition triggers. Presuppositions are connected intrinsically with propositions

and assumptions which are taken for granted and therefore do not need to be made
explicit. They are organized and administered in the framework of context sets
(Stalnaker 1999), which serve as common ground in communication. Consequently, pragmatic presupposition may be seen as a restriction on common ground.
Common ground is indispensable for philosophical and cognitive conceptions
of knowledge where it serves as background for reasoning and for retrieving
speaker-intended meaning and other types of implicit meaning, such as indexical
expression or implicature. In the field of computer science, common ground and
world knowledge are frequently conceptualized as a database. It is seen as comprising a set of propositions, which serve as a resource for the understanding of utterances. Common ground is implicit but can be made explicit via propositions,
and model users presuppose its validity and fall back on it when they retrieve implicatures (cf. Bublitz 2006).
Common ground in the sense of background also plays a fundamental role in
dialogue system modelling. According to Vanderveken and Kubo (2002), model
users negotiate the compatibility of background with utterances and their felicity
and satisfaction conditions in and through the process of communication. However, background and context are not identical because possible contexts of utterance can have different backgrounds. As a consequence of this, background contains not only mutual knowledge of facts about the conversational background but
also knowledge about the world and of the world, such as ethical norms and sociocultural values, transcending the common sense notion of context. Searle (1995)
considers background to be a necessary condition both for literal and non-literal
meaning, thus assigning it the status of a basic premise for felicitous communication. He defines background as an open-ended set of skills, pre-intentional assumptions and practices, which are not representational but rather enable intentional acts and states to be made manifest.
In Relevance Theory (Sperber and Wilson 1995), common ground is conceived
of as a common set of premises for inference rules. The relevance-theoretic conception differs from the traditional notion of mutual knowledge by its attempt to
avoid the logical consequence of infinite regress, which follows from the code
model of communication. To avoid infinite regress, Sperber and Wilson base their
theoretical framework on an approximation of mutual knowledge, namely cognitive environments and mutual manifestness.
Common ground is a presupposed common knowledge base, which is a necessary condition for felicitous communication. To that base, model users anchor their
communicative action and they fall back on it, should they require further information, which may not be encoded explicitly. The overall encompassing concept of common ground has been given a context-dependent interpretation as conversational record (Thomason 1992) based on Lewiss concept of accommodation
(Lewis 1979). Conversational record contains public information, presumptions,
34
Anita Fetzer
and an update operator. Other context-dependent subsets of common ground are

Clarks notions of personal common ground, which contains a model users subjective experience, and communal common ground, which stores social experience
(Clark 1996), and Fetzers dialogue common ground, which is differentiated in individual dialogue common ground, containing a model users representation of
her/his dialogue common ground, and collective dialogue common ground, viz. the
model users representation of the set of model users representation of dialogue
common ground (Fetzer 2004). All of the subsets need to be interconnected, constantly updated and, if necessary, revised.
3.2.
Context
The analysis of context-dependent meaning is at the heart of pragmatics, and for

this reason context is one of its key objects of investigation. The theory, practice
and implementation of context are also of relevance to diverse fields of investigation, ranging from philosophy and computer-mediated communication to cognitive science, in particular dialogue management, artificial speech production,
artificial intelligence, distributed knowledge representation, robotics and information technology.
The heterogeneous nature of context and the context-dependence of the concept itself have made it almost impossible for the scientific community to agree
upon one commonly shared definition or theoretical perspective, and frequently,
only a minute aspect of context is described, modelled or formalized (cf. Akman et
al. 2001; Blackburn et al. 2003; Bouquet et al. 1999). Because of its multifaceted
nature and inherent complexity, context is no longer considered an analytic prime
but rather seen from a parts-whole perspective as an entity containing sub-entities
(or sub-contexts).
The multilayered outlook on context contains a number of different perspectives. First, context is conceived as a frame of reference whose job it is to frame
content by delimiting the content while at the same time being framed and delimited by less immediate adjacent frames. The nature of the connectedness between the different frames is a structured whole composed of interconnected
frames (Goffman 1986). The gestalt-psychological figure-ground scenario prevails
in psychological and psycholinguistic perspectives on context. It is also adopted in
cognitive pragmatics as is reflected in the relevance-theoretic conception of context as an onion, metaphorically speaking. The individual layers are interconnected
and their order of inclusion corresponds to their order of accessibility (Sperber and
Wilson 1995), as is reflected in inferencing and other kinds of reasoning.
Second, context is seen as a dynamic construct, which is interactionally organized in and through the process of communication. This view prevails in ethnomethodology (Garfinkel 1994; Goodwin and Duranti 1992; Heritage 1984), interactional sociolinguistics (Gumperz 1996, 2003) and sociopragmatics (Bublitz 2009;
35
Fetzer 2007, Schmid 2003), where context is assigned the dual status of process
and product. The dynamic outlook is based on the premise of indexicality of social
action, and the joint construction of context. In the primarily qualitatively oriented
paradigms, context is connected intrinsically with adjacency pair, conditional relevance and the turn-taking system on the micro level, and with institution on the
macro level, whose order is captured through context-independent and contextsensitive constraints and requirements. Closely related to the conception of context
as a dynamic construct is its relational conception, which conceives it as relating
communicative actions and their surroundings, relating communicative actions,
relating individual participants and their individual surroundings, and relating the
set of individual participants and their communicative actions to their surroundings (Fetzer and Akman 2002).
Third, context is seen as given as is reflected in the presuppositional approach
to context, which is also referred to as common ground or background information
(Stalnaker 1999). Here, context is seen as a set of propositions, which participants
take for granted in interaction. This allows for two different conceptions of context: a static conception in which context is external to the utterance, and an interactive one, in which context is imported into the utterance while at the same time
invoking and reconstructing context.
The context-dependence of context is reflected in its status as given and external to the utterance, reconstructed and negotiated in communication, indexical, and
never saturated. A further classification of context is anchored to a holistic conception of context embedding its constitutive parts of model user, conversational contribution, surroundings and their presuppositions, viz. cognitive context, linguistic
context and social context (Fetzer 2002, 2004).
Linguistic context comprises language use and is delimited by the constraints
and requirements of genre. Language is composed of linguistic constructions (or
parts) embedded in adjacent linguistic constructions (further parts), composing a
whole clause, sentence, utterance, turn or text. Linguistic context is functionally
synonymous to co-text (de Beaugrande and Dressler 1981; Widdowson 2004), denoting a relational construct composed of local and global adjacency relations.
Analogously to speech-act-theoretic constitutive rules and regulative rules (Searle
1969), linguistic constructions constitute text. The production and interpretation of
linguistic constructions is constrained by the rules of grammar, and the production
and interpretation of speech acts are constrained by felicity conditions.
Cognitive context is the foundation on which inference and other forms of reasoning are based. Constitutive elements of cognitive context are mental representations, propositions, contextual assumptions and factual assumptions. Since cognitive contexts are anchored to an individual but are also required for a cognitively
based outlook on communication, they need to contain assumptions about mutual
cognitive environments. Thus, cognitive context is not only defined by representations but also by meta-representations. In the social-psychological paradigm,
36
Anita Fetzer
cognitive context is conceptualized along the lines of the gestalt-psychological distinction between figure and ground and the related metacommunicative concepts
of frame and framing (Bateson 1972; Goffman 1986). Frame is seen as a delimiting
device, which is (or delimits) a class or set of messages (or meaningful actions)
(Bateson 1972: 187). Because of its delimiting function, psychological frames are
exclusive, i.e. by including certain messages (or meaningful actions) within a
frame, certain other messages are excluded and they are inclusive, i.e. by excluding certain messages certain others are included (ibid.). This also holds for context which, analogously to frame is also structured and metacommunicative, or to
use Batesons words: the hypothesis depends upon the idea that this structured
context also occurs within a wider context a metacontext if you will and that
this sequence of contexts is an open, and conceivably infinite, series (Bateson
1972: 245).
Social context comprises the context of a communicative exchange and is defined by deducting linguistic context and cognitive context from a holistic conception of context. Constituents of social context are, for instance, model users, the
immediate concrete, physical surroundings including time and location, and the
macro contextual institutional and non-institutional domains. The connectedness
between language and language use on the one hand, and linguistic context, social
context and cognitive context on the other is reflected in deixis, viz. temporal deixis, local deixis, participant deixis, discourse deixis and social deixis (Hanks
1996). Furthermore, the category of model user (as speaker and hearer) can no
longer be conceived of as an analytic prime but needs to be refined by the accommodation of footing (Goffman 1981; Levinson 1988). The importance of social
context to communication is spelled out by Hanks as follows: Hence it is not that
people must share a grammar, but that they must share, to a degree, ways of orienting themselves in social context. This kind of sharing partial, orientational and
socially distributed may be attributed to the habitus, or relatively stable schemes
of perception to which actors are inculcated (Hanks 1996: 235).
3.3.
Cooperation
The pragmatic principle of cooperation is connected intrinsically with coordination and collaboration on the one hand (Clark 1996; Cummings 2005; Grosz and
Sidner 1992), and with the pragmatic principles of rationality and intentionality
(Cohen, Morgan and Pollack 1992; Searle 1983) on the other. The principles are
examined in philosophy as regards their contents, function and status (Brandom
1994; Penco 1999), and they are analysed in psychology as mental representations,
mental states and beliefs (Bouquet et al. 1999; Sperber and Wilson 1995). Against
this background, language and language use are seen as fundamentally cooperative, and subsequently as intrinsically discursive and social (e.g., Grice 1975;
Lewis 1979; Clark 1996).
37
The connectedness between cooperation and intentionality, and the differentiation between I-intentions and we-intentions are at the heart of Searles conception
of speech act theory. To use his own words: Collective intentionality presupposes
a Background sense of the other as a candidate for cooperative agency; that is, it
presupposes a sense of the others as more than just conscious agents, indeed as actual or potential members of a cooperative activity (Searle 1991: 414). In his
work, Searle is very explicit about the nature of the parts-whole connectedness of
individual speech acts and I-intentions on the one hand, and of a conversation or
discourse and we-intentions on the other: The reason that we-intentions cannot be
reduced to I-intentions, even I-intentions supplemented with beliefs and beliefs
about mutual beliefs, can be stated quite generally. The notion of a we-intention of
collective intentionality implies the notion of cooperation (Searle 1991: 406).
In the Gricean approach to natural-language communication, in which language is seen as both purposive and social, and rational and cognitive, cooperation
counts as the fundamental premise of communication. As a philosophical concept,
cooperation is anchored to the premise of rationality and intentionality. Consequently, conversations are rational endeavours and cooperative efforts, in which
model users realize and recognize a common purpose or set of purposes, or, at
least, some mutually accepted direction. This is reflected in the Gricean cooperative principle which reads as follows: Make your conversational contribution
such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged (Grice 1975: 45). The cooperative principle is a necessary requirement for the calculation of pragmatic
meaning or meaningnn, and is for this reason not open for negotiation. Model users
may, however, employ the pragmatic strategy of opting out, which is generally accompanied by explicit references to some higher-order moral principle, such as
confidentiality or trust. When they opt out, they generally spell out the reasons for
their non-compliance, thus acting in accordance with the cooperative principle on a
deeper level. The cooperative principle subsumes four conversational maxims, the
maxim of quality, the maxim of quantity, the maxim of relation, and the maxim of
manner (Grice 1975: 4546). Further constitutive parts of the cooperative principle are the conventional and conversational implicatures, which are dealt with
below in the section 4.2 on linguistic pragmatics.
A cooperation-anchored outlook on language and language use not only is
manifest in a conversation as a whole, but also in its building blocks of speech act
and in the act of referring and predicating. Reference operates on the phrasal level
and involves a speakers use of a linguistic expression, which tends to be a noun
phrase, to induce a hearer to access or create some entity in her/his mental model of
the discourse. Reference may be performed directly with a noun phrase and indirectly as deferred reference in the contexts of pars-pro-toto configurations, meaning transfer and figurative meaning. Referring seen from a cooperation viewpoint
is not restricted to the semantic domain of a proposition but also applies to the in-
38
Anita Fetzer
terpersonal and interactional domains, viz. negotiation of meaning and common

ground.
The status of cooperation as a universal pragmatic principle has been examined
extensively in general pragmatics, and there seems to be a general agreement on its
validity. However, the claim that the Gricean cooperative principle is universal has
been under severe attack. There have been modifications as regards the adoption of
a further constitutive pragmatic principle on the same level, the politeness principle (Leech 1983), and there have been various modifications of the number of
conversational maxims required for the calculation of pragmatic meaning as well
as their speaker- and hearer-oriented conceptualizations and interpretations (Horn
1984; Lakoff 1973; Levinson 2000).
Pragmatics is a field of research which intersects with various paradigms. As
regards meaning, it cannot but consider semantics. As regards rationality, it cannot
but consider cognitive science, and as regards cooperation, it cannot but consider
philosophy. Thus, pragmatics needs linguistics and linguistics needs pragmatics.
For a thorough investigation of meaning, we need to examine how things are done
with words, and how conversational contributions are structured to achieve particularized goals in context. This is done in linguistic pragmatics, where generalpragmatic principles and pragmatic universals undergo language- and culture-specific modification.
4.
Linguistic pragmatics
Linguistic pragmatics is defined as the science of language use, and [i]n the same
way as human actions change existing reality, linguistic actions also change the
world (Marmaridou 2000: 22). Linguistic pragmatics and general pragmatics
share almost identical goals: general pragmatics examines pragmatic principles,
mechanisms and universals in the context of action theory, rationality and intentionality, while linguistic pragmatics focuses on their instantiation in language and
language use. Hence linguistic pragmatics overlaps with general pragmatics, sharing its generalized principles, mechanisms and universals, and departs from the
generalized framework by concentrating on language as a general construct and on
languages as particularized instantiations.
In linguistics, language tends to be examined in its own right, accounting for
language use, but not for the model user. The language system comprises the constitutive subsystems of phonology, morphology, syntax and semantics. These may
be assigned the status of autonomous modules, as is the case in the formal paradigms, they may be conceived of as connected subsystems, as is the case in systemic functional grammar (Halliday 1994), or as dialectically connected subsystems with fuzzy boundaries, as for instance the cognitive paradigm (Givn 1993,
2005), where language use is connected dialectically with the language system and
39
the model user. Language is seen as a dynamic construct changing through patterned use in context, in which grammatical and pragmatic aspects of interpretation
interface (Ariel 2008: 257259).
The dialectical connectedness between languages constitutive subsystems
requires the explicit accommodation of pragmatic interfaces. The pragmatic perspective adopted towards the object to be examined does not consider the object in
isolation but rather its connectedness with its conditions of use and with the user.
This is reflected in grammaticalization or grammaticization (Traugott 1988) and in
pragmaticalization (Aijmer 1997). Pragmatic principles shape the lexicon, as is reflected in Horns division of pragmatic labour, viz. his hearer-oriented Q-principle
and speaker-oriented R-principle (Horn 1984), and their impact on the structure
and use of lexical items, e.g. pragmatic strengthening, scalar implicature or polarity items. It is also manifest in deictic expressions and tense and aspect markers,
whose meanings are constant but whose referents vary with the speaker and the
hearer, and with time and place of the utterance, and on the style, register or purpose of a speech act (Ariel 2008; Huang 2007). Furthermore, pragmatic principles
have a strong influence on the information status of constituents within and across
sentences regarding focus, topic (or theme) and comment (or rheme). Deviations
from the default generate implicatures, such as emphasis implicating a set of alternatives (Birner and Ward 1998; Lambrecht 1994).
4.1.
Speech act
Speech act theory is connected intrinsically with J.L. Austins groundbreaking

lectures on ordinary language philosophy, edited posthumously (Austin 1971).
Austin scrutinized the then prevailing view that language describes the world in
terms of true and false, pointing out its shortcomings if applied to natural (or ordinary) language. He compared and contrasted the constative view, according to
which sentences are true or false, with the performative approach, according to
which sentences are used to perform speech acts, and came to the conclusion that
model users do not just describe the world, but rather do things with words, even
when describing the world. This is further refined by J.R. Searles modifications
and refinement in his research on speech acts (Searle 1969, 1991), systematizing
the cognitive, linguistic and social contexts, in which speech acts are performed
by differentiating between constitutive rules, which are a necessary part of the
language game and make speech acts count as particularized speech acts, e.g.
promise, threat or rejection, and regulative rules, which are normative in nature.
Considering language use from an action-theoretic frame of reference has paved
the way for investigating linguistic meaning as well as speaker-intended meaning. It has shifted the focus of investigation from the rigid framework of formal
semantics to the action-theoretic premises of rationality, intentionality and communication.
40
Anita Fetzer
Speech act theorys basic unit of investigation is the speech act, and depending
on the frameworks employed, speech acts are divided into locutionary, illocutionary and perlocutionary acts (Austin 1971), or propositional and illocutionary acts
(Searle 1969), and their respective sub-acts, viz. reference and predication in the
Searlean paradigm, and the phonetic act, phatic act and rhetic act in Austins frame
of reference. Speech acts and their constitutive acts are performed simultaneously,
and a speech act may be realized as a direct speech act, as an indirect speech act
and as a conventionally indirect speech act. To account for a speech acts perlocutionary effects as regards the social and interpersonal planes of discourse, speech
acts have been further refined and redefined as face-threatening acts, in which the
impact of a speech acts perlocutionary effects on the interpersonal plane is calculated with respect to the model persons, to employ Brown and Levinsons terminology, negative and positive face wants (Brown and Levinson 1987).
Speech act theory is anchored in action theory on the one hand, and in language
and language use on the other. Speech acts are described in various ways: they are
classified as explicit performative utterances, viz. utterances in which the speech
act contains a performative formula realized by a first-person pronoun and a
speech-act verb, e.g. I/we hereby close the meeting or we hereby request you to pay
your dues; they are classified as first-order speech act and second-order speech act,
accounting for the differentiation between direct and indirect speech acts; they are
classified as micro speech act and macro speech act, accounting for the larger unit
of genre (van Dijk 1981; Fetzer 2002), and as nth order speech acts, capturing the
multi-layered nature of performing speech acts in context. Furthermore, speech
acts are categorized with respect to possible directions of fit between language and
the world, that is the words-to-world direction of fit for assertion, the world-towords direction of fit for directive and commissive, the double direction of fit for
declaration and the empty direction of fit for expressive (Searle 1979). In addition
to the illocutionary-point anchored taxonomies, speech act verbs are classified and
systematized according to the way they are used in a particular language (cf. Collavin this volume).
Speech act theory combines language-internal and language-external factors
by systematizing them in the framework of felicity conditions, which are classified
with regard to propositional content conditions, preparatory conditions, essential
conditions and sincerity conditions. While propositional content conditions generally specify restrictions on the content of the proposition, preparatory, essential
and sincerity conditions explicate language-external conditions concerning the
language game as well as speaker- and hearer-specific requirements.
Speech acts are neither true nor false but rather felicitous or infelicitous. To be
felicitous, they need to be produced (and interpreted) in accordance with generalized felicity conditions, viz. normal input and output conditions, the essential and
sincerity conditions, and with particularized felicity conditions, viz. proposition
content condition.
41
Speaking a language is seen as engaging in a (highly complex) rule-governed form of behavior (Searle 1969: 12). The rule-governed behaviour is specified in speech act theorys felicity conditions and in the requirement that the model
user in her/his role as speaker intends that her/his utterance will produce in a model
user in her/his role as hearer a belief that the sincerity and essential conditions obtain by means of the recognition of the intention to produce this belief, and that the
speaker intends that recognition to be achieved by means of the recognition of the
utterance as one conventionally used to produce such beliefs. Secondly, that the
semantic rules of the language spoken by the model users are such that the utterance is uttered sincerely and such as is required just in case all of the previous conditions obtain.
4.2.
Implicature
The cooperative principle counts as a universal principle in pragmatics, where it

represents the solid base to which communication in general and the formulation
and interpretation of communicative action in particular are anchored. This holds
especially for the calculation of context-dependent communicative meaning and
the necessary processes of inferencing required for the contextualization and enrichment of underspecified conversional contributions. To use Ariels words,
[u]nderdeterminacy is an inherent characteristic of human language, since no
natural language sentence can encode interlocutors intended statements fully
(Ariel 2008: 265). In a similar vein, but more explicitly, Levinson (1995) argues
that intentionality is a fundamental premise of natural-language communication.
In his words, human interaction, and thus communication depends on intentionascription. Achieving this is a computational miracle: inferences must be made
way beyond the available data. It is an abductive process of hypothesis formation,
yet it appears subjectively as fast and certain the inferences seem determinate,
though we are happy to revise them when forced to do so (Levinson 1995: 241).
Underdeterminacy and the necessary processes of inferencing are connected
with the Gricean paradigm and its differentiation between what is said and what is
meant, which both refer to utterances produced in context (Grice 1975); and they
are also connected with Relevance Theory (Sperber and Wilson 1995). In both
frames of reference, communication is seen as a context-dependent endeavour, in
which communicative meaning may go beyond the level of what has been said.
Hence, what is said cannot be equated with pure linguistic meaning but rather is
closely related to the conventional meaning of the words (the sentence) [] uttered (Grice 1975: 44). Unlike the rather controversial status of what is said in
semantics and pragmatics, what is meant has always been equated with what is
implicated.
Relevance Theory differentiates between a code model of communication and
an inference model. While the former is conventional, the latter is inferential and
42
Anita Fetzer
hence pragmatic by definition. Sperber and Wilson define inference as follows:

Inference is the process by which an assumption is accepted as true or probably
true on the strength of the truth or probable truth or other assumptions. It is thus a
form of fixation of belief (Sperber and Wilson 1995: 68). Their theory uses demonstrative inference as is employed in various types of logic, and non-demonstrative inference, which is employed in the process of inferential comprehension.
Unlike local deductive reasoning, inferential comprehension is global and based
on inductive rules.
Pragmatic inferencing occurs at various levels of comprehension. It is connected with the differentiation between explicature, which is a combination of linguistically encoded and contextually inferred conceptual features (Sperber and
Wilson 1995: 182), and different kinds of implication, viz. trivial implication, analytic implication, synthetic implication and contextual implication. Against this
background, implicature is a contextual assumption or implication which a
speaker, intending her utterance to be manifestly relevant, manifestly intended to
make manifest to the hearer. We will distinguish two kinds of implicatures: implicated premises and implicated conclusions (Sperber and Wilson 1995: 194195).
Implicated premises must be supplied by the hearer, who must either retrieve them
from memory or construct them by developing assumption schemas retrieved from
memory, and implicated conclusions are deduced from the explicatures of the
utterance and the context. What makes it possible to identify such conclusions as
implicatures is that the speaker must have expected the hearer to derive them, or
some of them, given that she intended her utterance to be manifestly relevant to the
hearer (1995: 195). With these tools, Relevance Theory explicates the relevant inferencing processes involved in communication. They delimit themselves from the
Gricean paradigm where, from their perspective, the successes of human non-demonstrative inference must be explained by appealing not to logical processes of
assumption confirmation, but to constraints on the formation and exploitation of assumptions (1995: 81), viz. the cooperative principle and the maxims.
Grice (1975: 4344) differentiates between implicate and the related nouns implicature (cf. implying) and implicatum (cf. what is implied). He distinguishes between two basic types of implicature: conventional implicature and conversational
implicature. The latter is subdivided into generalized conversational implicature
and particularized conversational implicature. Generalized conversational implicature is also referred to as default implicature or pragmatic regularities (Bach
2006). This is in line with Levinsons claim that utterance-types carry generalized
implicatures []: rational speakers meannn both what they say (except in non-literal uses of language) and what that saying implicates; different layers of meaning
all come under the umbrella of meaningnn (Levinson 2000: 373). While conventional implicature is connected closely with linguistic form, for instance with connectives (e.g., but), implicative verbs (e.g., manage, forget to), honorifics or nonrestrictive relative clauses, conversational implicatures are essentially connected
43
with certain general features of discourse, viz. the Gricean conversational maxims.
Both conversational and conventional implicatures can be suspended: conventional implicatures are detachable, but not cancellable and conversational implicatures are cancellable but not detachable (for a thorough analysis of implicature, cf.
Huang this volume).
Conversational implicatures are generated or triggered by model users exploiting a maxim. That is to say, a model user gets in a conversational implicature if s/he
flouts a maxim, blatantly failing to fulfil it. Grice (1975: 4950) characterizes the
notion of conversational implicature as follows:
A man [or person, A.F.] who, by (in, when) saying (or making as if to say) that p has implicated that q, may be said to have conversationally implicated that q, PROVIDED THAT
(1) he is to be presumed to be observing the conversational maxims, or at least the cooperative principle; (2) the supposition that he is aware that, or thinks that, q is required
in order to make his saying or making as if to say p (or doing so in THOSE terms) consistent with this presumption; and (3) the speaker thinks (and would expect the hearer to
think that the speaker thinks) that it is within the competence of the hearer to work out,
or grasp intuitively, that the supposition in (2) IS required.
The conversational implicature is characterized by some degree of indeterminateness because it is, by definition, defeasible, non-detachable, calculable and nonconventional (cf. Bach 2006; e.g., Levinson 1979, 1983). Moreover, conversational implicatures are universal and reinforceable (Ariel 2008: 13). That is to say,
they can be asserted explicitly without bringing about a redundancy effect. It needs
to be pointed out, however, that the universality claim only holds for the inferential
mechanism and not for the relevant background assumptions.
Implicature is a cognitive mechanism anchored to model users in their roles as
speakers or producers of conversational contributions, and inference is a cognitive
mechanism anchored to model users in their roles as hearers. Cognitive processes
can be local, they can be global or they can be both. Not only natural model users,
but also non-natural intelligent agents may perform the cognitive operations of implicating and inferring, thus drawing invited inferences. For the operations to be
felicitous, compatible contexts for reasoning and drawing analogies are required
(cf. e.g., Bouquet et al. 1999; Cohen, Morgan and Pollack 1992).
In Gricean pragmatics, the conversational maxims are seen as specifications of
some unmarked communicative context representing the ground on which the conversational implicature is calculated. For this reason, deviations, however common, are seen as special or marked and signify that the speaker intends to communicate conversationally implicated meaning. The signalling is only possible
against the background of the Gricean premise of a reflexive intention, which is
succinctly explicated by Levinson: [w]hat distinguishes a Gricean reflexive intention from other kinds of complex reflexive intention is that the communicators
goal or intention is achieved simply by being perceived: recognition exhausts or
realizes the intention (Levinson 1995: 228).
44
Anita Fetzer
5.
Outlook
Neither general pragmatics nor linguistic pragmatics examine its objects of investigation in isolation but rather focus on their conditions of use, the connectedness
with their surroundings, and the necessary and sufficient conditions which assign
the object, e.g. intentionality, rationality, model user or action, the status of a particular object and make it count as that object. While general pragmatics concentrates on the analysis of these fundamental premises of practical action, identifying
their necessary and sufficient conditions, linguistic pragmatics establishes the explicit connection between those foundations and their language-specific and language-use specific constraints and requirements.
Pragmatics is more of a perspective towards an object under investigation than
the examination of the object as such. For this reason, it needs to touch on and interface with neighbouring disciplines, in particular philosophy, cognitive science,
neuroscience, linguistics and the social sciences. In spite of the diversity of the
field, the key research question of pragmatics is concerned with communicative
action, particularly with the expression and interpretation of meaning in context.
To tackle that question felicitously, pragmatics needs to accommodate extra-linguistic world knowledge, cultural and social stereotypes, and situation of discourse on the one hand, and word meaning, sentence structure and the cognitive
system, especially inference and abduction on the other. The generalized knowledge is stored as typed world knowledge and typed discursive knowledge. These
defaults represent pillars, against the background of which the argument from ignorance holds:
The argument from ignorance, it thus emerges, is an integral part of the inferential
mechanism by means of which implicatures are recovered in conversation. This argument permits us to conclude that conversational principles and maxims are being observed on the grounds that there is no reason to believe that they are not being observed.
Such a conclusion is at all possible because these principles and maxims have the status
of presumptions in communication principles that stand in the absence of counterindications. (Cummings 2005:109)
Presumptions or defaults concern both linguistic and extra-linguistic knowledge,

which need to be stored as social and cultural world-knowledge defaults on the one
hand, and as linguistic and discursive defaults on the other. Only then is it possible
to account for a pragmatic perspective which accommodates the inherent dialogic
and pragmatic principles of speech act and uptake, speech act type and cognitive
reality, and speech act type and situation (mis)match, which is calculated against
the background of the argument of ignorance. If a conversational contribution is in
accordance with the presumptions, no counterindications are found. Consequently,
the contribution is assigned the status of a match and interpreted accordingly. If a
conversational contribution is in disaccordance, there are counterindications and a
particularized process of inferencing is generated to calculate nonce meaning.
45
The dialogue-oriented pragmatic mechanisms of inferencing and abduction

require not only a dynamic framework accounting for the parts-whole connectedness, but also a conscious model user who is accountable for her/his actions (Fetzer
2004). At the same time, s/he makes use of automatic processing and default interpretations which figure as salient and strong interpretative probabilities, unless
counterindications in the context signify a non-default match. Only an integrated
framework, making allowances for methodological compositionality and synthesizing relevant findings (Fetzer 2004; Jaszcolt 2005; Marmaridou 2000) may capture the multifaceted nature of pragmatics.
Notes
1. I am grateful to the anonymous reviewer and the editors of this volume for helpful comments on the first version of this article, and I would like to thank Varol Akman and Teun
van Dijk for discussing previous versions of this chapter.
2. The seemingly clear-cut distinction between semantics and pragmatics, though useful for
analytic reasons, is highly idealized and has been strongly contested (cf. Bublitz 2009).
References
Aijmer, Karin
1997
I think an English modal particle. In: Toril Swan and Olaf Jansen (eds.), Modality in Germanic Languages. Historical and Comparative Perspectives,
Akman, Varol, Paolo Bouquet, Richmond Thomason and Roger A. Young (eds.)
2001
Modeling and Using Context. Third International and Interdisciplinary Conference, CONTEXT 2001, Dundee, UK, July 2001, Proceedings. Springer: Heidelberg.
Ariel, Mira
2008
Pragmatics and Grammar. Cambridge: Cambridge University Press.
1992
Austin, John L.
1971
How to Do Things with Words. Cambridge: Cambridge University Press.
Bach, Kent
1992
Communicative intentions, plan recognition, and pragmatics: comments on
Thomason and Littman and Allen. In: Cohen, Philip R., Jerry Morgan and
Martha E. Pollack (eds.), Intentions in Communication, 389400. Cambridge,
MA: MIT Press.
Bach, Kent
2006
The top 10 misconceptions about implicature. In: Betty Birner and Gregory
Ward (eds.), Drawing the Boundaries of Meaning. Neo-Gricean Studies in
Pragmatics and Semantics in Honour of Laurence R. Horn, 2130. Amsterdam: Benjamins.
46
Anita Fetzer
1971
Out of the pragmatic wastebasket. Linguistic Inquiry 2: 401407.
Bateson, Gregory
1972
Steps to an Ecology of Mind. Toronto: Chandler Publishing Company.
Beaugrande, Robert de and Wolfgang Dressler
1981
Einfhrung in die Textlinguistik. Tbingen: Niemeyer.
Birner, Betty and Gregory Ward
1998
Information Status and Noncanonical Word Order in English. Amsterdam:
Benjamins.
Birner, Betty and Gregory Ward (eds.)
2006
Drawing the Boundaries of Meaning. Neo-Gricean Studies in Pragmatics and
Semantics in Honour of Laurence R. Horn. Amsterdam: Benjamins.
Blackburn, Patrick, Chiara Ghidini, Roy M. Turner, and Fausto Giunchiglia (eds.)
2003
Modeling and Using Context. 4th International and Interdisciplinary Conference, Context 2003, Proceedings. Heidelberg: Springer.
Bouquet, Paolo, Patrick Brezillon, Massimo Benerecetti, Francesca Castellani and Luciano
Serafini (eds.)
1999
Modeling and Using Context. Second International and Interdisciplinary Conference on Modeling and Using Context, Context99, Trento, Italy, September
1999, Proceedings. Heidelberg: Springer.
Brandom, Robert B.
1994
Making it Explicit: Reasoning, Representing, and Discursive Commitment.
Brown, Penelope
1995
Politeness strategies and the attribution of intentions: the case of Tzeltal irony.
In: Esther Goody (ed.), Social Intelligence and Interaction, 153174. Cambridge: Cambridge University Press.
1987
Bublitz, Wolfram
2006
Bublitz, Wolfram
2009
Englische Pragmatik. (2nd rev. ed.). Berlin: Erich Schmidt.
Carston, Robyn
2002
Clark, Herbert H.
1996
Cohen, Philip R., Jerry Morgan and Martha E. Pollack (eds.)
1992
Intentions in Communication. Cambridge, MA: MIT Press
Collins Cobuild Essential English Dictionary
1990
Glasgow and London: Collins.
Cummings, Louise
2005
47
Duranti, Alessandro
1997
Linguistic Anthropology. Cambridge: Cambridge University Press.
Fairclough, Norman
1992
Discourse and Social Change. Cambridge: Polity Press.
Fetzer, Anita and Varol Akman
2002
Contexts of social action: guest editors introduction. Language and Communication 22(4): 391402.
Fetzer, Anita
2002
Fetzer, Anita
2004
Recontextualizing Context: Grammaticality Meets Appropriateness. Amsterdam: Benjamins
Fetzer, Anita
2007
Non-acceptances in context. Intercultural Pragmatics 4(4): 493520.
Garfinkel, Harold
1994
Studies in Ethnomethodology. Cambridge: Polity Press.
Givn, Talmy
1989
Mind, Code and Context. Hillsdale: Laurence Erlbaum.
Givn, Talmy
1993
English Grammar: A Function-Based Introduction. Amsterdam: Benjamins.
Givn, Talmy
2005
Context as Other Minds. Amsterdam: Benjamins.
Goffman, Erving
1981
Forms of Talk. Oxford: Blackwell.
Goffman, Erving
1986
Goodwin, Charles and Alessandro Duranti
1992
Rethinking context: an introduction. In: Alessandro Duranti and Charles
Goodwin (eds.), Rethinking Context. Language as an Interactive Phenomenon,
Grice, Herbert Paul
1975
Grosz, Barbara J. and Candace L. Sidner
1992
Plans for discourse. In: Philip R. Cohen, Jerry Morgan and Martha E. Pollack
(eds.), Intentions in Communication, 417444. Cambridge, MA: MIT Press.
Gumperz, John J.
1992
Gumperz, John J.
1996
The linguistic and cultural relativity of inference. In: John J. Gumperz and
48
Anita Fetzer
Gumperz, John J.
2003
Response essay. In: Susan L. Eerdmans, Carlo L. Prevignano and Paul J. Thibault (eds.), Language and Interaction. Discussions with John J. Gumperz,
Habermas, Jrgen
1998
On the Pragmatics of Communication. Cambridge, MA: MIT Press.
1994
Introduction to Functional Grammar. London: Arnold.
Hanks, William F.
1996
Language form and communicative practices. In: John J. Gumperz and
Heritage, John
1984
Horn, Laurence R.
1984
Toward a new taxonomy for pragmatic inference: Q-based and R-based Implicature. In: Deborah Schiffrin (ed.), Meaning, Form and Use in Context: Linguistic Applications, 1142. Washington: Washington University Press.
Horn, Laurence R.
1989
A Natural History of Negation. Chicago: Chicago University Press.
2004
Huang, Yan
2007
Jaszczolt, Kasia
1999
Discourse, Beliefs, and Intentions: Semantic Defaults and Propositional
Attitude Ascription. Oxford: Elsevier Science.
Jaszczolt, Kasia
2005
Default Semantics: Foundations of a Compositional Theory of Acts of Communication. Oxford: Oxford University Press.
Lakoff, Robin
1973
The logic of politeness: or minding you ps and qs. Papers from the 9th Regional Meeting of the Linguistic Society of Chicago: 292305.
Lambrecht, Knud
1994
Information Structure and Sentence Form. Topic, Focus and the Mental Representation of Discourse Referents. Cambridge: Cambridge University Press.
Leech, Geoffrey
1983
1979
1983
1988
Putting linguistics on a proper footing: explorations in Goffmans concepts of
participation. In: Paul Drew and Anthony Wootton (eds.), Erving Goffman. Exploring the Interaction Order, 161227. Cambridge: Cambridge University
Press.
49
1995
Interactional bias in human thinking. In: Esther Goody (ed.), Social Intelligence and Interaction, 221260. Cambridge: Cambridge University Press.
2000
Lewis, David
1979
Scorekeeping in a language game. Journal of Philosophical Logic 8: 339359.
Longman Dictionary of Contemporary English
2003
Harlow: Pearson Education Limited.
Marmaridou, Sophia
2000
Mey, Jacob
2001
Morris, Charles
1971
Penco, Carlo
1999
Objective and cognitive context. In: Paolo Bouquet et al. (eds.), 2nd International and Interdisciplinary Conference on Modeling and Using Context
(Context99), 270283. Heidelberg: Springer.
Recanati, Franois
2004
Reddy, Michael
1979
The conduit metaphor. In: Andrew Ortony (ed.), Metaphor and Thought,
Sbis, Marina
2002
Speech acts in context. Language and Communication 22(4): 421436.
Schmid, Hans-Jrg
2003
Introduction: context and cognition a concern shared by cognitive sciences.
In: Ewald Mengel, Hans-Jrg Schmid and Michael Steppat (eds.), Anglistentag
2002 Bayreuth. Proceedings, 363367. Trier: Wissenschaftlicher Verlag Trier.
Searle, John R.
1969
Searle, John R.
1975
Searle, John R.
1979
Searle, John R.
1983
Intentionality. Cambridge: Cambridge University Press.
Searle, John R.
1991
Collective Intentions and actions. In: John R. Searle, Herman Parret and Jef
Verschueren (eds.), (On) Searle on Conversation, 401415. Amsterdam: Benjamins.
Searle, John R.
1995
The Construction of Social Reality. New York: The Free Press.
50
Anita Fetzer

1995
Stalnaker, Robert
1999
The New Encyclopaedia Britannica.
1976
The University of Chicago, Vol 14. Chicago: Helen Hemmingway Benton.
Thomason, Richmond H.
1992
Accommodation, meaning, and implicature: interdisciplinary foundations for
pragmatics. In: Philip R. Cohen, Jerry Morgan and Martha E. Pollack (eds.),
Intentions in Communication, 325363. Cambridge: MIT Press.
1988
Vanderveken, Daniel and Susumo Kubo (eds.)
2002
Essays in Speech Act Theory. Amsterdam: Benjamins.
Vanderveken, Daniel and Susumo Kubo
2002
Introduction. In: Daniel Vanderveken and Susumo Kubo (eds.), Essays in
Speech Act Theory, 121. Amsterdam: Benjamins.
Van Dijk, Teun
1981
Studies in the Pragmatics of Discourse. The Hague: Mouton.
Van Dijk, Teun
2008
University Press.
Verschueren, Jef
1999
Watts, Richard, Sachiko Ide and Konrad Ehlich (eds.)
1992
Politeness in Language. Berlin: de Gruyter.
Widdowson, Henry
2004
Text, Context, and Pretext. Critical Issues in Discourse Analysis. Oxford:
Blackwell.
1958
Yule, George
1996
The Study of Language. Cambridge: Cambridge University Press.
2.

Piotr Cap
1.
Introduction
The present chapter discusses micropragmatics (the pragmatics of utterance based

concepts such as deixis, anaphora, presupposition, etc.) and macropragmatics (the
pragmatics of discourse based concepts such as speech events, global intentionality or macro speech acts). Its aim is to show how micropragmatics and macropragmatics (have come to) complement each other, both conceptually and in terms of
the division of labor across the field of contemporary linguistic pragmatics (henceforth: pragmatics).1
After a brief overview of the perspectivist approach to pragmatics in section 2.,
I turn to the discussion of context (section 3.) and discuss how its potential and
actual manifestations affect pragmatic inferences at different levels of the utterance and discourse structures. In 4., the perspectivist approach and the considerations of context are bridged to constitute two crucial factors defining micro- and
macropragmatics, in addition to several other factors and developments that have
led to such a distinction. In section 5., I deal with the four concepts central to the
methodology of pragmatics in general: deixis, presupposition, implicature and
speech acts. I discuss them from a micropragmatic perspective, showing their contribution to the proposition of an utterance, its illocutionary force, and the range
and kinds of effects the utterance is able to bring about. At the same time, however,
I introduce the macropragmatic approach, by raising three observations. The first is
the different potential each of the four concepts possesses to contribute to the description of function and effects of an utterance. The second is the fact that, in the
process of describing an utterance, some of the concepts go together and form
more sophisticated methodological tools, better than others. The third is the different range of utterance contexts each of the concepts is able to account for, which
opens up prospects for their hierarchical application in analysis. In 6., I take these
observations as prompts for an integrated study of intentionality at the macro level
of discourse. Addressing such notions as speech event and macro speech act, I
revisit the potential of individual speech acts for combining interacting discourse
topics into larger functional units. I discuss the conceptual infiniteness pertaining
to speech events in terms of their composition and hierarchy, and show the resulting relativity of the concept of the macro speech act.
52
2.
Piotr Cap
Pragmatics offers a unique, function-based perspective on all aspects of human

(linguistic) communication (cf. Verschueren 1999; Fetzer 2002). It is concerned
with all facets of a communicative act or a series of acts, such as the speaker, his/
her background knowledge and contextual assumptions, the lexical and grammatical constituents of an utterance, the hearers interpretations and patterns of inferencing, etc. All these are studied against a network of social factors, preconditions,
norms and expectations that govern communication, both within a culture and
across cultures. Since communicative acts involve linguistic units, whose choice is
dictated by language-internal rules, as well as their interpersonal, social and cultural embedding, pragmatic studies bridge the system and the use side of language.
They examine what is lexically and grammatically available for a speaker to accomplish a communicative goal, and at the same time explore the ways in which
the linguistic potential is realized in a specific social context.
Since the arsenal of language forms and the catalogue of social embeddings in
which these forms are used in communication are virtually infinite and heterogeneous, pragmatic scholarship naturally draws upon a number of diverse disciplines
in linguistics (and beyond). It is generated collectively by phonologists, morphologists, syntacticians, discourse analysts but also psychologists, sociologists and
other scientists in the Humanities as long as their work has a functional-communicative bent. Pragmatics is thus defined by its perspective more than by a set
number of its objects of investigation.2 It offers explanations which apply to different lexical and structural levels of language word, phrase, sentence, discourse.
3.
Context
The conception of pragmatics as a (functional) perspective foregrounds and endorses its preoccupation with context. Different types of context are naturally salient at different levels of communication (utterance, series of utterances, discourse, genre) but there are also manifestations of different aspects of context at
any specific level. Hence, context is a fundamental concept to account for in describing the micro-macro dialogue in pragmatic studies.
In a nutshell (cf. e.g. Bazzanella 2002), one can speak of a static and a dynamic conception of context. The static view (cf. e.g. Halliday and Hasan 2000;
Cummings 2005) sees context as a set of variables (linguistic, cognitive, social)
that surround strips of text. They constitute an a priori, static infrastructure of resources which are there for interlocutors to determine the meaning of utterances at
hand. On the dynamic view (cf. e.g. Duranti and Goodwin 1992; Bublitz 2003,
2006), context is never given and interlocutors keep creating current contexts for
current utterances the moment a sentence is uttered it becomes part of the envi-
53
ronmental resources on which the contextual interpretation of the following as well

as the preceding utterances has to draw (Bublitz 2003: 383).
As the next section will show in detail, the dynamic view of context is especially
important in elucidating the relationship between micropragmatics and macropragmatics the prerequisite being, of course, the move from a discourse-participant to
a discourse-analytic perspective. The complementarity of micropragmatics (pragmatics of the utterance) and macropragmatics (pragmatics of the discourse) consists in the micro-level analysis getting verified from the macro standpoint, a procedure often leading to the redefinition of the original analytic track. Text and context are, not only in interlocution but also in analysis, constantly matched with and
against each other, both prospectively and retrospectively. Any micropragmatic
analysis of an utterance, reshaped as a result of macropragmatic considerations of
the discourse to which the utterance contributes, is thus an example of reinterpretation from hindsight endorsed by the dynamic view of context. Such a reinterpretation is often indicative of the change in the number and kind of textual cues
which are used before and after the reshaping takes place. It is likely that in the
original micro-pragmatic analysis the reliance on the most prominent textual cues
will be significant, since the macro factors and constraints are, as yet, disregarded.
However, the reshaped micro-analysis may well see some of the initial cues
brushed aside, often for the benefit of the cues originally neglected (see section 4.).
All in all, the dynamic notion of context is a vital concept for micro- and macropragmatics since it allows speculations over, a) the extent to which the interpretation of meaning is guided by text; b) what factors may cause such an interpretation
to be insufficient; c) at which level of analysis such a deficit becomes visible.
The dynamic conception of context recognizes one more important premise for
the existence of the micro-macro dialogue, i.e. the different degrees of accessibility of contextual cues, at different stages of interpretation.3 The linguistic cues
(often referred to as co-text) are immediately accessible, but some of the cognitive and social cues might not be. This, altogether, makes discourse participants abstract from the holistic context (the context which is often defined at the macro
level, according to the characteristics of an attitudinal frame, genre, speech event,
discourse type) the cues that seem actually relevant to a given communicative situation (micro-level). The above process has been extensively described in terms of a
number of methodologies, which account for the relation between context and subcontexts (cf. Goffman 1986), figure and ground (cf. Langacker 1987, 2001), frame
and framing (cf. Bateson 1972; Fetzer 2007), generalized (default, unmarked) context and particularized (marked) context, global context and local context, etc.
Their common assumption is that the concept of context is best investigated from a
parts-whole perspective, which elucidates its function of delimiting content. Such
a perspective is naturally in line with the perspectivist view of pragmatics in general and, at the same time, congruent with the scope of micropragmatics and macropragmatics.
54
4.
Piotr Cap
Micropragmatics and macropragmatics: conceptual origins

and scope
As has been indicated before, micropragmatics can be defined as the study of illocutionary force at the utterance level. In contrast, the focus of macropragmatics is
not on the utterance, but on a series or sequence of utterances which form discourses/texts, seen as bearers of global intentionality of the speaker (i.e. the intentionality resulting from different speech act configurations, often referred to as
speech events) and as initiators of complex effects.
The distinction between micro- and macropragmatics follows from several factors as well as research developments. First, it has to do with flexible definitional
boundaries of pragmatics. If pragmatics involves the study of meaning in context,
then context is not limited in formal ways, such as the length of sentence/utterance
or the amount of its discourse history. Contextual considerations which are used for
determining the function and effect of an utterance may or may not stop within the
boundaries of the language form used. Thus, we need a conceptual handle on the interface between smaller and larger functional units of discourse (e.g. speech act
versus speech event), coded in smaller or larger forms (e.g. utterance versus text).
Second, since context is complex, dynamic and multilayered (cf. section 3.),
the accessibility of contextual cues is a matter of degree (cf. Marmaridou 2000).
The more access to contextual (including co-textual) embedding of an utterance or
a series of utterances, the better profiled the meaning. Hence a tendency on the part
of the analyst to pile up contextual cues and to add to the picture as much of the
prehistory of the discourse situation (cf. Mey 1993) as possible (and feasible).
But since access to contextual cues is each time different and invariably limited by
a dynamic network of constraints, the analyst can hardly ever claim to have exhausted all context to determine the meaning. In this sense, the micro-macro dichotomy involves a relation between the minimal and the maximal contextual
input into analysis, irrespective of its scope, i.e. the length of the language form
under scrutiny.
The domains of micro- and macropragmatics are complementary in terms of
their contribution to analytic labor. There is no micropragmatic analysis that would
not provoke a macropragmatic extension of scope; similarly, there is no macropragmatic study that would not question, retrospectively, its micropragmatic components, thus prompting revision or modification of the original analytic track. Altogether, we arrive at a bottom-top-bottom cycle of upgrades on the explanatory
power of both micropragmatic and macropragmatic concepts. This cycle involves
looking at different aspects of context (linguistic, cognitive, social) from two different perspectives, a more linguistic bottom-up perspective and a more social topdown perspective.
By way of illustration, consider a sample analytic procedure whose objective is
to account for the pragmatics of speeches of the American cold war presidents (say,
55
Harry S. Truman or Dwight D. Eisenhower). Imagine that a micro-level analysis of

utterances making up these speeches points to a large number of deictic, referential
and anaphoric markers, embedded in a multitude of direct speech acts. In contrast,
some other units or phenomena explored in such an analysis, for instance implicatures and presuppositions, turn out to be underrepresented. The next step is to classify these findings under controlling categories, such as speech events or macro
speech acts, to establish the performative denominators and thus postulate global
function(s) of the discourse genre (i.e. the presidential cold war speeches) as a
whole. This macro-analytic task requires inspecting the extralinguistic context,
which informs and complements the linguistic observations. The likely outcome of
the extralinguistic considerations is that the cold war presidents address a world
that is ideologically divided and respond to a need to clearly define to their audience the essence of this divide, which eventually leads them to construct the wellknown us and them opposition. Hence the rhetorical urge for directness and
clarity, rationalizing the opulent use of indexicals or anaphoric expressions and
banning most forms of implicit communication such as implicatures or presuppositions. Reaching this conclusion means that the analyst was able to use micropragmatic considerations as building blocks of a functional hypothesis, which he or she
then tested against contextual factors (including his or her expert knowledge), in
order to generate a macro-function of the genre. This is, roughly, how the analytic
curiosity about regularities governing the salience of some phenomena observed at
an utterance level can provoke the macropragmatic extension of scope. However,
the upgrade cycle does not stop here, there is still room for a reverse procedure,
which has the macropragmatic findings relate to the original component premises.
Accordingly, the analyst may want to return to the bottom level of an utterance to
look for more data, to further strengthen the macro conclusion(s). This may mean
leaving out some of the now useless forms (e.g. markers of implicitness) and
delving more deeply into those which have passed the extralinguistic context
verification. As a result, further micropragmatic activity can follow, with a view to,
either narrowing down the study of the original parameters, e.g. deixis or direct
speech acts, to only those aspects which are in line with the macropragmatic conclusion, or, possibly, identifying more bottom-level forms which are relevant, for
instance more cohesive devices (cf. Halliday and Hasan 1976), relational propositions (cf. Mann and Thompson 1983), etc. Which is where the first round of the
bottom-top-bottom cycle ends, but, of course, more rounds can still take place,
with the micro-macro conclusions refining each other virtually ad infinitum. The
existence of such a dialogue is probably among the most significant methodological arguments for having the two concepts, micropragmatics and macropragmatics,
in the theoretical framework of the field.
The distinction between micro- and macropragmatics has been strongly influenced by work in disciplines which pragmatics addressed while defining its
scope, and upon which it imposed its functional perspective. The ethnography of
56
Piotr Cap
communication (cf. Hymes 1974; Saville-Troike 1989; Gumperz 1992) has provided insights into communicative acts seen as socially situated cultural forms, and
micro-specifications of the more general social and linguistic contexts. Interactional socioliguistics (cf. Gumperz 2003) distinguishes between two levels of inference in analyses of interpretation processes. The higher level accommodates
global inferences of what the exchange is about and what mutual rights and obligations apply, what topics can be brought up and what reply is desired by the cultural macro-convention, and the lower level contains local inferences concerning
what is intended with the speakers own move and what is required by way of a response. Research in corpus linguistics (cf. Biber 1988) has adapted social and institutional factors to the examination of variation in language, both spoken and
written. It is argued (cf. Biber 1988) that conversational and discourse goals comprise both conventional social and institutional goals, as well as personal microgoals. Furthermore, the micro-macro interface has been at the heart of studies in
activity types (cf. Levinson 1979) and communicative projects (cf. Linell 1998).
These studies have paved the way for bridging the micro with the macro on the
plane of the explicit or implicit accommodation of context in interpretative processes. As it seems, the only way in which a distinction between the macro and the
micro domains can be attempted is indeed by looking into the different degrees of
lexical and grammatical coding of context in an utterance or a discourse vis a vis its
extralinguistic coding, i.e. the import of mostly non-indexical cues. Such analytic
endeavors are central to the most recent approaches to the analysis of communicative genres (cf. Bauman 1992; Luckmann 1995; Thibault 2003; Fetzer 2007).
Genre (recall our sample analysis above) is a classificatory concept; it specifies
typical ways in which lexicogrammatical resources are deployed to enact a communicative goal, by matching them against its own delimiting frame. Thus, micro
analysis is an analysis of individualistic phenomena (tokens), which macro analysis verifies against the generic requirements (type). The token contributions are not
merely cumulative, they can get encapsulated in one another or form a hierarchy
as will be shown in the next section.
5.
A micropragmatic perspective on deixis, presupposition, implicature

and speech acts with implications for macropragmatics
While the basic scope of micropragmatics is an utterance, any micropragmatic

analysis is, ultimately, a journey into the macro domain. This follows from the conceptual and methodological characteristics of the most common parameters utilized in description: deixis, presupposition, implicature, speech act. They are never
static components of the analytic enterprise; instead, they prompt an analysis
which is essentially cumulative, interactional and hierarchical. At the bottom of the
hierarchy is the input obtained from the interpretation of the lexical and grammati-
57
cal coding of the utterance (immediate) context. This input (from deixis, presupposition lexical triggers, etc.) is usually insufficient for the utterance interpretation
since it does not include non-textual cues. Accordingly, the analysis proceeds to
the study of inferencing (pragmatic presuppositions, implicatures), which bridges
the linguistic cues with the extralinguistic ones. Finally, the speech act level sees
the function of the utterance emerge from the premises and interpretations (speaker-, hearer-, content-oriented) that have accumulated. Such a function remains,
however, tentative, until a wider discourse context has been addressed for verification (viz. the sample analysis in 4.). Thus, the category of the speech act sits at
the methodological borderline between the micro and the macro domain, which the
latter involves generic, institutional, etc., characteristics and constraints. The
speech act is at the same time an umbrella category for deixis, presupposition, implicature, and, potentially, a component category in the macro considerations. Altogether, micropragmatics feeds sequentially into macropragmatics because the
consecutive-collective use of micropragmatic tools in analysis entails a continual
extension of scope, which eventually straddles the original utterance boundaries.
The four subsections below unfold this sequence, showing, ultimately, the inherent
relativity of the micro-macro distinction.
5.1.
Deixis
The outset of micropragmatic analysis is marked by establishing a tangible, lexicogrammatical architecture of the utterance, and especially the explicit manifestations of the relation between the words used and the context. The analysts task is
to identify the lexicalized pointers which indicate who utters the words to whom
to accomplish what, when and where. This task endorses the status of deixis as the
initiator of the (micro)pragmatic analysis since deixis is the concept that captures
the relationship between the language form and the context in the most evident and
direct manner. It can be viewed as the main phenomenon whereby features of context are encoded in utterances by primarily lexical (e.g. demonstratives) but also
grammatical (e.g. tense) means.4
While the analysis of deixis sets the stage for a more complex scrutiny of the
utterance, its potential to interact with the other parameters of description (presupposition, implicature) is limited. As a concept involving lexicalized or grammaticalized forms of expression, it makes a relatively small contribution to the illocutionary force of the utterance, especially in implicit communication. Naturally,
there are exceptions. For instance, deictic projections (the cues are processed not
from the speakers own perspective, but from the perspective projected on the
hearer, cf. Levinson 1983; Fillmore 1997; Marmaridou 2000; Manning 2001) enable deictic expressions to generate implicatures. If, at 4:01 p.m. on Friday, I send
an e-mail to my publisher which says:
58
Piotr Cap
(1) I will submit the manuscript on Saturday.

I have good reason to expect the publisher (who works until 4 p.m. and only on
weekdays) to read the e-mail no earlier than Monday and, seeing no attachment
containing the ms. in the mailbox, to be puzzled did I mean the Saturday that has
just passed or the one to come? Of course, I retain the rhetorical comfort of denying
whichever interpretation. In fact the comfort starts the moment the Friday e-mail
is sent out I can feel free to submit the manuscript on whichever of the two Saturdays I prefer. There is no way I can be blamed for not submitting the ms. on the
first of the two Saturdays, as I can always respond that I was taking my addressees
perspective (i.e. receiving time), not mine.
Most types of deixis (person, time, place, for an overview cf. Hanks this volume) are analyzed at the utterance level. However, a notable exception is discourse
deixis, which crosses the utterance-discourse boundary, thus becoming a macropragmatic worktool. Discourse deixis involves the use of a lexical item within an
utterance to point to the preceding or following utterances in the same discourse
situation (speech event). The backward and forward reference can be illustrated by
(2) and (3), respectively, where already refers to the earlier stretch of discourse and
here anticipates the upcoming stretch:
(2) As already indicated, all languages possess deictics.
(3) Here goes my argument.
There is no rule to how much textual distance should hold between the deictic expression and its referent. While it is reasonable to expect that the referent of here
will be the immediately following utterance(s), the referent of already is surely not
the closest, preceding utterance, but rather an utterance made much earlier in the
unfolding discourse. This brings us to an important conclusion regarding the pragmatics of discourse deixis. By using a specific number of deictic expressions, the
speaker is able to control the overt connectedness of discourse and, in consequence, its comprehension by the hearer. The presence of deictic markers in utterances which make up a discourse situation where a specific topic is pursued
usually contributes to explicitness and clarity. On the other hand, the speaker may
choose to withhold the use of deixis to purposefully obscure the message.
These exceptions (deictic projections contributing to implicatures; discourse
deixis as controllers of discourse comprehension), however important, do not detract from the core function of deixis in pragmatic studies. Deixis is an essentially
micropragmatic phenomenon which sets analysis of the utterance in motion, but,
by itself, adds little to the characteristics of the speech act(s) residing in it. The
deictic framework of the utterance is sketched to establish the basic, lexically encoded relations between the referent(s) of the utterance and the common ground
of knowledge the utterance assumes to exist between the speaker and the hearer.
Yet, a full account of these relations (and of the utterance function(s) the relations
59
contribute to) needs complementation from concepts that go beyond lexically encoded relations (e.g. presupposition, implicature).
5.2.
Presupposition
Presupposition can be defined as a mechanism whereby the speaker addresses a

body of knowledge and experience, involving both linguistic and non-linguistic
contexts, which he or she assumes to be common to him-/herself and the hearer.
The assumption of the existence of the shared knowledge may cause the speaker
not to grammaticalize (or lexicalize) it in the utterance. This characterization takes
presupposition to be a phenomenon lying at several intersections: the encoded and
the assumed, the semantic and the pragmatic, (or even) the linguistic and the nonlinguistic.5
Presupposition comes in contact with deixis on the plane of its partial anchoring in lexical and structural forms. However, since many instances of presupposition can only be approached with reference to (non-linguistic) context, presupposition also reaches out in the direction of the implicit, constituting, in a sense, a
shared knowledge prerequisite for communicating messages whose final destination is their inference by the hearer. Hence its feasible combination with the apparatus of implicature and, altogether, its relevance to the hierarchy of micropragmatic analysis, which derives its output from both accumulation and interaction of
descriptions offered by the individual conceptual tools. As one of the latter, presupposition targets the communicative act at the stage where it develops upwards
from the lexicogrammatical coding of context to its further abstraction and elaboration by the speaker, with a view to producing a speech act. Throughout this
stage, the speaker decorates the deictic framework of the utterance with instantiations of knowledge shared by the speaker and the hearer with regard to all entities
indexed, referred to, or implied in the utterance.
Traditionally, the more a presupposition was linked with a lexical item or a linguistic construction generating it, the more it was treated as a semantic phenomenon, the other cases deemed pragmatic and worth less attention precisely because of the absence of fixed language forms responsible for enacting particular
presuppositions.6 This view has produced multiple typologies of presupposition,
based on its embedding in lexicogrammatical forms called presupposition triggers.
Furthermore, a number of properties have been assigned to presuppositions, including cancellability (a possibility of denying a presupposition, usually by adding
more content to the utterance where it occurs) and constancy under negation (negating the predication carrying a presupposition does not detach or change the presupposition) (cf. e.g. Beaver 2001).
From the perspective of micropragmatic analysis oriented toward the speech
act characterization of the function of the utterance, as well as the macropragmatic
perspective of the discourse, a rigid distinction between semantic and pragmatic
60
Piotr Cap
presupposition seems far from necessary. It offers little explanatory power compared to an integrated, global view of presupposition as a concept which should be
studied against the utterance and discourse goals it serves. Such a view is quite naturally pragmatic, because even though there are specific lexical items (in fact,
many) associated with specific assumptions (for instance, the assumptions of existence), their descriptive capacity does not expire within the structural boundaries of
the linguistic expression. On the contrary, their significance goes much beyond as
they are able to combine with primarily experiential premises and thus successfully contribute to a network of contextual, often non-linguistic beliefs making up
the entire load of knowledge shared by the speaker and the hearer (see fn. 5).
Most communicative goals served by presupposition (relevant to both the
micro- and macropragmatic analysis) have to do, in one way or another, with economy of expression, though the latter is rarely the only goal sought. If a speaker
could not rely on shared assumptions, the lexical and grammatical load of his/her
utterance would grow in size, potentially obstructing its comprehension. Although
economizing on the linguistic form for the benefit of unobstructed communication
could be an utterance goal in itself, it is often a contribution to a larger utterance or
discourse function. This becomes clear when analyzing, at the macro level, a series
of utterances containing consecutive acts of adjustment to shifting discourse expectations. If I am overweight and say I started jogging after visiting my doctor
and my confession meets with a blatant attack like So it took you going to the doctors to work out like you should, I can always follow a defense line in continuing
Well, to be honest, I tried to do some jogging a few times before, but now I do it
regularly. Since the short form of the initial utterance makes it underdetermined
in terms of meaning, the denial of the only then and never before presupposition
comes rather easy, contributing to the overall explanation and justification (see
Huang 2007 on cancellability of presuppositions).
5.3.
Implicature
The relevance of the analysis of implicature7 to micropragmatics (and, potentially,

macropragmatics) is twofold. First, implicature studies account for the further processing of information that has been encoded by the speaker based on his/her presuppositions. Second, in doing the latter, they eventually recognize the contribution implicature makes to the update of the utterance-discourse context.
From the analytic standpoint, the explanatory powers of presupposition and
implicature are inherently complementary, shedding light on both the speaker and
the hearer side of the speech act formation. They are suited to cover, in combination, the whole process of encoding messages by speakers and decoding them by
their hearers. This process is essentially a continuum, where making a presupposition paves the way for the utterance before it takes on a linguistic form, in which
the presupposition is lexically or non-lexically salient. From that point on, i.e. the
61
point of making the utterance by the speaker, the recovery of the implicature(s) by
the hearer may begin of course, if the hearer senses a prompt to search for it/
them. The inference of the implicature(s), whether in accordance with the speakers expectations or not, finalizes the entire process, thus updating the status of the
interaction and creating a new contextual basis on which to build presuppositions
for further utterances in the exchange. The cycle in question corroborates the dynamic view of context (see section 3.) and endorses the intrinsic relativity of the
micro-macro dichotomy. While the update takes place, technically speaking,
within the utterance, its effect is on the prospective discourse.
Delving deeper, implicatures created within the boundaries of the utterance
(the micro level) are often returned to or readdressed purposefully later on in
the unfolding discourse (the macro level). Since implicature is rarely encoded in
language form,8 it involves a virtually indeterminable number of more or less complex contextual inferences. As such, it constitutes a valuable rhetorical tool
whereby the speaker can control the flow of discourse, adopting his or her consecutive utterances to the current goals. This is due to a central property of implicature,
cancellability (cf. Sadock 1978; Levinson 2000, etc.), which makes it possible for
the speaker to deny, at any moment of speech situation, any implicature he or she
apparently created. We have seen this property as partly relevant to presuppositions, but it is implicature that permits its broadest manifestation. Indeed, a great
many implicatures are cancelled to re-establish adherence to the conversational
norms (see the Cooperative Principle and the Maxims of Conversation by Grice
(1975, 1989) as well as their numerous reformulations and supplements9) but also
to play with the addressee, pull a trick on him/her, or simply annoy him/her. Following on this note, many implicatures are cancelled for ironic or sarcastic effects,
which is well documented in humor studies (cf. Attardo 1990, 1993). Another area
where implicatures and their cancellations surface, is public (especially political)
discourse. Implicatures contribute to the rhetorical safety of public speakers, who,
on the one hand, wish to make statements that are universally acceptable, but on
the other, want to retain the possibility of refining or fine-tuning these statements for hearers who hold different expectations of the meanings conveyed in
them (cf. Cap 2008, 2010). Altogether, the phenomenon of the cancellability of implicature belongs to macropragmatics, since, first, the context that determines the
cancellations is made up of a heterogeneous number of social and institutional factors, second, the distance between implicature and its cancellation is a matter of
discourse, rather than utterance.
5.4.
Speech acts towards macropragmatics
The three brief subsections above have shown that deixis, presupposition and implicature make their distinctive micropragmatic contributions to understanding
how an utterance is built, what its referents are and how they are encoded, what as-
62
Piotr Cap
sumptions are made before the utterance is produced, what effects can be expected
after it has been produced and what inferential processes determine these effects.
They partake in the process of enacting goal(s) of the utterance, from the speakers
intention to realize its envisaged function via application of specific indicators of
force, to the hearers successful (or not) recognition of this function and its results.
A procedure this complex needs a controlling, umbrella parameter of description.
It needs a conceptual tool that is able to cover both speaker and hearer related aspects of the utterance function, and, while doing so, draw upon and thus systematize the particular contributions from deixis, presupposition, and implicature in
order to make them fit for macropragmatic work at the discourse level. The concept
of the speech act seems an excellent theoretical candidate to take up this task.10
The orientation of speech acts to both parties of a verbal exchange, as well as to
its linguistic matter, is visible at a glance from the traditional distinction between
the locutionary, the illocutionary, and the perlocutionary aspects of a speech act.11
While the locutionary aspect is the most objective since it concerns the stable language form of the utterance, the illocutionary and perlocutionary aspects involve a
dynamic negotiation of meaning between the speaker and the hearer. In saying Its
hot in here a speaker may be producing an (implicit) illocutionary act requesting
the hearer to open the window, and the perlocutionary act (effect) might be that the
hearer indeed opens it, but it might also be that he or she turns on the air-conditioning instead. Thus, the illocutionary-perlocutionary relation not only mirrors the
complex process of meaning evolution as sketched at the beginning of this subsection; it also inscribes in the distinction between explicit (direct) and implicit (indirect) ways of communicating a speech act. Consequently, it invokes the notions of
deixis, presupposition, and implicature, since they all situate themselves at some
specific yet different points of the conceptual axis which links what is said with
what is effected.
The classificatory, controlling power of the speech act is further reflected in its
network of felicity conditions, i.e. the conditions that underlie a successful, logical,
felicitous production of different acts.12 For example, a speaker cannot make a
successful order if he or she does not sincerely want the order to be followed, or if
he or she deems the hearer incapable of following it. These two felicity conditions
are excellent illustrations of the connection that holds between the concept of the
speech act and the other micropragmatic concepts a relation we have postulated
at the beginning of this subsection. The speakers awareness of cognitive and social context obtaining at the moment of producing a speech act gives rise to pragmatic presuppositions underlying the utterance that contains this act. Then, once
the act is accomplished, the speakers presuppositions can be assessed against the
effectiveness of implicatures they helped to create. While this part is naturally the
domain of implicature studies, we do get an extra insight from the cover framework of the speech act theory and, specifically, from its contribution to the research
in direct versus indirect acts.
63
Ironically, the greatest advantage of this research is that it almost removes one
of the two groups of acts from our scope of interest, at both the micro and the
macro level of description. As pointed out in several studies, most speech act
usages are indirect (cf. Bertolet 1994; Holdcroft 1994, etc.) and indeed, even the
apparently straightforward act of asking for salt in Can you pass the salt? is, formally speaking, indirect. This, however, should not be discouraging. The orientation of the speech act theory to indirect usages is only reflective of the visible emphasis on indirectness that is salient in the other (micro-)pragmatic concepts; after
all, our discussion of deixis, presupposition, and implicature has been mostly preoccupied with hidden meanings. Thus, speech act theory turns out to be nothing
but congruent with the other apparatuses. Moreover, it extends over all of their
fields of application, from the stage where knowledge is assumed to get encoded in
the utterance, to the stage where the utterance is interpreted. Conceivably, a single
speech act can be realized through recourse to an x number of presuppositions,
an x number of deictic markers in the utterance, which the latter could produce an
x number of implicatures. We have thus arrived at a micropragmatic hierarchy of
analysis, with the speech act constituting a category superordinate over the other
micropragmatic categories. At the same time, however, the speech act should not
be considered the top-most variable of description; in order to account for discourses, rather than individual utterances, we are in need of yet higher-rank concepts.
6.
Macropragmatics
Most speech acts residing in individual utterances tend to combine into larger functional units, thus paralleling the combination of utterances into texts and discourses. This process can be seen from the analysis of (4). Assume that instead of
reading out students names from the attendance list at the beginning of a seminar,
I simply ask:
(4) Hello, are we all here?
In uttering (4), I perform two direct speech acts (greeting and asking), as well as an
indirect act of requesting my students to reveal the names of the absentees. The
speech acts involve a deictic anchoring (e.g., here recognized as classroom), presuppositions (e.g., of some students still missing), and the indirect act involves a
process of inference. This is, with some simplification, where a micropragmatic
analysis of (4) can get us.
However, being myself the producer of (4), I can enjoy the privilege of stating
with absolute certainty that, most of the time, I do not open the class with this utterance for the mere sake of greeting, asking, and requesting, as described above.
That would be unlikely considering that (4) happens regularly and has thus become
64
Piotr Cap
(a part of) expectable routine, a pattern that arises, usually, from some kind of a
larger intention.13 Mark an analogy: If I follow the routine of buying each morning
a newspaper (an action which involves micro-actions analogical to the acts defined
in (4)), the larger intention can be described as a general aspiration to stay updated with the current news. The larger intention in (4) is to begin the seminar,
making sure all things are in place for a productive meeting. This finding does not
sound terribly original in itself, but we will take it as a starting point to show,
eventually, that the nature of intentionality is such that it cannot be a matter of the
individual utterance and thus the scope of inquiry must be extended accordingly.
More often than not, I may need to say more than just (4), to successfully begin
the seminar. I may choose from an apparently infinite catalogue of other utterances
which are able to serve the same intention. Some typical cases may be: Its getting
late, Would you close the doors please?, Now, Adam, listen up, will you?,
Right, shall we begin?, And the chalk is missing as usual!, I cant remember
when we last started on time, etc. All these seemingly diverse utterances contribute, in one way or the other, to realization of the principal intention. Interestingly, they do so while performing different speech acts (assertives, directives, expressives) whose force may also be different, direct or indirect. Consequently, each
of these acts may accommodate a different input from deixis, presupposition, and
implicature.
The routine of performing individual speech acts (which may in turn involve
the other micropragmatic categories) to realize a more general intention, can be defined as a speech event.14 Speech events may contain, as we have seen, a potentially
infinite number of utterances, depending on how many are actually necessary to
carry out the intention. If, for instance, my seminar group enters the classroom discussing a just-finished exam in which they participated, I may be in for a longer
stretch of talk to set up the stage for my own class. Otherwise, it may take a few
short remarks to cover the technicalities and initiate the topic proper. Whatever
happens in actuality is thus dictated, as anything in pragmatics, by context. A
speech event can comprise just as many utterances (and speech acts) as needed to
match the contextual preconditions. These utterances do not have to come in a
monologic pattern. I may keep performing the principal speech event by producing
a few utterances in a row, then pausing, then, possibly, reacting to a question that
comes in the meantime, then resuming the monologue, etc. Thus, my performance
is a genuine discourse performance, which exists in and responds to a social setting, though, of course, some parts of it can mirror textual units and their connectedness patterns.
The concept of infiniteness resurfaces once we go on to discuss speech events
as building blocks for realization of a still larger intention. To readdress the
case initiated by (4), the speech event which realizes the intention to successfully
begin the seminar is not only an umbrella category for a series of individual acts,
but may itself be subordinate to a larger discourse goal be it, for instance, con-
65
ducting a productive seminar meeting as a whole. This goal entails that an apparently infinite number of speech events are carried out on the way, from a speech
event of providing a complex, multi-act explanation to a problem that has arisen
during the session, to a speech event of assigning homework, which may again involve a number of component speech acts. The latter (event) may include: an expressive (e.g., reprimanding students for not completing the previous assignment),
a commissive (e.g., threatening to fail students at the end of the course, if they keep
neglecting their assignments), a directive (e.g., telling students to do a particular
task for the next meeting), an assertive (e.g., describing a rationale for the task),
etc. The variety and diversity of the acts involved is, here, no smaller than in the
case of the speech event performed to begin the seminar.
In this clearly bottom-up fashion, we have approached the problem of the uppermost or global category of intentionality enactment, one whose promise
would be to encompass all the subordinate intentions realized in speech events and
their attendant acts. Such a promise is partly fulfilled by the classical concept of the
macro speech act, which is, in van Dijks words, a global speech act performed by
the utterance of a whole discourse, and executed by a sequence of possibly different speech acts.15 Still, while doing some useful job in the way of systematizing
speech events, the macro speech act suffers from a problem of an inherent relativity as regards the range of its own operation. The definition of the macro act only
corroborates the problem; we do not get to know how much is a whole discourse.
Is it, to return to our example above, the whole body of discourse produced within
the duration of the seminar, controlled by the global intention to make it a productive meeting? Then the relationship between the macro speech act and the component speech events (and their individual acts) seems analytically elegant. But, does
this account exhaust the potential of the macro speech act to combine with further
macro speech acts, to serve a yet-more-global intention? Apparently not. The intention to carry out an academically rewarding seminar can be considered subordinate to the intention to conduct the entire course as planned, which in turn partakes in the intention to perform my (academic) duties properly as a whole, etc.,
etc., which, naturally, makes the consecutive macro acts accumulate accordingly.
To envisage the highest-rank intention, pursued in the highest-rank speech act,
turns, then, a philosophical undertaking.
One of the undisputed methodological values of the search for the clear-cut categories signposting the particular levels of intentionality and its enactment lies,
paradoxically, in the recognition of limits to which intentionality could be accounted for in larger stretches of discourse. Thus, endeavors like the macro speech
act theory should not be carelessly brushed off, as they eventually motivate research, however minimalist, in better demarcated and better empirically equipped
areas. Apart from the aforementioned research in genres, a prominent example of
such an area is Conversation Analysis.16 Conversation analysts have elaborated an
impressive arsenal of techniques for the description of speech act deployment,
66
Piotr Cap
though in necessarily limited contexts, often just co-texts. As a result, the apparatus of macropragmatics has been endowed with a number of relevant concepts,
such as floor, topic, turn, turn-taking, transition relevance place, etc., each
of which contributes its share of insight in how people manage their intentions and
goals in particularized (conversational) settings.
A related and very important significance is that a macro perspective on the
pragmatics of discourse makes the analyst approach intentionality as continually
re-shaped and updated by extralinguistic context and thus work out methodological tools to handle the variability of enactment of the speakers intention(s) over an
extensive period of time. Let us illustrate this phenomenon with an example from
political discourse, a domain rich in speakers attempts to continue with an enactment of a global intention, though often in consecutively modified patterns, responding to the changing extralinguistic reality. Consider an excerpt from one of
G.W. Bushs speeches legitimizing US involvement in the recent Iraq war:
(5) By advancing freedom in the greater Middle East, we help end a cycle of dictatorship and radicalism that brings millions of people to misery and brings
danger to our own people. By struggling for justice in Iraq, Burma, in Sudan,
and in Zimbabwe, we give hope to suffering people and improve the chances
for stability and progress. [] Had we failed to act, the dictators programs for
weapons of mass destruction would continue to this day. For all who love freedom and peace, the world without Saddam Husseins regime is a better and
safer place.17
This speech comes eight months into the war, which started on the assumption that
the Iraqi regime (and its alleged terrorist allies) had developed access to weapons
of mass destruction (WMD), thus becoming a world threat and unwilling to disarm
unless forced to. In his address, Bush attempts to maintain the aura of legitimization of the US intervention, against the increasingly evident collapse of the original premise no weapons of mass destruction were found in Iraq ever since the
US and the coalition troops entered it in March 2003. Since he can no longer employ the single WMD argument to execute his global intention (to keep the legitimization process intact), he deftly switches to a more universal, ideological
rationale. Instead of invoking a direct threat from a destructive impact of WMD, he
concentrates on building up an ideological framework for a potential growth of
such a threat in the future and presents the US strike as part of the necessary means
to ensure that the antagonistic ideologies (dictatorship, radicalism, regime)
do not materialize in the form of physical impact. The localization of these ideologies in more countries than Iraq itself (viz. Burma, Sudan, Zimbabwe) is
in fact a clever rhetorical ploy. By extending the spectrum of the ideological conflict and, thus, the spectrum of US foreign involvement, he encourages the construal of the WMD intelligence failure in terms of an isolated incident, fully justifiable given the range of the American mission as a whole. Altogether, Bush
67
keeps up with the global intention to continue with legitimization of the Iraq war,
now on ideological grounds, but the pragmatic patterns (and, consequently, lexical
choices) responsible for realizing this intention within a macro-temporal reality
need to be updated to meet the extralinguistic developments.
In section 4., I argued that macropragmatic analyses often lead to redefinitions
of the micropragmatic input that has been utilized to build up their tentative versions. They also tend to review the micropragmatic concepts and point to those
whose descriptive potential has not yet been fully exhausted. The analysis of
Bushs speech is a good case in point. At its macro-level, we consider the general
patterns of adaptation of the political speaker to contextual conditions. This leads
us to the more focused question which of the micro-level concepts are able to account for such processes in the possibly richest lexical way, i.e. which of them are
most frequently reflected in lexical items and structures that eventually make up
the matter of the macroanalysis. The analysis of (5) reveals, for instance, that much
explanatory power is yet to be drawn from a careful scrutiny of implicature forms
in the text. If we gather from macro-contextual considerations that Bush is forced
to switch to a new argumentative strategy, yet, as seems logical, without discrediting the previous argumentation, what better way to accomplish the goal than
through implied meanings, which are always subject to cancellation as the speaker
sees fit? Recall the phrase programs for weapons of mass destruction. It is flexible enough to concede that Saddam did not indeed possess WMD understood as a
product ready for use, at the outset of the war, but it does not completely detract
from the original assumption that he did. It is quite likely that a microanalysis of
this implicature (as well as of many other bottom-level forms in (5)) could be overlooked but for the prompt from the macropragmatic approach, involving a vast
range of contextual factors affecting the entire discourse of the Iraq war.
In this way we have returned to the question of the micro-macro dialogue,
which I defined in the fourth section of this chapter as one of the most significant
methodological arguments for having the [distinction between the] two concepts,
micropragmatics and macropragmatics, in the theoretical framework of the field.
It should be said in closing that prompts for such a dialogue need not come from
pragmatic analyses alone; in fact, the dialogic relationship between micropragmatics and macropragmatics is kept alive by insights from approaches which, at
least definitionally, go much beyond linguistics-based disciplines or perspectives.
An example of such an approach is Critical Discourse Analysis (CDA; cf. e.g. Fairclough 1995; Wodak and Chilton 2005), which extends onto domains traditionally
associated with various branches of sociology and social psychology. Rooted in the
conception that discourse is just as much reflective of the existing reality as it is potentially constitutive of a new reality, CDA offers a number of useful ways for the
analysis of the reality (which pragmatics would rather term extralinguistic aspects of context). In so doing, it provides (macro-)pragmatic considerations with
socio-cognitive grounding of discourse (cf. van Dijk 1995, 2002), thus signposting
68
Piotr Cap
the downward, top-down analytic procedure. Two examples of how such a procedure could operate have been, in this article, the suggested refinements and extensions of the analysis of deixis and direct speech acts in the genre of the cold war
presidents speeches, as well as the just discussed hint for a fully fledged study of
implicature in (5). Of course, CDA has much more to offer, as does any approach
or discipline that recognizes the paramount importance of (holistic, dynamic) context in human (linguistic) behavior. In whatever kind of acting, including pragmatic acting, we adapt ourselves to a context as is, but at the same time we
change it with a view to future effects. As Mey (1993: 263) writes, all this is done
through the intermediate use of language as the tool helping us to select the relevant features of any situation in the total context. Micropragmatics and macropragmatics are two perspectives which, on the one hand, statically account for the
different size manifestations of this total, or holistic, context. On the other
hand, they dynamically inform each other (and thus the pragmatic analysis as a
whole) of, first, which instances of the utterance-level use of language are worth a
broader contextual verification to determine their function, second, which parts of
the broader contextual check are applicable as determiners of further analytic activity at the utterance level.
Notes
1. See Fetzer (this volume) for the distinction between general (non-linguistic) pragmatics
and linguistic pragmatics.
2. See Fetzer (this volume) or Huang (2007) for a discussion of perspectivist versus componential views of pragmatics.
3. Cognitive pragmatics (e.g. Sperber and Wilson 1986) illustrates this by the onion metaphor. The pattern of inclusion of the individual layers, which are interconnected, reflects
their order of accessibility in the process of inferencing.
4. Foundations of research in deixis are presented in Bhler (1934/1982). Among the most
useful modern overviews are Nunberg (1993), Fillmore (1997), and Levinson (2004); cf.
also Hanks (this volume).
5. The standard view of presupposition (see Huang this volume) is that (i) it involves a proposition whose truth is taken for granted in the utterance of a sentence and remains in force
even if the sentence is negated, and that (ii) it is engendered by certain lexical and structural triggers. This definition goes back to philosophical and semantic views of Frege
(1892) and Strawson (1952), later enriched with semantic-pragmatic analyses (cf. Karttunen 1973; Stalnaker 1973; Gazdar 1979). Though these characterizations of presupposition are still generally accepted (cf. Saeed 2009), many linguists have pointed to the
existence of counter-examples deeming the mainstream approach (necessarily) imperfect (cf. Huang this volume). In the present chapter, presupposition is treated as a phenomenon which does not rely for its existence on linguistic form alone, but on a complex
relation between the linguistic form, the speakers and the hearers (common) understanding of the world (cf. Grundy 1995: 78), as well as the speakers fulfillment of
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
69
felicity conditions for performance of an act involving presupposition(s). This stance

integrates the standard view with several other research perspectives: the speech act approach to presupposition (cf. Grice 1981; Green 1989), the cognitive approach (cf. Carston 1996), the experiential approach (cf. Marmaridou 2000), or the common ground
approach (cf. Grundy 1995).
Examples of battles over the status of presupposition are Stalnaker (1973), Soames
(1989); while Horn (1996), Atlas (2004) and Huang (2007) offer well-balanced overviews.
In the present chapter the use of the term implicature is synonymous with conversational implicature. See Huang (this volume) for a comprehensive account of various
types of implicature.
We do not deal with conventional implicatures here (cf. e.g. Davis 1998; Bach 1999).
See reductionist attempts such as Horns (1984) Q-R model or Levinsons (2000) Q-I-M
principles; for supplements and general overviews see Levinson (2000), Horn (2007,
2009), Huang (2009a, b and this volume).
See Austin (1962/1975) and Searle (1975, 1979) for the foundations of the Speech Act
Theory. Tsohatzidis (1994), Thomas (1995), Marmaridou (2000), and Sadock (2004)
are excellent, objective overviews of both the classical ideas and later developments.
See also Collavin (this volume).
See Austin (1962/1975) and a review of Austins ideas in e.g. Alston (1994).
Cf. Austins (1962/1975) account of felicity conditions; see also Searle (1979).
Despite a temptation to the contrary, I save the term global for the discussion of macro
acts. I refer to this intention as larger not just to say that it is more indirect than the
intentions to greet, grasp students attention, or work out the number of absentees. My
point is that all of the latter can accumulate into a bigger intentional schema and that
there is no limit to the number of micro-intentional components making up such a
schema. Hence the use of the comparative larger, which allows for a downward
specification (i.e. the specification of what the larger intention is supposed to comprise), as well as an upward specification (i.e. the specification of what higher or highest level of intentionality it could contribute to). Interestingly, van Dijks (1980)
speech event approach to intentionality levels, as well as the activity type and genre
oriented approaches to intentions and context (cf. Levinson 1979; Bauman 1992; Luckmann 1995; Thibault 2003; discussed in section 4), are far from elucidating this bi-directional possibility, which is an important element of the micro-macro interface.
See van Dijk (1980) and Mey (1993). Levinson (1979) uses the term activity type. The
most frequent, however, is the term genre (cf. fn. 13). Recall (section 4) that [genre]
specifies [multiple] typical ways in which lexicogrammatical resources are deployed to
enact a communicative goal, so even such short one-utterance forms as (4) apply, since
the ways in which they could be supported by other forms (recall from the above the potentially infinite catalogue of expressions such as its getting late!) are controlled by
common, conventional, and thus generic constraints. Unlike speech events, which combine into macro speech acts (cf. van Dijk 1980), genres do not presuppose a methodologically superordinate concept. Thus, they constitute a more elegant framework in
terms of their downward specification, but fall short of describing the infiniteness of
consecutive layers in the whole intentionality structure.
See mainly van Dijks work, from classical contributions to text grammar (cf. van Dijk
1977, 1980), to recent research in critical discourse analysis (cf. e.g. van Dijk 2008).
70
Piotr Cap
See also his joint work with Walter Kintsch (e.g. van Dijk and Kintsch 1983), which in
my view represents the best step ever taken to integrate the concepts of discourse and
text in a pragmatic analysis.
16. Its protagonists include Emanuel Schegloff, Harvey Sacks, Gail Jefferson, Jonathan
Potter; cf. Ten Have (2007) for an overview.
17. The Whitehall Palace address, November 19, 2003. The analysis of (5) shows that macropragmatic studies often proceed in a top-down manner (cf. section 4); they draw on
social and cognitive context accessible to the analyst and employ the analysts expert
knowledge to an a priori formulation of the functional thesis. The source of the analysts expert knowledge is not only his/her cognitive experience, but also the data provided by disciplines outside the principal domain of linguistics. Thus, macropragmatics
fosters interdisciplinarity, a notable example of which is Critical Discourse Analysis.
References
Alston, William
1994
Illocutionary acts and linguistic meaning. In: Savas Tsohatzidis (ed.), Foundations of Speech Act Theory: Philosophical and Linguistic Perspectives,
Atlas, Jay
2004
Presupposition. In: Laurence Horn and Gregory Ward (eds.), The Handbook of
Attardo, Salvatore
1990
The violation of Grices maxims in jokes. In: Kira Hall, Michael Meacham and
Richard Shapiro (eds.), Proceedings of the 16th Annual Meeting of the Berkeley
Linguistics Society, 355362. Berkeley, CA: University of California Press.
Attardo, Salvatore
1993
Violation in conversational maxims and cooperation: The case of jokes. Journal of Pragmatics 19: 532558.
Austin, John
1962/1975 How to Do Things with Words. Oxford: Oxford University Press.
Bach, Kent
1999
The myth of conventional implicature. Linguistics and Philosophy 22:
327366.
Bateson, Gregory
1972
Steps of an Ecology of Mind. Toronto: Chandler Publishing Company.
Bauman, Richard
1992
Contextualization, tradition, and the dialogue of genres: Icelandic legends of
the kraftskald. In: Alessandro Duranti and Charles Goodwin (eds.), Rethinking
Context: Language as an Interactive Phenomenon, 125146. Cambridge:
Bazzanella, Carla
2002
The significance of context in interpretation: The we case. In: Bruce Edmonds
and Varol Akman (eds.), Context in Context. Special issue of Foundations of
Science 7, 239254.
71
Beaver, David
2001
Presupposition and Assertion in Dynamic Sentences. Stanford: CSLI.
Bertolet, Rod
1994
Are there indirect speech acts? In: Savas Tsohatzidis (ed.), Foundations of
Speech Act Theory: Philosophical and Linguistic Perspectives, 335349. London: Routledge.
Bezuidenhout, Anne
1997
Pragmatics, semantic underdetermination and the referential/attributive distinction. Mind 106: 375409.
Bezuidenhout, Anne
2002
Truth-conditional pragmatics. Philosophical Perspectives 16: 105134.
Biber, Douglas
1988
Press.
Blakemore, Diane
1992
Understanding Utterances: An Introduction to Pragmatics. Oxford: Blackwell.
Bublitz, Wolfram
2003
Bublitz, Wolfram
2006
Bhler, Karl
1934/1982 Sprachtheorie. Die Darstellungsfunktion der Sprache. Stuttgart/New York:
Fischer.
Cap, Piotr
2008
Towards the proximization model of the analysis of legitimization in political
discourse. Journal of Pragmatics 40: 1741.
Cap, Piotr
2010
Axiological aspects of proximization. Journal of Pragmatics 42: 392407.
Carston, Robyn
1996
Metalinguistic negation and echoic use. Journal of Pragmatics 25: 309330.
Carston, Robyn
2002
Cole, Peter (ed.)
1978
Syntax and Semantics 9: Pragmatics. London: Academic Press.
Cummings, Louise
2005
Davis, Wayne
1998
Implicature: Intention, Convention, and Principle in the Failure of Gricean
Theory. Cambridge: Cambridge University Press.
72
Piotr Cap

1992
Rethinking Context. Language as an Interactive Phenomenon. Cambridge:
Elugardo, Reinaldo and Robert Stainton
2004
Shorthand, syntactic ellipsis, and the pragmatic determinants of what is said.
Mind and Language 19: 44271.
Fairclough, Norman
1995
Critical Discourse Analysis: The Critical Study of Language. London: Longman.
Fetzer, Anita
2002
Fetzer, Anita (ed.)
2007
Context and Appropriateness: Micro Meets Macro. Amsterdam: Benjamins.
Fillmore, Charles
1997
Lectures on Deixis. Stanford: CSLI.
Frege, Gottlob
1892
ber Sinn und Bedeutung. Zeitschrift fr Philosophie und Philosophische Kritik 100: 2550. Trans. as On sense and reference. In: Peter Geach and Max
Black (eds.), 1952 Translations from the Philosophical Writings of Gottlob
Frege. Oxford: Blackwell.
Gazdar, Gerald
1979
Goffman, Erving
1986
Green, Georgia
1989
Erlbaum.
Grice, Paul
1975
Semantics 3: Speech Acts, 4158. New York: Academic Press.
Grice, Paul
1981
Presupposition and conversational implicature. In: Peter Cole (ed.), Radical
Grice, Paul
1989
Grundy, Peter
1995
Doing Pragmatics. London: Edward Arnold.
Gumperz, John
1992
Gumperz, John
2003
Response essay. In: Susan Eerdmans, Carlo Prevignano and Paul Thibault
(eds.), Language and Interaction. Discussions with John J. Gumperz, 105
73

1976
2000
System and text: making links. Text 20: 201210.
Holdcroft, David
1994
Indirect speech acts and propositional content. In: Savas Tsohatzidis (ed.),
Foundations of Speech Act Theory: Philosophical and Linguistic Perspectives,
Horn, Laurence
1984
Toward a new taxonomy for pragmatic inference: Q-based and R-based implicature. In: Deborah Schiffrin (ed.), Meaning, Form, and Use in Context: Linguistic Applications, 1142. Washington, DC: Georgetown University Press.
Horn, Laurence
1996
Presupposition and implicature. In: Shalom Lappin (ed.), The Handbook of
Contemporary Semantic Theory, 299320. Oxford: Blackwell.
Horn, Laurence and Gregory Ward (eds.)
2004
Horn, Laurence
2007
Istvan Kecskes and Laurence Horn (eds.), Explorations in Pragmatics: Linguistic, Cognitive and Intercultural Aspects, 3969. Berlin: Mouton de
Gruyter.
Horn, Laurence
2009
1: 334.
Huang, Yan
2007
Huang, Yan
2009a
1: 118153.
Huang, Yan
2009b
Hymes, Dell
1974
University of Philadelphia Press.
Karttunen, Lauri
1973
1999
Philosophy in the Flesh: The Embodied Mind and its Challenge to Western
Thought. New York: Basic Books.
Langacker, Ronald
1987
Foundations of Cognitive Grammar. Vol. I: Theoretical Prerequisites. Stanford, CA: Stanford University Press.
Langacker, Ronald
2001
Discourse in Cognitive Grammar. Cognitive Linguistics 12: 143188.
74
Piotr Cap
Levinson, Stephen
1979
Levinson, Stephen
1983
Levinson, Stephen
2000
Linell, Per
1998
Approaching Dialogue. Amsterdam: Benjamins.
Luckmann, Thomas
1995
Interaction planning and intersubjective adjustment of perspectives by communicative genres. In: Esther Goody (ed.), Social Intelligence and Interaction:
Expressions and Implications of the Social Bias in Human Intelligence, 175
Lyons, John
1995
Linguistic Semantics. Cambridge: Cambridge University Press.
Mann, William and Sandra Thompson
1983
Relational propositions in discourse. Technical Report ISI/RR-83115. University of Southern California: Information Sciences Institute.
Manning, Paul
2001
On social deixis. Anthropological Linguistics 43: 54100.
Marmaridou, Sophia
2000
Mey, Jacob L.
1993
Pragmatics. Oxford: Blackwell.
Nunberg, Geoffrey
1993
Indexicality and deixis. Linguistics and Philosophy 16: 144.
Recanati, Franois
1993
Recanati, Franois
2001
What is said. Synthese 128: 7591.
Recanati, Franois
2004
Saeed, John
2009
Sadock, Jerrold
1978
Semantics 9: Pragmatics, 281298. London: Academic Press.
Sadock, Jerrold
2004
Speech acts. In: Laurence Horn and Gregory Ward (eds.), The Handbook of
Saville-Troike, Muriel
1989
The Ethnography of Communication. Oxford: Blackwell.
Searle, John
1975
A taxonomy of speech acts. In: Keith Gunderson (ed.), Minnesota Studies in
the Philosophy of Science 9: 344369.
75
Searle, John
1979
Soames, Scott
1989
Presupposition. In: Dov Gabbay and Franz Guenthner (eds.), Handbook of
Philosophical Logic, 553616. Dordrecht: Reidel.
1986
Stalnaker, Robert
1973
Presuppositions. Journal of Philosophical Logic 2: 447457.
Ten Have, Paul
2007
Doing Conversation Analysis. A Practical Guide. Thousand Oaks: Sage.
Strawson, Peter
1952
Thibault, Paul
2003
Contextualization and social meaning-making practices. In: Susan Eerdmans,
Carlo Prevignano and Paul Thibault (eds.), Language and Interaction. Discussions with John J. Gumperz, 4162. Amsterdam: Benjamins.
Thomas, Jenny
1995
Tsohatzidis, Savas (ed.)
1994
Foundations of Speech Act Theory: Philosophical and Linguistic Perspectives.
London: Routledge.
Van Dijk, Teun
1977
Text and Context. London: Longman.
Van Dijk, Teun
1980
Macrostructures. An Interdisciplinary Study of Global Structures in Discourse,
Interaction, and Cognition. Hillsdale, NJ: Erlbaum.
Van Dijk, Teun
1995
Discourse, semantics and ideology. Discourse and Society 6: 243259.
Van Dijk, Teun
2002
Ideology: Political discourse and cognition. In: Paul Chilton and Christina
Schaeffner (eds.), Politics as Text and Talk, 3357. Amsterdam: Benjamins.
Van Dijk, Teun
2008
University Press.
Van Dijk, Teun and Walter Kintsch
1983
Strategies of Discourse Comprehension. New York: Academic Press,
Verschueren, Jef
1999
Understanding Pragmatics. London: Edward Arnold.
Wodak, Ruth and Paul Chilton (eds.)
2005
A New Agenda in (Critical) Discourse Analysis: Theory, Methodology and Interdisciplinarity. Amsterdam: Benjamins.
3.

Sophia Marmaridou
1.
Introduction
Within a broad definition of pragmatics as the scientific study of all aspects of linguistic behaviour, the distinction between pragmalinguistics and sociopragmatics
is intended to focus on two methodological approaches to pragmatic analysis.
Pragmalinguistics typically concerns the study of the particular resources that a
given language provides for conveying pragmatic meaning (illocutionary and interpersonal), whereas sociopragmatics relates pragmatic meaning to an assessment
of participants social distance, the language communitys social rules and appropriateness norms, discourse practices, and accepted behaviours. This methodological distinction, initially launched by Leech (1983) and Thomas (1983) and widely
adopted in subsequent work in pragmatics, is distinct from, though not totally unrelated to, earlier work in Marxist pragmatics, in which the term pragmalinguistics
essentially marks the pragmatic turn in linguistic analysis and is broadly associated, or even identified, with sociolinguistics, stylistics, or text linguistics (see
Mey 1979; Prucha 1983). Moreover, the term sociopragmatics has been used in
contrast to psychopragmatics (Dascal and Franozo 1989), also independently
from the previous distinction just mentioned.
In this paper I shall argue that, as defined above, the distinction between pragmalinguistics and sociopragmatics suggests a conceptual dichotomy in our understanding of language either as a system used for communication purposes, or as an
action domain reflecting, or co-constructing, social order in culturally sanctioned
ways. Regardless of whether this conceptual dichotomy is systematically reflected
in the methodological distinction it motivates, the latter has been particularly productive. In fact, it has both affected the internal organisation and theoretical refinement of other concepts, such as pragmatic failure and pragmatic transfer, and also
informed particular methodological approaches to second language teaching and
testing. Moreover, it has been of particular relevance to historical pragmatics and
historical corpus linguistics. The apparent usefulness of the distinction between
pragmalinguistics and sociopragmatics seems to be at odds with the fact that the
borderline between the two is often fuzzy, to the extent that they are often thought
of as the two end points of a continuum. Given the variety of studies straddling different points along this continuum, I propose to use the figure/ground schema afforded us by psychology to further argue that the two concepts under investigation
essentially reflect a methodological choice in current practices to focus on one area
of pragmatic research by conferring background status to another. The use of the
78
Sophia Marmaridou
figure/ground relationship in this context seeks to explain the bibliographically attested fluidity of the content of these concepts and to present them as distinct, but
interacting, instances of perspectivisation.
It has become obvious from the above that the discussion of the concepts
underlying the terms under investigation will adopt a historical approach to their
analysis. This is expected on the one hand to place pragmalinguistics and sociopragmatics in the wider framework of pragmatic study that was flourishing in the
two crucial decades of its theoretical formation, and, on the other, to show how
these concepts have gradually permeated more recent trends in pragmatic study.
In the next section reference will be made to the ways in which the terms pragmalinguistics and sociopragmatics have been used independently of each other and
prior to the methodological distinction they have been mainly associated with as
introduced by Leech (1983) and Thomas (1983). In particular, the term pragmalinguistics as used in Marxist linguistics in the late seventies and early eighties will
be shown to be broadly equivalent to pragmatic linguistics or societal pragmatics
(Mey 1993). In this framework, the term covers research which emphasises functional aspects of language use within various paradigms (Marxist, or otherwise).
Moreover, sociopragmatics, as a theory of the ways in which the non-linguistic environment affects language use, will be contrasted with psychopragmatics as a theory of how the linguistic environment affects thought (Dascal and Franozo 1989).
In the third section the distinction between pragmalinguistics and sociopragmatics
will be discussed from the perspective of Leech (1983) and Thomas (1983). More
specifically, the conceptual underpinnings of this distinction will be set off against
established research practices concerning issues of contrastive linguistics, pragmatic failure, pragmatic transfer, pragmatic development, and the teaching and
testing of pragmatics in second language learning settings. In the fourth section the
contribution of the pragmalinguistic and sociopragmatic perspective to historical
pragmatics and historical corpus linguistics will be examined with a view to highlighting more recent trends in pragmatic research. In the last section of this paper
an overall assessment of the concepts of pragmalinguistics and sociopragmatics
will be attempted and their distinction will be evaluated with respect to the fields of
study in which it has been adopted.
2.
The borders between pragmalinguistics and sociopragmatics
In order to appreciate the conceptual dichotomy encoded in the terms pragmalinguistics and sociopragmatics, one should be reminded of the two main trends that
have dominated the field of pragmatics ever since its inception as a distinct, established field of linguistic enquiry. In the Anglo-American tradition, pragmatics is
mainly concerned with speakers meaning as intentionally communicated to an addressee. As Jucker (2008: 894) observes, pragmatics in this tradition is a rescue
79
plan for syntax and semantics that deals with a systematic account of implicit
meanings and deals with intentional human action. Levinson (1983: 732), while
espousing this tradition, has identified, discussed, and amply demonstrated problems of defining pragmatics in this narrow sense, to conclude that the best way to
delimit this field of study is to actually observe what practitioners do, or provide a
list of the phenomena which a pragmatic theory must account for, namely deixis,
implicature, presupposition, speech acts, and aspects of discourse structure (Levinson 1983: 27). He also points out that the upper bound of pragmatics is provided
by the borders of semantics and the lower bound by sociolinguistics (and perhaps
psycholinguistics too), and admits that drawing a boundary between sociolinguistic and pragmatic phenomena is likely to be an exceedingly difficult enterprise
(Levinson 1983: 29).
Within the tradition of Levinsons topical definition of pragmatics, the term
pragmalinguistics has been used in the area of literary linguistics, or stylistics, to
refer to the study of Gricean maxims, speech acts and politeness phenomena in a
literary work. An example is Haverkates (1994) study of the dialogues of Don
Quixote de la Macha in Cervantess well known novel. Apparently, this pragmalinguistic frame of reference provides new insights into the interactional roles and
personality traits of Don Quixote and Sancho Panza, which traditional literary theory does not typically address.
The Anglo-American perspective on pragmatics has been criticised as leaving
out, or backgrounding, the larger social context of communication, which is of
central importance in the European tradition (Jucker 2008: 895). The latter derives more directly from Morriss (1938) definition of semiotics, distinguishing between syntactics as the formal relation of signs to one another, semantics as the relation of signs to the objects to which the signs are applicable, and pragmatics as
the relation of signs to interpreters. In Morriss account, the user/speaker is seen in
a wider perspective, since the human being, as an interpreter of signs, does not only
have a mental identity, but also a biological and social identity, which affects her
interpretation of signs (Marmaridou 2000: 18). In fact, Morris acknowledges the
dependence of semiosis on the human being as an entity that is simultaneously the
source and the subject of all psychological, biological and sociological phenomena
that have to do with the functioning of signs comprising language use:
It is a sufficiently accurate characterization of pragmatics to say that it deals with the
biotic aspects of semiosis, that is, with all the psychological, biological, and sociological phenomena which occur in the functioning of signs. (Morris 1938: 108)
Within the context of these two traditions in pragmatics, the terms pragmalinguistics and sociopragmatics have both been used to refer broadly to pragmatic phenomena that characterise the use of language for communication purposes. In what
follows, it will be shown that, while pragmalinguistics has been associated with a
societal and Marxist perspective on language use, sociopragmatics generally refers
80
Sophia Marmaridou
to the study of the external pragmatic factors affecting language use. In this sense,
sociopragmatics has also been juxtaposed to psychopragmatics, the latter focusing
on internal, cognitive operations such as reasoning and problem solving. It should
be pointed out that pragmalinguistics and sociopragmatics, as defined above, constitute two different and unrelated projects. Therefore they do not reflect the distinction and complementariness they have been associated with in the work of
other scholars, such as Leech (1983) and Thomas (1983) (see section 3). Rather,
the use of these two terms in the frameworks discussed below is indicative of the
various developments of pragmatics within the two decades of its initial expansion.
Focusing on sociological phenomena, Mey (1979) uses the term pragmalinguistics to redefine the object and goal of linguistic enquiry in a Marxist framework
as the study of language (the linguistic object) in a variety of pragmatic settings. In
his analysis he identifies pragmatics with the description of action in general and
pragmatic linguistics (or pragmalinguistics) with the study of linguistic action as
determined by the social conditions for speaking and understanding. As he claims,
from a pragmatic viewpoint, the conditions for speaking and understanding, for
production and consumption of language, cannot be divorced from the conditions of production and consumption in the society in general (Mey 1979: 11).
The totality of conditions that are active in the production and consumption of
texts, including co-text, constitute the context, which is dynamic, rather than static,
and creative, rather than passive.
Following Meys understanding of pragmalinguistics, Olsen (1979) raises the
issue of linguistic meaning as located in, and defined by, language use in its social
context. As he observes,
the meanings of words and utterances were not universals which merely were differently labelled in different languages. The meanings were dependent on, and part of, the
different social systems []. Utterances are produced and understood within the definition of the situation and it is the rules by which we define the situation that are shared
by the members of a social system (Olsen 1979: 247).1
That linguistic meaning is a function of language use and that it is being constructed all the time by the process of interaction are facts that, according to Olsen,
must be accounted for by pragmatic linguistics. It transpires that, in his analysis,
pragmalinguistics is to be broadly understood as the pragmatic turn, or a pragmatic
perspective on linguistics as informed by social theory.
As already mentioned, pragmalinguistics is also understood to generally cover
the study of functional aspects of language use as practised in Eastern Europe in
the late seventies. The relevant studies spread over the areas of functional stylistics, text linguistics, rhetorics, psycholinguistics, sociolinguistics, social communication theory, and semiotics (see Prucha 1983). Within such a broad framework, pragmalinguistics emerges as a theory of the pragmatic component of a
grammar and as a theory of verbal activity in social intercourse. These two per-
81
spectives are reflected in the analysis of the meaning of utterances as comprising

contextual and situational aspects respectively, as in Bang and Drs (1979) work.
Contextual meaning is determined by the lexicon and syntax, style and genre, and
textual implications, and it is to a high degree reproducible and general. Situational
meaning expresses an interpretation of the world, society and the human being, and
carries an evaluative orientation. To accept a situational meaning is to subscribe to
a certain set of values. In this understanding, contextual meaning belongs to the
pragmatic component of grammar, whereas situational meaning develops from
verbal activity in social intercourse. These two kinds of utterance meaning are said
to have a dialectal relation to one another and to define a continuum ranging from
the unique and unrepeatable (situational) to the general and reproducible (contextual) (Bang and Dr 1979: 46).
It transpires from the above that pragmalinguistics in the Eastern European and
Marxist traditions encompasses aspects of language use which range from grammatically and textually encoded elements to socially constructed discursive practices. The emphasis is typically placed on the latter, which figures prominently in
this tradition against the background of the grammatical and textual aspects of language. This theoretical turn is to be interpreted both as a reaction to explicitly formal accounts of language in the various structural approaches of the time (see, for
example, the work of Chomsky), and as an attempt to articulate an ideological
stance towards the epistemological foundations of linguistics. The realisation that
utterance meaning, as the domain proper of pragmalinguistic investigation, is a
function of linguistic means in interaction with, or in a dialectal relation to, the social conditions of language use creates a conceptual frame for what is later to be established as a distinction between pragmalinguistics and sociopragmatics, albeit
outside explicitly stated ideological agendas.
Apart from the societal perspective on language and the association of pragmatics with the social conditions of language use, the cognitive perspective on language contributes to its construal as an instrument of thought. The latter line of investigation has led to the use of the terms sociopragmatics and psychopragmatics
to carry the distinction between the use of language for communication purposes
and the use of language as an instrument of thought, respectively. More specifically, in their attempt to shed light on the pragmatic turn in psycholinguistics, Dascal and Franozo (1989) refer to pragmatics as the field of study concerned with
conversational maxims, politeness requirements, contextual circumstances of utterances, and other factors that impose constraints on the selection and understanding, in a given context, of a particular linguistic expression for a certain cognitive
content (Dascal and Franozo 1989: 5). In many cases, the pragmatic turn in psycholinguistics amounts to the study of the acquisition of pragmatics by young
children and, in particular, the study of how children acquire and perform speech
acts and how they establish interpersonal communicative relations (Bates 1976).
However, Dascal and Franozo (1989) observe that language plays an important
82
Sophia Marmaridou
role in our mental processes and higher level cognitive processes or states, such as
reasoning, problem solving, storing and retrieving information, believing, etc.
They further claim that the use of language for such purposes also belongs to pragmatics: we use language not only to communicate, but also to do our reasoning, to
solve problems, etc. They therefore use the term psychopragmatics to refer to the
study of such mental use of language and distinguish it from the study of its use for
communicative purposes, i.e. from sociopragmatics. By the former term they refer
to internal pragmatic factors that affect the performance of the cognitive operations
themselves which in some way involve language. By sociopragmatics they refer to
external pragmatic factors that concern the perception and the production of linguistic signs in a particular situation, such as indirectness in the performance of
speech acts. It follows that psychopragmatics is a theory of the way in which the
linguistic environment of thought influences thought, whereas sociopragmatics is a
theory of the ways in which the non-linguistic environment (which includes both
thought and situation) affects language interpretation and production (Dascal and
Franozo 1989: 13).
The proposed distinction between sociopragmatics and psychopragmatics
serves to compare and contrast language use in jokes and dreams, topics mainly addressed in the framework of Freuds writings (see Freud 1900; 1905). More specifically, Dascal (1985: 98) argues that jokes depend for their effectiveness on the
existence of the sociopragmatic phenomenon of indirectness; they are intended to
be understood as such by an audience and therefore perform a social function. As
opposed to jokes, dreams are mental objects, not meant to be communicated to an
audience, i.e. they have an asocial character. Therefore, they are not socially constrained, their only constraint being the existence of some kind of associative link,
perhaps arbitrary and fortuitous, that provides some path for the expression of
a problematic content. In this respect the psychopragmatic role of language
in dreams is [] similar to its use for another cognitive purpose namely, as a mnemonic device (Dascal 1985: 104). In this analysis, Dascals sociopragmatics
covers the broader area of general pragmatics in both the Anglo-American and the
European traditions with the aim of distinguishing it from psychopragmatics, the
latter illustrating a pragmatic turn in psycholinguistics.
3.
The pragmalinguistics/sociopragmatics distinction

in a cross-linguistic and cross-cultural perspective
The distinction between pragmalinguistics and sociopragmatics was essentially established within the framework of applied linguistics studies on which it was to
exert a strong influence. It was initially introduced in the work of Thomas (1981),
further explored in Thomas (1983), and explicitly spelt out as a methodological approach in pragmatics in Leech (1983). In the latters programme, General Prag-
83
matics is that aspect of human language which concerns the use of language in
communication. More specifically, the understanding of the nature of language
depends both on the study of grammar, as the abstract, formal system of language
(comprising phonology, syntax and semantics), and on the study of pragmatics, as
the interaction of the Cooperative and Politeness Principles partly characterizing
language use. Moreover, Leech (1983: 7) views semantics as a level of contact between grammar and pragmatics, so that linguistic meaning can be fruitfully studied
by a combination of semantic and pragmatic approaches, the former targeting the
sense of an utterance and the latter its force.
Within Leechs project, General Pragmatics is a fairly abstract study of the general conditions of the communicative use of language, excluding more specific
local conditions on language use: The latter may be said to belong to a less
abstract field of SOCIO-PRAGMATICS, for it is clear that the Cooperative Principle and the Politeness Principle operate variably in different cultures or language
communities, in different social situations, among different social classes, etc.
(Leech 1983: 10). Alongside sociopragmatics, Leech introduces PRAGMALINGUISTICS as the study of the more linguistic end of pragmatics where we consider the particular resources which a given language provides for conveying particular illocutions (Leech 1983: 11). It follows that sociopragmatic studies are
culture-specific, while pragmalinguistic studies are language-specific. Apparently,
Leechs observation that sociopragmatics is the sociological interface of pragmatics is implicitly an attempt to address an issue that Levinson (1983: 29) also
raises when he observes that the boundary between sociolinguistic and pragmatic
phenomena is hard to define.
Since Leech (1983) is concerned with establishing the Cooperative and Politeness Principles as the two cornerstones of General Pragmatics and with arguing
their universality, he does not pursue the pragmalinguistics/sociopragmatics distinction any further. Nevertheless, he views both principles as the natural constraints of universality. For example, while claiming the universality of the principles, he concedes that they do not apply in an identical manner in all societies:
Indeed, one of the main purposes of sociopragmatics [] is to find out how different societies operate maxims in different ways, for example by giving politeness a higher rating than cooperation in certain situations, or by giving precedence to one maxim of the
Politeness Principle rather than another (Leech 1983: 80).
Similarly, he observes that cultural stereotyping (e.g. the British are more tactful
that the Americans) is related to pragmalinguistic strategies such as strategies of
indirectness, and the norms observed in the performance of these strategies in different speech communities (Leech 1983: 231).
Leechs (1983) distinction between pragmalinguistics and sociopragmatics is
the natural outcome of his claim about the universality of the principles of pragmatics he proposes. By prioritizing conversational principles, he proposes a rhe-
84
Sophia Marmaridou
torical model of pragmatics referring to the study of the effective use of language
in communication and focusing on the goal-oriented speech situation, whereby the
speaker uses language in order to produce a particular effect in the mind of the
hearer. Leechs principles of general pragmatics are to be held constant and form
the basis, or the ground, against which all other language- and culture-specific
pragmatic variables can be studied as distinct figures. This position lends itself to
cross-linguistic and cross-cultural comparisons of communicative behaviour with
obvious applications for language teaching/learning and testing. Moreover, the distinction proposed within this framework seems compatible with the aims and
methods of the study of discourse and historical pragmatics in particular. In what
follows, these areas of investigation will be related to the conceptual and methodological distinction between pragmalinguistics and sociopragmatics as introduced in this section.
4.
Applications of the pragmalinguistics/sociopragmatics distinction
4.1.
Preliminary investigations in contrastive linguistics
The impact of the pragmalinguistics/sociopragmatics distinction has been particularly felt in the conceptual organisation of some theoretical constructs it has motivated, such as pragmatic failure, pragmatic transfer, and pragmatic development.
This development in the area of applied linguistics is to some extent foreshadowed
by Riley (1979) advocating a contrastive pragmalinguistics.
Taking the broad view of pragmalinguistics as pragmatic linguistics, Riley
(1979) defines its aim as the study of communicative acts. These may be realised
verbally, paralinguistically, or non-verbally, and rest on a theory of illocution; illocutionary force is extended in this system to all kinds of communicative acts. Focusing on linguistic acts, he notes that what is central to contrastive pragmalinguistics is language functions rather than linguistic structures, discourse, not
grammar, the communicative act in context, not the sentence in isolation (Riley
1979: 57). Drawing in this way a methodological line between form and function,
he highlights instances of same form and different function (e.g. the possibility of
using Youre not going out to express the illocutions of prohibiting, confirming,
threatening, expressing surprise, or stating), and instances whereby the same function is performed by different forms (e.g. the function of agreeing as expressed by
the forms Yes, of course, I agree, nodding, or repetition of the previous speakers utterance). In his model, form-function pairings constitute the ground against
which cross-linguistic contrasts are to be drawn.
In order to place this model in a wider framework encompassing all aspects of
pragmatic study, Riley (1979) further proposes cross-linguistic contrasts across
what he calls discourse structure, consisting of three sub-levels of structure:
85
Firstly, formal structure includes all aspects of message-bearing elements, e.g. verbal, paralinguistic, or non-verbal, in a particular situation. Secondly, illocutionary
structure consists of sequences of illocutionary acts, e.g. inviting accepting, confirming thanking, etc. Thirdly, interactive structure refers to linguistic organization in terms of interactional tactics such as turns (e.g. opening, reply, closing, address, exchange, transaction, etc.). Evidently, Rileys work is an attempt to include
in his model all topics of pragmatic interest and put them to the forefront of contrastive linguistics. However, his efforts to systematise the proposed model lead to
a schematic and idealised understanding of communicative acts. For example, the
sequences of illocutionary acts seem to refer to idealised prototypes. In reality, invitations are typically, but not always, followed by acceptances, or refusals, confirmations by thanks, etc.
This apparent theoretical weakness of Rileys model, though, is counterbalanced by its applicability in contrastive analysis and language teaching in particular. More specifically, the model allows for the identification and the teaching of interactional styles and illocutionary point in relation to sociolinguistic variables
such as roles, status, formality, and discursive control, as for example, in a typical
teacher student illocutionary sequence, or in a typical asking-for-street-instructions exchange between strangers. In this way, Rileys proposal addresses newly
created needs in the methodology of communicative language teaching research
that was being carried out at the time (see, e.g., Candlin, 1976; Brumfit and
Johnson, 1979; Widdowson, 1979). Perhaps the most important contribution of
Rileys work in introducing contrastive pragmalinguistics is that he creates an
awareness of the significance of pragmatic knowledge to language teaching and
emphasises the systematicity of such knowledge so that it can be actually taught.
His suggestions clearly encourage a methodological turn, to be further explored in
the context of the pragmalinguistics/sociopragmatics distinction introduced by
Thomas (1981; 1983) and Leech (1983).
4.2.
Pragmatic failure
Following Leechs (1983) division of linguistics into grammar and pragmatics and
adopting his goal-oriented understanding of language use (see above), Thomas
(1983) first construes linguistic competence as consisting of grammatical competence (abstract knowledge of phonology, syntax and semantics) and pragmatic
competence as the ability to use language effectively in order to achieve a specific
purpose and to understand language in context (Thomas 1983: 92). Within this
framework, pragmatic failure is defined as an inability to recognise the force of
the speakers utterance when the speaker intended that this particular hearer should
recognise it (Thomas 1983: 94). Notably, whereas grammatical error indicates the
speakers restricted linguistic proficiency, pragmatic failure reflects badly on her as
a person. It follows that, in the context of cross-cultural communication, pragmatic
86
Sophia Marmaridou
failure may lead to a misunderstanding of the foreign speakers intentions. Awareness-raising tasks aiming to develop foreign students metapragmatic ability to
consciously analyse language use are commendable in this case, but inadequate in
Thomass (1983: 99) view, unless informed by the distinction between pragmalinguistic and sociopragmatic failure.
The proposed categorisation of types of pragmatic failure, consistent with
Leechs (1983) dyad of language vs. culture specificity, is not intended to impose
rigid boundaries. Thomas (1983: 101) explicitly states that the two concepts form a
continuum from what is language-specific to what is culture-specific. Pragmalinguistic failure results from the speakers mapping of pragmatic force to an utterance
in a way that is systematically different from a native speakers. Sociopragmatic
failure results from the speakers miscalculation of the social conditions placed on
language in use, such as the size of the speakers imposition on the addressee in
performing a particular speech act, a cost/benefit scale of acts in the particular culture, the social distance between interlocutors, and relative, culture-specific rights
and obligations of interlocutors to each other in specific situations. As one moves
from the communicative intent of a speakers utterance to the interpersonal dynamics between speaker and addressee, one is also moving from the pragmalinguistic to the sociopragmatic end of the continuum.
In providing examples of pragmalinguistic failure, Thomas (1983: 101) concedes to the systematicity of formfunction pairings in a particular language and
culture in the sense that native speakers assign pragmatic force to certain utterances fairly predictably. Unlike Riley (1979), she seems to restrict the scope of this
systematicity to highly conventionalised utterances, such as the often-quoted can
you do X English utterance used to issue a request rather than ask for ones ability.
The fact that in other languages, such as French or Russian, the opposite is true
may lead the foreign language learner to an inappropriate transfer of a formfunction pair, and hence to pragmalinguistic failure. A further cause of pragmalinguistic failure may be the inappropriate transfer of speech act strategies from L1 to
L2, such as the use of direct vs. indirect forms of issuing requests politely. For
example, it is reported that polite, indirect requests for street instructions (e.g. excuse me, could you please tell me how to ) would be counterproductive in Russian, while they are pretty standard in English. Hence, the Russian learner of English using direct means to perform a request might be misunderstood as being
impolite (Thomas 1983: 102).
Politeness is indeed an area in which pragmalinguistic failure merges with sociopragmatic failure. Whereas the choice of polite forms and strategies seems to
belong to the pragmalinguistic end of the continuum, when and to whom to be polite concerns the sociopragmatics end. In Thomass view, the size of imposition of a
speech act, the awareness of taboos in a particular cultural context, and differing
assessments of relative power or social distance between interlocutors may lead to
sociopragmatic failure. One may add to these sources of sociopragmatic failure the
87
orientation of a culture to positive or negative politeness (see Sifianou 1992). However, it is not always clear whether an instance of pragmatic failure can be assigned
to the pragmalinguistic or to the sociopragmatic level. An example that Thomas provides has to do with the conception of what constitutes free goods in
the former Soviet Union as opposed to Britain. Asking for a cigarette in the Soviet
Union requires a minimal degree of politeness, as these items are free goods in
this cultural setting. A Russian asking for a cigarette in Britain and using a similar
strategy may have either encoded wrongly the correct amount of politeness she
intended, or misjudged the amount of imposition of her request. The former case
would represent an instance of pragmalinguistic, while the latter of sociopragmatic
failure. The effect of either of these two types of pragmatic failure may be the same
for the addressee and their causes difficult for the analyst to distinguish. However,
the distinction is essential in the context of language teaching, as it reflects two different types of decision-making. As Thomas observes, pragmalinguistic decisions
are language-specific and the teacher can correct them in a fairly straightforward
manner. Sociopragmatic ones are culture-specific and sensitive to the learners
own system of values; hence they should not be corrected, but only pointed out
and discussed (Thomas 1983: 109).
It is therefore clear that the discussion of pragmatic failure in the context of language teaching relies heavily on the distinction between pragmalinguistics and sociopragmatics. However, given that sociopragmatic decisions are social before
they are linguistic (Thomas 1983: 104), the importance of the distinction between
sociopragmatics and sociolinguistics becomes apparent. The identification of social roles and stereotypes, power relations between interlocutors relating to age,
social status, or dialect, socioculturally-based evaluations of situations, and linguistic value judgements have been typically considered social variables and have
traditionally concerned sociolinguistic research (see, e.g. Holmes 2008). How are
all these to be differently researched in sociopragmatics? Indeed, this question is
often raised, at least implicitly, in the relevant literature and is differently approached (see the discussion in Bou-Franch and Garcs-Conejos 2003; Roever
2006; Archer and Culpeper 2003, below). It transpires that the application of the
pragmalinguistics/sociopragmatics distinction to pragmatic failure on the one hand
enriches language teaching methodology, but on the other, raises the issue of the
theoretical distinction between sociopragmatics and sociolinguistics, to be addressed in the next section.
4.3.
Pragmatic transfer
The pragmalinguistic/sociopragmatics distinction is also reflected in the study of

pragmatic transfer within the wider context of second language acquisition research. The relevant literature strongly supports the position that non-native speakers understanding and production of linguistic action is influenced by their L1
88
Sophia Marmaridou
pragmatic knowledge (Kasper 1992; Kasper and Rose 2002). Prominent in this
kind of research is the concept of interlanguage as a stage on a continuum representing the second language learners development towards the L2 system. In Selinkers (1972) definition, interlanguage is a psycholinguistic phenomenon which
mainly refers to the learning process, rather than the outcome of this process. However, the dominant practice in interlanguage pragmatics concerns the outcome, i.e.
the collection and comparison of data from native speakers linguistic performance
in the target language, the learners use of her native language, and the learners interlanguage (i.e. her use of the target language at a particular stage of development).2 In this context, pragmatic transfer concerns the possible influences from
the learners L1 on her interlanguage performance.
Even though in practice pragmatic transfer has been typically associated with
transfer of speech act knowledge, indirectness, and politeness, it may also make
reference to sociolinguistic, or discourse parameters in the same instance of language use. For example, transfer of address terms and register concern social variation in language use and are therefore to be considered in the domain of sociolinguistics. Yet, their strategic exploitation in pursuing some illocutionary goal
places their investigation in the area of pragmatics proper. Moreover, some speech
acts are characterized by their position in discourse: greetings occur discourse-initially and refusals occur as seconds in a conversational exchange. Furthermore,
the speech act force of an utterance may become evident only after the utterance is
responded to by the interlocutor, which points to its negotiable character and conversational grounding.
As in the case of pragmatic failure, the pragmalinguistics/sociopragmatics distinction has enabled a more detailed discussion of pragmatic transfer and has added
clarity to this concept by maintaining the corresponding form/function dyad. In
Kaspers (1992) definition, pragmalinguistic transfer shall designate the process
whereby the illocutionary force or politeness value assigned to particular linguistic
material in L1 influences learners perception and production of form-function
mappings in L2 (Kasper 1992: 209). However, as shown in the discussion of pragmatic failure, politeness value does not only depend upon linguistic material, but
also upon social parameters of the speech event, such as the relative social distance
of interlocutors and the power relations between them. For example, L2 learners
may be familiar with address terms of varying degrees of deference or intimacy in
L2, but their use depends upon the learners assessment of social roles, settings,
etc., which may be based on the their own culture. Therefore, politeness is not only
a candidate for pragmalinguistic, but also for sociolinguistic transfer. As Kasper
(1992: 209) observes, the latter is operative when the social perceptions underlying language users interpretation and performance of linguistic action in L2 are
influenced by their assessment of subjectively equivalent L1 contexts. It follows
that the form-function pair does not guarantee appropriate constraints on the discussion of pragmalinguistic and sociopragmatic transfer. Whether politeness
89
transfer occurs at the pragmalinguistic or the sociopragmatic level is largely a

matter of perspectivisation. If construed as a transfer of (in)appropriate forms, the
latter are foregrounded against the background of the social conditions of politeness. If the social conditions of politeness are focused upon, the formal aspects
constitute the background to the analysis.
The issue of pragmalinguistic and sociopragmatic transfer is also associated
with the distinction between positive and negative transfer. Positive transfer of elements from L1 in L2 performance is methodologically difficult to identify, given
that acceptable L2 performance may have sources other than positive transfer of
L1 elements, for example the application of general pragmatic principles, generalisation of L2 specific knowledge, and teacher input, among others. As Kasper
(1992: 213) correctly points out performance data alone cannot tell us whether
[] learners did indeed consult their L1 knowledge or operated strictly on their interlanguage pragmatic competence. More specifically, positive transfer has been
associated with the possibility of pragmatic universals (as in Grice 1975; Leech
1983; Brown and Levinson 1987) that are assumed to function cross-culturally. Interlanguage research has shown that L2 learners have access to the same range of
strategies as native speakers in implementing linguistic actions, such as requests,
suggestions, invitations, refusals, complaints, compliments, etc. (Wolfson 1981;
Trosborg 1987; Beebe and Takahashi 1989; Blum-Kulka 1991; Olshtain and Weinback 1992). This has been taken as evidence of the existence of pragmalinguistic
universals. Similarly, L2 learners sensitivity to social factors such as interlocutors relative status, or degree of imposition, etc., possibly indicates the availability
of sociopragmatic universals. However, in spite of the fact that the evidence is
based on a variety of languages and cultures, the existence of pragmatic universals
at the pragmalinguistic or sociopragmatic level is by necessity an empirical matter,
especially since the universality of theoretical concepts in general pragmatics, such
as politeness, or the co-operative principle, have been seriously questioned (see,
e.g. Keenan 1976; Rosaldo 1982; Wierzbicka 1987; Eelen 2001). Negative pragmalinguistic transfer occurs when inappropriate forms or strategies are used in performing particular speech acts affecting the politeness value of the utterance, or
even its illocutionary force. For example, it has been reported that, in expressing
refusals, Japanese learners of English use strategies such as generalisations (I
never yield to temptation), statements of philosophy (to err is human), or suggestions for alternative action (why dont you ask someone else?) that are atypical of native English speakers (Beebe et al. 1990). At the sociopragmatic level, L2
learners perception of contextual factors may originate in their L1 and affect politeness style. It is reported that Japanese learners of English selected their refusal
strategies on the basis of whether the refusers status was higher or lower than the
interlocutors, whereas American native speakers choice was made on the basis of
status-equal or status-unequal speaker addressee relationships. In spite of the
great number of studies that address the issue of negative pragmatic transfer, the
90
Sophia Marmaridou
systematicity of the relevant experiments, and the careful assessment of their results, L2 learners inappropriate performance may also be a function of other factors, such as learning context, shifts in the available target variety of the L2, little
exposure to the target language and culture, varying L2 development stages, and
lack of contextualisation cues, among others. To the extent that the pragmalinguistics/sociopragmatics distinction shows the need for more precise experimental
techniques in the study of pragmatic transfer, it can contribute to the validation of
hypotheses and is therefore to be understood as a useful methodological tool in interlanguage pragmatics.
4.4.
Pragmatic development
A further contribution of the pragmalinguistics/sociopragmatics distinction to

pragmatic research concerns the issue of the development of grammar and pragmatics in interlanguage research. An important argument in interlanguage pragmatics evolves round the primacy of grammatical or pragmatic knowledge in
interlanguage development. While focus on the development of L2 learners
pragmatics as an autonomous component of communicative competence has had a
serious impact on second language acquisition research, its relation to other aspects of communicative ability has also attracted attention (see Bardovi-Harlig
1999 and references therein). In this context, the pragmalinguistics/sociopragmatics distinction serves to investigate what happens when existing grammatical
knowledge (such as the semantics of the progressive as a grammaticalised aspectual category) is not put to target-like pragmalinguistic use (e.g. to mitigate imposition), as in I was wondering if I could have a word (Kasper and Rose 2002: 180);
moreover, this distinction serves to investigate what happens when grammatical
knowledge leads to target-like pragmalinguistic use, but non-target-like sociopragmatic use. Robinson (1992) reports that, although a Japanese learner of English
had the appropriate knowledge to form and use the I would like to form to express preference, she nevertheless used I want to on the sociopragmatic assumption that the former form would be too polite when addressing her American
friend.
Several studies show that L2 learners may have knowledge of a particular
grammatical structure, which, however, they do not put to native-like pragmalinguistic use. For example, Bodman and Eisenstein (1988) report on L2 learners use
of sophisticated grammatical knowledge which they put to unconventional L2
pragmalinguistic use, as in expressions of thanking like May God grant you a long
life. It is hypothesised that such instances also testify to L1 pragmatic transfer
phenomena. Similarly, Takahashi (1996; 2001) has found that advanced Japanese
EFL students avoided bi-clausal I was wondering whether you could VP and favoured the mono-clausal Could you (please) VP? even though they knew the semantics of the bi-clausal structure. On this evidence Takahashi (2001: 173) con-
91
cludes that the Japanese EFL learners lack the L2 pragmalinguistic knowledge
that an English request can be mitigated to a greater extent by making it syntactically more complex. Sophisticated as the above studies may be, they cannot exclude the possibility that the particular adult L2 learners production is constrained
by sociopragmatic, rather than pragmalinguistic, considerations. For example, it
could be argued that Bodman and Eisensteins subjects may have chosen the particular non-native-like forms of thanking as a way of maintaining their cultural
identity. Besides, it is possible that the Japanese EFL students may not have assessed the particular request(s) as requiring a comparatively higher level of mitigation than others. It is therefore obvious that the lack of sharp boundaries between
pragmalinguistics and sociopragmatics is also testified in pragmatic development
studies.
Knowledge of a grammatical structure and its pragmalinguistic functions, accompanied by lack of familiarity with the sociopragmatic conditions of target-like
use, is also reported in the relevant literature. Bardovi-Harlig and Hartford (1991)
show that, in a student-advising session at an American University, a Japanese student wishing to reject the faculty advisors suggestion uses a series of questions as
a strategy of indirectly conveying negative intent. In the researchers analysis, the
student clearly had the grammatical knowledge to form questions and the pragmalinguistic knowledge to use questions in order to indirectly convey intent, but did
not have the sociopragmatic knowledge to assess when the particular strategy is
appropriate and effective. It is also reported that this question strategy was not observed in the rejections of American-English speaking students. The particular
data fall short of providing evidence for the appropriate use and effectiveness of
the question strategy in conveying intent by L2 learners or by native speakers of
the L2. For example, they do not show whether American-English speaking students employ the question strategy to indicate negative intent in other social settings, or to what extent this can happen. If it turned out that the particular strategy
to convey negative intent is not favoured by native speakers of the L2, then it is
possible that the Japanese student has erroneously associated a grammatical form
with a pragmatic strategy, which is evidence for insufficient development of her
pragmalinguistic rather than her sociopragmatic knowledge. Moreover, even if the
question strategy was associated with the same function by American speakers, it
is not clear whether the understanding of negative intent would result from the operation of sociopragmatic factors or from on-line inferencing heuristics, i.e. pragmalinguistic factors. As previously noted with respect to pragmatic failure and
pragmatic transfer, it is sometimes difficult to diagnose the source of a learners inappropriate use of a speech act strategy at any particular stage of L2 development.
92
4.5.
Sophia Marmaridou
Second language teaching and testing
The most widely stated aim of L2 teaching is to enable successful communication

in the foreign language. This involves not only the mastery of the phonology, syntax and semantics of the L2, but also its pragmatics. The question that is typically
addressed in this context is whether pragmatic knowledge can be taught in the
classroom, i.e. not necessarily by cultural immersion in the native environment,
and how it can be taught to learners of different ages, educational backgrounds, etc.
The applications of the pragmalinguistics/sociopragmatics distinction in second
language acquisition research and the consequent refinement of the concepts of
pragmatic failure and pragmatic transfer relate in an obvious way to second language teaching and testing. If this distinction is useful in, perhaps selectively, identifying causes of pragmatic failure and areas of pragmatic transfer, then L2 teaching and testing should be oriented towards the development of pragmalinguistic
and sociopragmatic appropriateness (Harlow 1990: 328).
Following Leech (1983), Kasper and Rose (2001) claim that communicative
action includes not only using speech acts (such as apologising, complaining,
complimenting, and requesting), but also engaging in different types of discourse
and participating in speech events of varying length and complexity. More importantly, speakers do not only want to perform speech acts in order to achieve
their goals, but also wish to attend to interpersonal relationships with the other
speech participants in the communicative event. Within this context, pragmalinguistics refers to the resources for conveying communicative acts and encoding interpersonal meanings. According to Kasper and Rose (2001: 2), these resources include pragmatic strategies such as directness and indirectness, speech act routines,
and the intensification or softening of communicative acts. The social perceptions
underlying participants interpretation and performance of communicative action
fall within the area of sociopragmatics and refer to speakers and hearers social
distance and social power, their respective rights and obligations, and the degree of
imposition involved in particular communicative acts.
It transpires from the above that in language teaching research, as in second
language acquisition research, pragmalinguistics and sociopragmatics focus on the
area of speech acts and, more specifically, on their linguistic realisation and their
strategic use in particular social circumstances. In this view, teaching L2 pragmatics primarily involves teaching the means and strategies of performing particular speech acts, i.e. the pragmalinguistics of L2 (see McNamara and Roever 2006),
and teaching politeness, i.e. teaching how to assess social distance between interlocutors, social power and degrees of imposition in the L2 sociocultural setting
(see Bou-Franch and Garcs-Conejos 2003). Clearly, to be pragmatically competent in L2, learners need knowledge both at the pragmalinguistic and the sociopragmatic levels. For example, learners need to know not only how to express a
particular speech act indirectly, but also the relation between indirectness and po-
93
liteness, as well as when and to whom to be polite, the degree of politeness that is
required in a particular context of interaction, etc.
Apart from the linguistic expressions themselves, strategies of communicative
actions may vary not only according to context within a particular language, but
also across languages and cultures. However, as Kasper and Rose (2001) observe,
adult learners get a considerable amount of L2 pragmatic knowledge for free, since
some pragmatic knowledge is universal, while other aspects of such knowledge
may be successfully transferred from the learners L1. At the pragmalinguistic
level, L2 teaching is facilitated by the cross-cultural availability of the major categories of speech acts (e.g. Searles (1979) directives, commissives, expressives,
etc.), and specific acts such as greetings, leave-takings, etc. Moreover, it has been
found that the major realisation strategies of some speech acts occur across several
ethnolinguistically distant speech communities (see e.g. House and Kasper 1987;
Faerch and Kasper 1989; Rose 2000, among others). It is shown that in early learning states, learners may not have acquired the linguistic means necessary to implement such strategies, but once linguistic obstacles are removed by teaching, they
will use the appropriate strategies without instruction. Similarly, corresponding
form-function pairings between L1 and L2 used in the same contexts facilitate
pragmalinguistic performance.
Positive transfer and awareness of universally existing social variables, such as
social distance, social power and degree of imposition, often affected by other social parameters such as sex, age, and familiarity between interlocutors (Harlow
1990), can also facilitate the learners task in acquiring sociopragmatic knowledge
in L2. However, pragmalinguistic and sociopragmatic conventions make reference
to the lexical and grammatical constructions of particular languages, while successfully performing communicative acts often depends on the degree to which
a strategy is conventionalised in a particular language and culture. These parameters, then, point to the necessity of classroom instruction in L2 pragmalinguistics and sociopragmatics. Especially concerning the latter, it has been argued
that the promotion of sociopragmatic awareness is also likely to improve production and comprehension of pragmalinguistic resources (Bou-Franch and GarcesConejos 2003).
The apparent usefulness of the pragmalinguistics/sociopragmatics distinction
in the study of L2 teaching is not proportionately reflected in L2 testing research.
As in the discussion of pragmatic failure above, it is often difficult in practice to determine whether a given error is due to pragmalinguistic or sociopragmatic deficits
(McNamara and Roever 2006: 55). For example, if politeness markers are missing
from an utterance, this could be either because the learner does not know these
markers, or because she is not aware that these markers should be used in the particular situation. The former indicates pragmalinguistic, whereas the latter sociopragmatic, inadequacy. Therefore, it is difficult to devise a test that would assess
pragmalinguistic knowledge to the exclusion of sociopragmatic knowledge, or the
94
Sophia Marmaridou
reverse. The problem seems more acute when testing pragmatic strategies, such as
indirectness. While the linguistic expressions used to encode indirectness belong
to the pragmalinguistic level, the social target to which they are addressed is a sociopragmatic matter. Moreover, sociopragmatic appropriateness seems to be less
dependent on linguistic proficiency than pragmalinguistic knowledge.3 L2 learners
with a high general L2 competence are likely to build their pragmalinguistic competence more easily, but high general L2 competence does not guarantee correct
judgements about sociopragmatic appropriateness.
The problem of testing pragmalinguistics and sociopragmatics independently
from each other also relates to the types of tests devised to test pragmatic knowledge in general, as there is a significant tension between the construction of authentic assessment tasks and practicality (McNamara and Roever 2006: 54). The
two most frequent types are discourse completion tests (DCTs) and role plays. In
attempting to test sociopragmatic knowledge, multiple choice DCTs were shown
not to be reliable, as it is extremely difficult to devise distractors (i.e. wrong
items) that are totally unacceptable by all members of the target language community without being wrong in a very obvious way (Brown 2001). Open DCTs can
only provide evidence of the L2 learners sociopragmatic ability concerning the
particular items in the test (e.g. apologies, requests, formulaic implicatures, etc.)
rather than her general sociopragmatic ability. Additionally, they do not provide
evidence of the learners ability to participate in conversations in real time, or to
take turns, especially given that speakers normally distribute strategies over various turns, monitoring their interlocutors reactions and adjusting their talk accordingly. Although role plays are more reliable in this respect, they are not preferred as
testing devices because they take substantial time to conduct, they are difficult to
monitor, and they require multiple rating. Besides, role play cannot establish context in the way that authentic communication does in the real world, nor does it represent the speakers or the addressees actual face needs in real terms. However, it
is reported that role plays can assess learners ability to make judgments about sociopragmatic appropriateness at discourse level (Cook 2001).
By contrast, it is possible to test L2 pragmalinguistic knowledge practically
and reliably, even though creating items for pragmalinguistic instruments is not
easy. Roever (2006), for example, constructed a test to assess interpretation of implicature, recognition of routine formulae, and knowledge of speech act strategies.
While these three types of knowledge differ, they are based on general pragmatic
and linguistic input. Interpretation of implicature was expectedly found to rely on
the learners proficiency and deduction abilities. The effect of L2 learners proficiency in their knowledge of routines was almost negligible, given that routines are
not creatively constructed, but rather learned as holistic items to be used in specific
situations. As opposed to routines, the production of speech acts and the pragmalinguistic strategies employed therein correlated with level of L2 proficiency, as
expected. Some of the limitations of assessing L2 learners pragmalinguistic
95
knowledge are similar to the ones of testing sociopragmatics. The written format of
the tests restricts the number of contextual cues that would be available in a real encounter. The lack of extended negotiation of speech acts in the context of a DCT is
another limiting factor. However, studies of pragmalinguistic assessment in L2 can
benefit both from the development of testing techniques in general and from
further cross-cultural study of pragmalinguistic parameters of communicative acts
in particular.
It becomes obvious from all the above that the pragmalinguistics/sociopragmatics distinction is better exploited in the area of L2 teaching methodology than
in studies of pragmatic failure, pragmatic transfer, pragmatic development and L2
testing. What the latter have in common is their diagnostic methodology in assessing L2 learners approximations to the target language. The sources of particular
approximations are not easy to detect. By contrast, a theoretical categorisation into
pragmalinguistic and sociopragmatic parameters of language use can fairly easily
be mapped onto teaching methods, materials and awareness raising tasks in which
L2 learners may be involved.
5.
Historical sociopragmatics4
The concepts of pragmalinguistics and sociopragmatics have also informed the

field of study generally known as Historical Pragmatics. Still a relatively young
field of linguistics, historical pragmatics takes a broad, sociologically based view
of patterns of intentional human interaction, as these are determined by the conditions of societies of earlier periods (see Jucker 2006 for a brief overview).5 One approach to historical pragmatics, pragmaphilology, describes the contextual aspects
of historical texts and is essentially synchronic. It refers to the identities of speakers and addressees, their social and personal relationship, the physical and social
setting of text production and reception and the goals of the text (Jacobs and Jucker
1995). Another approach, diachronic pragmatics, concerns form-to-function and
function-to-form mappings across different historical stages of the same language.
Apparently, form-to-function mappings correspond to a pragmalinguistic perspective on the analysis of phenomena such as deixis (Fries 1993), discourse markers
(Brinton 1990), or interjections (Taavitsainen 1995). Diachronic function-to-form
studies, adopting a sociopragmatic perspective, compare the realisation of speech
acts, politeness formulae, text types, or forms of dialogue, at different points in the
development of a language (Arnovick 1999; Jucker and Taavitsainen 2008; Fritz
1995; Biber and Finegan 1992). Since the mid-1990s the investigation of the relevant topics is based on available historical corpora (e.g. the Helsinki Corpus of
English Texts), which are either read and annotated in an old-fashioned and time
consuming way, or analysed by corpus-linguistic methods of data retrieval. Apart
from written genres, such as essays, fiction, personal letters, and manuals for good
96
Sophia Marmaridou
behaviour, speech-based genres have also been analysed, e.g. dialogues from plays
and fiction, court proceedings, political debates, public speeches, and sermons
(Biber and Finegan 1992; Jacobs and Jucker 1995).
Within this framework, historical sociopragmatics focuses on the interaction
between specific aspects of social context and particular historical language use
that leads to pragmatic meanings. More specifically, it concerns language use in its
situational, local context and the ways in which situational contexts generate
norms which interlocutors employ or exploit for pragmatic purposes. A synchronic
approach to historical sociopragmatics consists in showing how language use
shapes and is shaped by context at a particular historical period, while a diachronic
perspective involves the investigation of how shifts in language use affect shifts in
contexts, or how shifts in contexts shape language use over time. An important line
of investigation in historical sociopragmatics concerns the reconstruction of contexts on the basis of historical texts; the latter are viewed as carrying evidence of,
or projecting, their own contexts (Culpeper 2009: 182183).
As already mentioned, the local context, or the sociological context, of language use is of primary concern to sociopragmatics. For example, Nevalainen and
Raumolin-Brunberg (1995) have studied the sociopragmatics of terms of address
in Early English correspondence. Given that the way people address each other in
interaction depends on social variables such as age, status, dialect, etc., the question arises as to the relation between sociopragmatics and sociolinguistics. According to Nevalainen and Raumolin-Brunberg (2003), sociolinguistics comprises
three areas of research, namely, social dialectology, interactional sociolinguistics
and the sociology of language. Parallel to this distinction, Culpeper (2009: 180)
identifies three levels of what may constitute the context against which texts may
be understood: the most local, immediate text and co-text of interlocutors, the
medial level of social situation (including speech events, activity types, frames,
etc.), and the most general level, making reference to national and regional cultures, institutional cultures, etc. As the author notes (Culpeper 2009: 181)
sociopragmatics should primarily, though not exclusively, concern itself with the medial
context and the phenomena that constitute it. Social situations can provide a link between micro, more linguistically-oriented considerations (the typical focus of pragmalinguistics), and macro, more sociologically-oriented considerations (the typical focus
of a field such as Critical Discourse Analysis).6
According to this definition, sociolinguistics is concerned with mapping regular

patterns of usage in interaction, while sociopragmatics is concerned with the ways
in which these regular patterns are used and exploited in particular interactions.
Given the concern of historical sociopragmatics for interactional data and the
obvious non-availability of recorded forms of interaction of earlier periods, the
question arises as to the availability of historical corpora that lend themselves to
sociopragmatic research and the search tools required for this type of investigation.
97
Recent developments in historical corpus linguistics have allowed more pragmatic

research questions in a way that is amenable to corpus-linguistic methodologies
(see, e.g., Kohnen 2008; Jucker et al. 2008); but the need is still felt for the appropriate pragmatic tagging of relevant pragmatic units, such as those of interest to sociopragmatic research. Within this framework, Culpeper and Kyts compilation
of the Sociopragmatic Corpus, a subsection of the Corpus of English Dialogues:
15601760, has made sociopragmatic annotation possible. Facing the issue of the
sociopragmatic variability of context, Archer and Culpeper (2003) devise and implement a tagging system that allows the annotation of speech changes that potentially affect the social meaning of interaction in drama and trial proceedings.7 The
variables include relatively static, sociolinguistic values, such as status (nobility,
gentry, professional, ordinary commoners, lowest groups) and age (young, adult,
old), as well as dynamic values, such as activity role (witness, defendant, customer), kinship role (father, daughter, mother, son, father-in-law), social role (surgeon, baker, friend), and dramatic role (fool, villain, seducer, etc.). The significance of this type of work lies in capturing the utterance-by-utterance interaction
between speakers and their addressees in terms of sociopragmatic variables, which
enhance the dynamic aspects of the analysed texts.
Working with the same corpus, Archer and Culpeper (2009) develop the notion
of keyness as the identification of keywords, key parts-of-speech and key semantic
fields that are statistically characteristic of the speech of dyads in interaction, e.g.
master/mistress with servant and examiner with examinee in trials. Examples include the use of imperative verbs directed to servants by their masters in the partsof-speech category and the use of the semantic domain of documents and writing in
the speech of the same dyad. These social role dyads are a specific part of particular
social situations, the latter constituting the local context relative to all the social
roles and situations that make up the two genres under examination, i.e. drama and
trial proceedings. Therefore, keyness analysis is a tool for the identification of the
characteristic discourse norms of particular local contexts. While the authors use
the term sociophilology to refer to their approach, their concern for the particular
type of annotation and the analysis of the local conditions of language use indicates
the deployment of a sociopragmatic orientation to their work.
The concept of sociopragmatics has also informed further studies focusing on
the social conditions that affect the use of texts. Adopting the model of critical discourse analysis and frame analysis, Wood (2009) examines personal letters from
late 15th century English taken from the two volumes of the Paston Collection, and
Margaret Pastons letters in particular. Wood (2009: 188) addresses the issue of
authorship, which is important in historical sociopragmatics: since some of Margarets letters were penned by her own sons, might we not be justified in ascribing
the language of those letters to them, rather than representative of Margaret Paston
herself? Supporting the view that local contexts are culture specific, she shows
that local context does not only make reference to a particular interactional dyad,
98
Sophia Marmaridou
but also to a more extended speech community, to the discursive practices of that
community, the texts distribution and its consumption. Within this framework, she
shows that on the one hand these letters evidence the manipulation of power structures within a family and, on the other, they indicate Margaret Pastons control
over the form and content of the letters, thereby confirming authorship.
Viewed broadly, historical sociopragmatics includes the macro-level of social,
socio-cultural and sociological factors as well as the micro-level of personal, situational and stylistic factors. Nevala (2009) discusses the concept of person reference and social deixis in Late Modern English letters and journals. Her data consist
of the correspondence of some members of the Lunar Society of Birmingham and
the letters and journals of Agnes Porter, an eighteenth century governess. In particular, she concentrates on the interactional aspects of the referential term friend
as used by public figures and ordinary people, as well as on the way in which selfand addressee-oriented third person reference is used to convey contextual shifts in
interpersonal distance and authority. She demonstrates that the writers in the data
strategically use different terms in order to alternate between specific social positionings, but they also take into consideration the prevalent social and societal constraints.
Examining the letters subcorpus of the Network of Eighteenth-century English
Texts within a relevance theoretic perspective, Fitzmaurice (2009) focuses on the
sociopragmatic construction of implicature and inference in the illicit courtship
correspondence between Edward Wortley and Mary Pierrepont. She argues that
key historical and cultural reference points are necessary in understanding how
communicative practices are embedded in the local context. Such practices are
typically associated with certain activity types and social rituals including courting. Therefore, the understanding of implicated meanings in these letters makes
reference to the cotext created by the discourse of the letter, the situational context
constructed by the exchange of letters, and the broader historical context and the
social constraints in which the correspondence is embedded. Fitzmaurice (2009)
shows that it is at the discourse cotext that the participants in the interaction enact
their sociopragmatic roles and fight for dominance of their relationship.
It transpires from all the above that the concepts of pragmalinguistics and sociopragmatics have been particularly useful in historical pragmatics and enhance
the possibilities of historical corpus linguistics. Also notice that, within the framework of historical sociopragmatics, pragmalinguistic concerns are also accommodated, as, for example, the study of the forms used to perform a particular function
in Nevalas (2009) work. On this understanding, sociopragmatics is broadly construed as referring to both the micro and macro levels of pragmatic analysis. It is
possible that the success in the application of the two concepts under investigation
to historical pragmatics depends partly on the availability of appropriate corpora
and partly on appropriate retrieval methods that facilitate the analysis.
6.
99
Concluding remarks
As already shown, apart from Leechs (1983) and Thomass (1983) contrast and
distinction between pragmalinguistics and sociopragmatics, these terms have been
variously construed in the relevant literature and, used independently of each other,
they have emphasised the social conditions in which utterances are produced
and understood as verbal contributions in acts of communication. Regardless of
whether the pragmalinguistics/sociopragmatics distinction is theoretically justified, analytically possible, or ideologically desirable, it has motivated the corresponding methodological distinction in L2 acquisition research and L2 teaching
and testing in particular. The concepts of pragmatic failure, pragmatic transfer and
pragmatic development, relevant in these fields of research, explicitly make reference to this distinction. The research questions in the pragmatics of L2 teaching
and testing have to do with defining L2 learners verbal approximations to L2 as
detected in their performance and then attributing them to pragmalinguistic or sociopragmatic sources. The focus is on whether L2 learners utterances are appropriate communicative contributions in the social context in which they occur and
whether their inferencing is based on an understanding of social context. Appropriateness is an essentially evaluative term (see Fetzer 2004; 2007). In the framework of L2 acquisition research it refers to the learners speech actions, which are
constantly assessed as being in accordance with the linguistic norms and contextual constraints and requirements of particular communicative acts in the target language and culture. The appropriateness of L2 learners performance is not always
easy to evaluate on the basis of the pragmalinguistics/sociopragmatics distinction.
More specifically, the absence of sharp boundaries between these two concepts
weakens their methodological usefulness in diagnosing the source of pragmatic
failure, pragmatic transfer, or the stage of L2 learners pragmatic development, especially in the context of L2 testing, as already mentioned. It seems easier to maintain the pragmalinguistics/sociopragmatics distinction in the context of L2 teaching, in the sense that the methodology adopted, the materials used, and the tasks in
which learners are engaged can be designed so as to focus on this distinction and
raise the L2 learners awareness of the linguistic and social issues involved in their
use of L2.
In the area of historical pragmatics and historical corpus linguistics the situation is different. The texts examined belong either to the same period (as in synchronic studies) or to different periods (as in diachronic studies).8 Pragmalinguistic
diachronic studies proceed from form to function, whereas sociopragmatic diachronic studies follow the function-to-form process. Historical sociopragmatics
focuses on the interaction of context and language use either synchronically or
diachronically. In the context of historical pragmatics and historical corpus linguistics the distinction between pragmalinguistics and sociopragmatics can be
fruitfully maintained in spite of the lack of sharp boundaries between the two. His-
100
Sophia Marmaridou
torical texts are viewed as definitive and complete instantiations of a complex configuration of parameters, such as coparticipants and their social, interactional and
discursive roles, communicative action, genre, and cultural norms and strategies of
a speech community. Unlike historical pragmatics research, L2 acquisition research is based on texts as approximations of the target language, which makes it
extremely difficult to diagnose just which specific parameter is at issue in evaluating appropriateness.
On the basis of the above, it is fair to say that, even though the pragmalinguistics/sociopragmatics distinction was initially launched as a level of cross-linguistic application of the general pragmatic principles of cooperation and politeness (as in Leech 1983), and even though it was systematically taken up in the
context of L2 acquisition research, it seems to have been more relevant in the area
of historical pragmatics and historical corpus linguistics as a reliable methodological tool in the investigation of the pragmatics of historical texts. Needless to say,
further theoretical justification of the concepts of pragmalinguistics and sociopragmatics, and of the distinction between them, ultimately depends on future research
in pragmatics and related areas of linguistic enquiry.
Notes
1. It is interesting to note here that Olsens views remind one of Wittgensteins (1958) position that linguistic meaning resides in language use and Wierzbickas (1992) introduction
of a Natural Semantic Metalanguage to deal with the cultural relativity of word-meaning
pairings across languages. Evidently, the dependence of linguistic meaning on use in particular socio-cultural settings is a recurring theme in linguistic theory.
2. In fact, interlanguage, as an interim system of L2 learners, has some features of the L1
and some of the L2, but also features that are independent of their L1 and L2 (see Yule
2006: 167).
3. This is compatible with the view that sociopragmatic awareness is likely to improve pragmalinguistic performance.
4. See also Jucker A. H. and I. Taavitsainen (eds.) 2010. Historical Pragmatics, Vol. 8 of the
series Handbooks of Pragmatics (Berlin: Mouton de Gruyter).
5. Within the narrow, Anglo-American tradition to pragmatics, Traugott (2004: 538) defines
historical pragmatics as a usage-based, pragmatically motivated, approach to language
change.
6. Critical discourse analysis is fundamentally concerned with analysing opaque as well as
transparent structural relationships of dominance, discrimination, power and control as
manifested in language (Wodak 2001: 2).
7. These two text types offer interactive, face-to-face, speech-related data that only approximate authentic discourse. Drama consists of imaginary constructed dialogue and trial
proceedings constitute the record of a prior speech event.
8. Relevant to this point is the distinction between diachronic pragmatics and pragmaphilology (Jacobs and Jucker 1995) and diachronic sociopragmatics and sociophilology
(Archer and Culpeper 2009).
101
Acknowledgements
I would like to thank the three reviewers for their helpful suggestions and comments. I am also grateful to Jonathan Culpeper, Marcelo Dascal, Andreas Jucker,
and Carsten Roever for making their work available to me. This paper has benefited from very useful stylistic suggestions made by Eleni Antonopoulou, to whom
I am greatly indebted.
References
2003
Frankfurt am Main: Lang.
2009
Identifying key sociophilological usage in plays and trial proceedings (1640
1760). An empirical approach via corpus annotation. Journal of Historical
Pragmatics 10/2: 286309.
Arnovick, Leslie K.
1999
Diachronic Pragmatics. Seven Case Studies in English Illocutionary Development. Amsterdam: Benjamins.
Bang, Jrgen Chr. and Jrgen Dr
1979
Language, theory, and conditions for production. In: Jacob L. Mey (ed.), Pragmalinguistics. Theory and Practice, 2150. The Hague: Mouton.
Bardovi-Harlig, Kathleen
1999
Exploring the interlanguage of interlanguage pragmatics: A research agenda
for acquisitional pragmatics. Language Learning 49: 677713.
Bardovi-Harlig, Kathleen and Beverly Hartford
1991
Saying no in English: Native and nonnative rejections. In: Lawrence F. Bouton and Yamura Kachru (eds.), Pragmatics and Language Learning, 4157.
Urbana-Champaign, IL: Division of English as an International Language,
University of Illinois.
Bates, Elizabeth
1976
Language and Context. The Acquisition of Pragmatics. New York: Academic
Press.
Beebe, Leslie M. and Tomoko Takahashi
1989
Sociolinguistic variation in face-threatening speech acts. In: Miriam Eisenstein (ed.), The Dynamic Interlanguage, 199218. New York: Plenum.
Bebee, Leslie M., Tomoko Takahashi, and Robin Uliss-Weltz
1990
Pragmatic transfer in ESL refusals. In: Robin C. Scarcella, Elaine Andersen,
and Stephen C. Krashen (eds.), Developing Communicative Competence in a
Second Language, 5573. New York: Newbury House.
1992
The linguistic evolution of five written and speech-based English genres from
the 17th to the 20th centuries. In: Matti Rissanen, Ossi Ihalainen, Terttu Neval-
102
Sophia Marmaridou
ainen, and Irma Taavitsainen (eds.), History of Englishes. New Methods and
Interpretations in Historical Linguistics, 5573. Berlin: Mouton de Gruyter.
1991
Interlanguage pragmatics: The case of requests. In: Robert Phillipson, Eric
Kellerman, Larry Selinker, Michael Sharwood-Smith, and Merrill Swain
(eds.), Foreign/Second Language Pedagogy Research, 255272. Clevedon:
Bodman, Jean and Miriam Eisenstein
1988
May God increase your bounty: The expression of gratitude in English by
native and non-native speakers. Cross Currents 15: 121.
Bou-Franch, Patricia and Pilar Garcs-Conejos
2003
Teaching linguistic politeness: A methodological proposal. International Review of Applied Linguistics 41/1: 122.
Brinton, Laurel
1990
The development of discourse markers in English. In: Jacek Fisiak (ed.), Historical Linguistics and Philology, 4571. Berlin: Mouton de Gruyter.
Brown, James Dean
2001
Six types of pragmatics tests in two different contexts. In: Kenneth Rose and
Gabriele Kasper (eds.), Pragmatics in Language Teaching, 301325. New
York: Cambridge University Press.
1987
Politeness: Some Universals in Language Usage. New York: Cambridge University Press.
Brumfit, Chris and Keith Johnson
1979
The Communicative Approach to Language Teaching. Oxford: Oxford University Press.
Candlin, Chris N.
1976
Communicative language teaching and the debt to pragmatics. In: Clea Rameh
(ed.), 27th Round Table Meeting, 237256. Washington, DC: Georgetown University Press.
Cook, Haruko Minegishi
2001
Why cant learners of Japanese as a foreign language distinguish polite from
impolite speech styles? In: Kenneth R. Rose and Gabriele Kasper (eds.), Pragmatics in Language Teaching, 80102. New York: Cambridge University
Press.
Culpeper, Jonathan
2009
Historical sociopragmatics: An introduction. Journal of Historical Pragmatics
10/2: 179186.
Dascal, Marcelo
1985
Language use in jokes and dreams: sociopragmatics vs. psychopragmatics.
Language and Communication 5/2: 95106.
Dascal Marcelo and Edson Franozo
1989
The pragmatic turn in psycholinguistics: Problems and perspectives. Theoretical Linguistics 15/1,2: 123.
Eelen, Gino
2001
A Critique of Politeness Theories. Manchester: St Jerome.
103

1989
Internal and external modification in interlanguage request realization. In:
Shoshana Blum-Kulka, Juliane House, and Gabriele Kasper (eds.), Cross-cultural Pragmatics, 221247. Norwood, NJ: Ablex.
Fetzer, Anita
2004
Recontextualizing Context: Grammaticality Meets Appropriateness. Amsterdam: Benjamins.
Fetzer, Anita
2007
If I may say so: Indexing appropriateness in dialogue. In: Anita Fetzer (ed.),
Context and Appropriateness, 115145. Amsterdam: Benjamins.
Fitzmaurice, Susan M.
2009
The sociopragmatics of a lovers spat. The case of the eighteenth-century
courtship letters of Mary Pierrepont and Edward Wortley. Journal of Historical Pragmatics 10/2: 215237.
Freud, Sigmund
1900
The Interpretation of Dreams. London: Harvester Press, Hassocks.
Freud, Sigmund
1905
Jokes and their Relation to the Unconscious. London: The Hogarth Press.
Fries, Udo
1993
Towards a description of text deixis in Old English. In: Klaus R. Grinda and
Klaus-Dieter Wetzel (eds.), Anglo-Saxonica. Festschrift fr Hans Schabram
zum 65. Geburtstag, 527540. Mnchen: Fink.
Fritz, Gerd
1995
Topics in the history of dialogue forms. In: Andreas H. Jucker (ed.), Historical
Pragmatics. Pragmatic Developments in the History of English, 469498.
(Pragmatics and Beyond New Series 35). Amsterdam: Benjamins.
Grice, Paul
1975
Semantics, Volume 3: Speech Acts, 4158. New York: Academic Press.
Harlow, Linda L.
1990
Do they mean what they say? Sociopragmatic competence and second language learners. The Modern Language Journal 74: 328351.
Haverkate, Henk
1994
The dialogues of Don Quixote de la Mancha: A pragmalinguistic analysis
within the framework of Gricean maxims, speech act theory, and politeness
theory. Poetics 22: 219241.
Holmes, Janet
2008
An Introduction to Sociolinguistics. 3rd ed. London/New York: Longman.
House, Juliane and Gabriele Kasper
1987
Interlanguage pragmatics: Requesting in a foreign language. In: Wolfgang
Lrscher and Rainer Schulze (eds.), Perspectives on Language in Performance, 12501288. Tbingen: Narr.
Jacobs, Andreas and Andreas H. Jucker
1995
The historical perspective in pragmatics. In: Andreas H. Jucker (ed.), Historical Pragmatics. Pragmatic Developments in the History of English. (Pragmatics and Beyond New Series 35), 333. Amsterdam: Benjamins.
104
Sophia Marmaridou
Jucker, Andreas H.
2006
Historical pragmatics. In: Jan-Ola stman and Jef Verschueren (in collaboration with Eline Versluys) (eds.), Handbook of Pragmatics 2006, 114. Amsterdam: Benjamins.
Jucker, Andreas H.
2008
Historical pragmatics. Language and Linguistics Compass 2/5: 894906.
2008
Speech Acts in the History of English. Amsterdam: Benjamins.
2010
Historical Pragmatics. Handbooks of Pragmatics, Vol. 8. Berlin: De Gruyter.
Jucker, Andreas H., Gerold Schneider, Irma Taavitsainen, and Barb Breustedt
2008
Fishing for compliments: Precision and recall in corpus-linguistic compliment
research. In: Andreas H. Jucker and Irma Taavitsainen (eds.),
2008
Kasper, Gabriele
1992
Pragmatic transfer. Second Language Research 8/3: 203231.
2001
Pragmatics in Language Teaching. In: Kenneth R. Rose and Gabriele Kasper
(eds.), Pragmatics in Language Teaching, 110. New York: Cambridge University Press.
2002
Pragmatic Development in a Second Language. Oxford: Blackwell.
Keenan, Edward L.
1976
The universality of conversational postulates. Language in Society 5: 6780.
Kohnen, Thomas
2008
Tracing directives through text and time: Towards a methodology of a corpusbased diachronic speech-act analysis. In: Andreas H. Jucker and Irma Taavitsainen (eds.),
2008
Leech, Geoffrey N.
1983
Principles of Pragmatics. London/New York: Longman.
1983
Marmaridou, Sophia S. A.
2000
2006
56/2: 4379.
Mey, Jacob L.
1979
Mey, Jacob L.
1993
Morris, Charles W.
1938
Foundations of the theory of signs. In: Otto Neurath, Rudolf Carnap and
Charles W. Morris (eds.), International Encyclopaedia of Unified Science. Volume 1, 77138.
105
Nevala, Minna
2009
Altering distance and defining authority. Person reference in Late Modern
English. Journal of Historical Pragmatics 10/2: 238259.
1995
Constraints on politeness: The pragmatics of address formulae in Early English correspondence. In: Andreas H. Jucker (ed.), Historical Pragmatics. Pragmatic Developments in the History of English, 5776. Amsterdam: Benjamins.
2003
Historical Sociolinguistics. Harlow: Pearson.
Olsen, Svend Erik
1979
Psychopathology, interaction, and pragmatic linguistics. In: Jacob L. Mey
(ed.), Pragmalinguistics. Theory and Practice, 233253. The Hague: Mouton.
Olshtain, Elite and Liora Weinback
1992
Interlanguage features of the speech act of complaining. In: Gabriele Kasper
and Shoshana Blum-Kulka (eds.), Interlanguage Pragmatics, 195208. New
York: Oxford University Press.
Prucha, Jan
1983
Riley, Philip
1979
Towards a contrastive pragmalinguistics. Papers and Studies in Contrastive
Robinson, Mary Ann
1992
Introspective methodology in interlanguage pragmatics research. In: Gabriele
Kasper (ed.), Pragmatics of Japanese as native and target language Technical
report no. 3: 2782. Honolulu: University of HawaiI at Manoa, Second Language Teaching and Curriculum Centre.
Roever, Carsten
2006
Validation of a web-based test of ESL pragmalinguistics. Language Testing
23/2: 229256.
1982
Rose, Kenneth R.
2000
An exploratory cross-sectional study of interlanguage pragmatic development.
Studies in Second Language Acquisition 22: 2767.
Searle, John R.
1979
Selinker, Larry
1972
Interlanguage. International Review of Applied Linguistics 10: 209230.
Sifianou Maria
1992
Politeness Phenomena in England and Greece. A Cross-Cultural Perspective.
Oxford: Clarendon Press.
Taavitsainen, Irma
1995
Interjections in Early Modern English: From imitation of spoken to conventions of written language. In: Andreas H. Jucker (ed.), Historical Pragmatics.
Pragmatic Developments in the History of English, 439465. Amsterdam:
Benjamins.
106
Sophia Marmaridou
Takahashi, Satomi
1996
Pragmatic transferability. Studies in Second Language Acquisition 18: 189
223.
Tahahashi, Satomi
2001
The role of input enhancement in developing pragmatic competence. In: Kenneth R. Rose and Gabriele Kasper (eds.), Pragmatics in Language Teaching,
171199. New York: Cambridge University Press.
Thomas, Jenny
1981
Pragmatic failure. M.A. dissertation, Department of Language and Linguistics, University of Lancaster.
Thomas, Jenny
1983
Cross-cultural pragmatic failure. Applied Linguistics 4/2: 91112.
2004
Historical Pragmentics. In: Lawrence R. Horn and Gregory Ward (eds.), Handbook of Pragmatics, 530561. Oxford: Blackwell.
Trosborg, Anita
1987
Apology strategies in natives/non-natives. Journal of Pragmatics 11: 147167.
Widdowson, Henry G.
1979
Explorations in Applied Linguistics. Oxford: Oxford University Press.
Wierzbicka, Anna
1987
English Speech Act Verbs: A Semantic Dictionary. Sydney/New York: Academic Press.
Wierzbicka, Anna
1992
Semantics, Culture, and Cognition. Universal Human Concepts in CultureSpecific Configurations. New York/Oxford: Oxford University Press.
1958
Wodak, Ruth
2001
What CDA is about. A summary of its history, important concepts and its
developments. In: Ruth Wodak and Michael Meyer (eds.), Methods of Critical
Discourse Analysis, 113. London: Sage Publications.
Wolfson, Nessa
1981
Compliments in cross-cultural perspective. TESOL Quarterly 15: 117124.
Wood, Johanna L.
2009
Structures and expectations. A systematic analysis of Margaret Pastons formulaic and expressive language. Journal of Historical Pragmatics 10/2: 187214.
Yule, George
2006
The Study of Language. 3rd ed. Cambridge: Cambridge University Press.
4.
Metapragmatics
Axel Hbler
Introduction
The term metapragmatics is slippery, many meanings have been ascribed to it,
too many perhaps. A restrictive or selective attitude, however, does not seem appropriate for a handbook article. On the contrary, the survey should accommodate
as many perspectives as are possible and justifiable. The problems mainly arise
from two interrelated facts. (1) The Greek loan prefix meta- has different meanings
(in between, after, later or beyond, among others) and thus lends itself to forming
all sorts of terms. (2) The formative meta- is popular. Wordspy, the internet window
administered by Random, lists 75 new meta-formations.1 Its attraction presumably derives from its particular air of something intellectual, detached and with an
impressively broad view.
To reduce the risk of a sell-out of this formative device, it may be prudent to
control its deployment and avoid it where it is not necessary. A case in point is the
use of the term metapragmatics, where the term pragmatics actually suffices. If we
define pragmatics as the discipline which has interpersonal communication as its
object, then it goes without saying that all abstractions and theoretical conclusions
which help explain communicative behavior form part of this discipline, in fact
they are defining features of the discipline. This holds true for all the answers given
to questions such as how to cooperate, how to be kind, polite, etc. and all the
conditions found to explain aspects of interpersonal communication, i.e., the conditions of thinkability (e.g., the constitutive rules of Searles speech acts), the
conditions of feasibility (e.g., the Gricean maxims) and the conditions of recognizability (e.g., the linguistic and paralinguistic expressive means) (cf. Caffi 2006:
84). To separate such theoretical stances or constructs from the discipline that developed them seems to make little sense. Communication can be act and object,
as Hagemann (1997: 32) subtly observes, pragmatics cannot. It is certainly true
that communication shows properties that can be labeled pragmatic, but such
labeling originates in an external view. This then is a reading of metapragmatics
that deserves mentioning,2 but will not be part of the focus of the article; the theoretical views and notions as such, however, will, of course, be used where necessary.
What remains to be unpacked are the following four readings of the term and
the four corresponding forms of linguistic practice: metapragmatics as the study of
explicit metacommunication (section 1); metapragmatics as the study of implicit
metacommunication (section 2); metapragmatics as the study of peoples abstract-
108
Axel Hbler
ing from interacting (section 3.1); and metapragmatics as metatheory of pragmatics (section 3.2).
1.
Metapragmatics as the (pragmatic) study of explicit

metacommunication
The object of a (meta)pragmatic analysis is metacommunication. In a first go, it

can be defined as communication about (selected aspects of) communication.
Usually, the communication talked about is neither communication as such (the
abstract and general possibilities and impossibilities of communication) nor some
definite previous or future communication, but the ongoing communication. Paul
(1999: 98) characterizes it immanently as practical speech reflection. Quite original is Antons view contrasting it with communication.
[W]e routinely and commonly attend not to speech but simply from it. That is, when we
listen with [sic!] one another in our everyday encounters, we routinely listen from our
speech to the thought so intended. (Anton 1998: 198199)
This behavior contrasts with what happens in particular circumstances, where

we explicitly attend to the speech itself. In Leders [1990] terminology, moments of
breakdown manifest speech dys-appearance.
Reconsider the case of reading a novel: I may be engrossed in the story, simply attending from the sentences to the story, but then, a misspelled word, a foreign word, or perhaps a too-recondite one appears. Now, I consciously and explicitly reflect to the word
[]. Speech, then, is routinely an absent body, an intentional arc which disappears for
the sake of the meaning so intended. (Anton 1998: 199)
What Antons circumscription brings out nicely is a kind of break separating communication from metacommunication, a sort of strained dialogical structure, which
is not attributable to a change in speakers (or voices), but to a change in the level of
communication. This is mainly brought about by a functional change from practical acting to reflecting.3
The change in level is often notable in that it involves a change from ordinary
language to metalanguage. Metalanguage is commonly understood as the language
referring to language; it represents what Verschueren (1998: 55) calls the object
notion of metalanguage.4 It is, however, not an autonomous language (not even
where linguists use a metalanguage). When members of a speech community engage in metacommunication, the metalanguage used does not consist of complicated terminology but of ordinary words, mainly nouns (such as chat or denial) and
verbs (such as promise or argue), whose feature that distinguishes them from other
words is their function to refer not to something in the world out there but to some
aspect of speech.5
Metapragmatics
1.1.
109
On full forms of metacommunication
Though metacommunication is always embedded in primary communication, its

extension can vary considerably. For the purpose of this article, it will be sufficient
to simply differentiate between full and reduced (or abbreviated) metacommunicative forms. A full form consists of at least one metacommunicative utterance;
anything below utterance level can no longer be considered metacommunication
proper, but is, of course, still metacommunicative.6 If primary communication is
the on-line event, then metacommunicative utterances (and more extended forms)
cause such an articulate deviation from it that they appear as off-line, as it were
(cf. Hbler and Bublitz 2007: 12). The boundary between primary communication
and metacommunication is usually not marked in strictly formal terms (except for
those cases in which a speaker refers to some previous utterance within the current
communicative situation by using some speech reporting frame (I said , you
said ); for references to utterances outside the current situation, cf. below). The
shift in level is identified, rather, in pragmatic terms.
The following questions are meant to delineate the main pragmatic parameters
that apply for analyzing metacommunication. They pertain to the issue/topic
chosen, to the intention pursued, and to some frame conditions of metacommunication (for a more detailed treatment and examples of full-length analyses, cf. Bublitz and Hbler (eds.) 2007).7
(1) What are the topics of metacommunication? Though always referring to
particular speech events, either single or serial, the topical scope can be generic as
well as specific. Metacommunicative topics of a more generic kind may, for instance, concern the problem of effability (I dont have words which could adequately express how I feel or Love is just a gross label), the mode chosen (You
dont have to touch me when you want something from me or Could you speak up a
bit?), or questions of participating in a conversation (Why do you always want to
dominate conversation? or You never take part in this kind of talk). They may be
comments that operate with general principles, norms and maxims; thus Grices
maxims find rich application (e.g., the maxim of quality in Are you telling the
truth?, the maxim of quantity in Just give us the gist of it, the maxim of relevance
in Come to the point, please! and the maxim of manner in Let me just finish this
story before we leave) as well as Leechs politeness principle (Stop being so patronizing!).
More specific metacommunicative topics may concern illocutionary functions
(Is that a complaint?), acts of reference (Who are you talking about?) or predication (Thats a very euphemistic formulation or What do you mean by insinuation?). Or they may concern matters of discourse organization. These include
performing or holding back certain speech acts (Give me your permission to do so,
please), announcing a speech act (And now I will disclose my next plans), or steering an ongoing discussion (Weve got three requests to speak; first Lila, then John,
110
Axel Hbler
then Karen). Comments relating to the text structure (such as But the real sensation
is coming now) may occur within longer turns.
(2) What for or why? These are the central questions for any pragmatic approach to communication. The purposes that metacommunicative utterances in
particular pursue are in part prefigured by their topic.8 Thus they concern the organization of discourse, including the textual structure of longer stretches of utterance (So far we have been describing a state of affairs; now we may turn to );
they help secure an adequate understanding, including attempts at rendering a
speech act function more precise (Is this a complaint or an accusation?) and other
kinds of monitoring. Their sense is to prevent or repair misunderstandings.
Another common purpose of metacommunicative comments is evaluative,
mainly critically, in nature, in as far as what is topicalized marks or implies a
negative deviance from what is to be expected (But I am repeating myself). Where
such comments are self-critical, they often represent a preventive measure anticipating criticism from others.
Besides general communicative norms and principles, which provide the standard for (metacommunicatively) evaluating a given utterance, other standards
seem to have been availed of recently as well. They are based on what has become
known under the label of critical language awareness (cf. Fairclough ed. 1992),
which Coupland and Jaworski (2004) characterize (rather critically in tone) thus:
The growing obsession with good communication (Cameron 2000) is filled by a communication industry happily producing prescriptive orders of metadiscourse which are
constitutive of new social orders. These meta-discourses are based on various regimes
of verbal hygiene (Cameron 1995) and codifications of communication skills. (Coupland and Jaworski 2004: 39)
The underlying intentions are certainly honorable: to contribute on the mental

level to the social goal of emancipation.
Farther reaching objectives, i.e., objectives that go beyond the associative link
with the communicative aspect topicalized, may consist in securing the participants receptivity or even in raising conflict, in a constructive or destructive key.
The last example (But I am repeating myself) and its variant (You are repeating
yourself) can well illustrate these purposes. And in a still wider perspective, such
metacommunicative comments may be used for defending or questioning communicative norms (the speaker subscribing to a culture-specific norm of brevity
and conciseness) or even creating/modifying personal identities as communicators
(profiling him- or herself as a critical, assertive person with a social status that entitles him or her to utter things of that sort).
(3) Who is the target? Though comments on utterances may appear completely
depersonalized, the speaker of the utterance commented on is nonetheless the ultimate target of such comments. In principle, any participant in a communication
can be the target; we thus can roughly differentiate between comments directed at
Metapragmatics
111
oneself and comments directed at some other speaker (cf. Hbler and Bublitz 2007:
1516).
In principle, metacommunicative actions are reciprocal and egalitarian,
wherever symmetrical relations between participants in an interaction are given
or should be given (cf. Habermas 1971). There are interactive frames, however,
which do not allow for symmetry; in the classroom or the courtroom, for example,
the rights for metacommunicative utterances are asymmetrically distributed. This
pertains, of course, primarily to the right of commenting on the contribution of the
other party. But even self-commenting may show restrictions, especially where the
possibility for a certain communicative move (such as changing the topic, or
launching into a longer discourse) is not readily available; one may have to choose
the form of asking for permission to add something not really to the point instead
of simply announcing that one is going to do that.
(4) What are the forms of realization? The examples illustrating the various
points have already provided a variety of examples of how a speakers metacommunicative intentions can be realized. Since any metacommunicative activity is
based on a reflexive and diagnostic analysis, a commenting utterance seems to be
the most direct equivalent, either a categorical (I am talking too much or You are
talking too much respectively) or a modalized one (Do I talk too much? Maybe, I
am talking too much or I think you are talking too much respectively). Another option is to operationalize the analysis in terms of a corresponding action to be taken
(I should not talk that much or Stop talking that much, Could you stop talking too
much? respectively). Where the target is not the speaker him/herself, such metacommunicative utterances, especially if they are critical, are likely to provoke
some reaction (Thats ok or Im always supposed to keep my mouth shut!), and this
could even lead to extended sequences.
As to the position of metacommunicative clauses, it is most common that they
follow the utterance that they refer to. Only where the speaker takes him/herself as
target may we find a reversed order, in which the utterance referred follows. This
restriction, of course, is not surprising; it is, after all, only in this circumstance that
the speaker of a metacommunicative utterance knows what it will be that s/he (cataphorically) refers to.
A systematic treatment of how the various aspects interact would certainly be
desirable, but is currently not feasible.
Certain meta-phenomena have so far not been taken into account for good
reasons: Though they consist of reflexive language (in the sense of Lucy 1993) and
though they contribute to the ongoing communication, they do not result in metacommunication. Their common denominator consists in their being reportive in
some way or another, i.e., they refer to instances of direct speech outside the current communicative situation and either reproduce them or sum them up.9
The most common forms are direct and reported speech, which are usually
framed (introduced or rounded off) by a phrase containing a verb of saying (He
112
Axel Hbler
said, or She told me that ), a performative verb (He promised to stop smoking), or one of a more interpretative character (He insisted that ). They lack, to
quote Meyer-Hermann (1978: 128), the defining feature that the object referred to
and predicated about is a communicative interaction (sequence) or part belonging
to the same interactional unit as the metacommunicative speech act. These metarepresentations, as Wilson (2000) calls them, refer to speech events outside a
given communicative situation and can therefore be considered extracommunicative (cf. Ungeheuer 1970). Due to their extracommunicative character, these
speech references get absorbed in the primary discourse either as instrument or object. Newspaper articles are a common source for illustration; examples of another
kind are literary analyses or philosophical language essays.
A phenomenon that is not extracommunicative and still represents hardly an instance of metacommunication is the so-called echo-utterance, where the speaker
echoes the preceding speaker by repeating some linguistic material yet giving a
specific turn to it; cf. Sperber and Wilson (1986), Wilson (2000); Graf (2007).
Echo-questions (Youve got frequent-flyer status with which company?) try to reelicit a certain part of the preceding utterance that the addressee either did not
understand or feels provoked by. Echo-statements such as in the following example
He: Its a lovely day for a picnic
[They go for a picnic and it rains]
She: (sarcastically): Its a lovely day for a picnic, indeed.
(Sperber and Wilson 1986: 239)
usually just convey attitudes10 towards the propositional state of affairs quoted/
echoed. In neither case are we confronted with instances of metacommunication;
the echo-utterances are information-oriented (in that they elicit information or
evaluate information) and do not concern some formal aspect of communication.
Where they cause a break, as in the information-seeking reading of the echo-question above, they do not transpose the ongoing communication onto the meta-level.
Another quotative variant, which is not extracommunicative either but functions within an ongoing communicative situation, is at least a strong candidate for
metapragmatics. The reference is to parody, which is defined as imitating the characteristic style of a speaker (or an author or a work of his/hers) with the intention to
ridicule or achieve a comic effect. We usually associate a literary genre with it, but
parody also occurs in ordinary life and everyday conversation. The illustration that
follows is taken from a piece of literature, but the parodist here is not the author
himself but one personage in his (realistic) drama.
Nick:
George:
Nick:
George:
Nick:
George
I try not to
Get involved. Um? Isnt that right?
Yes thats right.
Id imagine not.
I find it embarrassing.
(sarcastic): Oh, you do, hunh?
Metapragmatics
Nick:
George
113
Yes. Really. Quite.

(mimicking him): Yes. Really. Quite. (Then aloud, but to himself:) Its disgusting!
(Edward Albee, Whos Afraid of Virginia Woolf? New York: Pocket
Books, 1966: 90)
The parodistic mimicking performs the metacommunicative act inexplicitly, but

could be made explicit as is illustrated in the given example by the aside with
which Georges turn ends; both modes convey his disgust at the way in which Nick
expresses himself. As is also shown by the example, parody at least in everyday
conversation where it is used only occasionally shows something similar to, but
weaker than, the break which marks the boundary between communication and
metacommunication: the ongoing dialogic flow comes to a halt. With Bally (1965),
we could characterize this inexplicit type of metacommunication as mode vcu,
contrasting with the mode pur of explicit metacommunication (cf. below).
Stylization, as treated by Coupland (2004), and some forms of irony provide
other instances of such inexplicit metacommunication. They represent doublevoiced utterances in the sense of Bakhtin (e.g., 1981) in that they are moving on
two different levels (the level of primary communication and the level of metacommunication, which consists in exaggerated imitation) and yet are united in one
utterance.
1.2.
On abbreviated forms of metacommunication
The subtitle already links this part directly to section 1.1 where full forms of metacommunication were tackled. The same (meta-)pragmatic aspects are applicable
here as well. They will, consequently, be expounded by the same guiding questions; merely the order will be slightly altered, starting this time with the most distinctive feature, i.e., the formal make-up.
(1) What are the forms of realization? The means allowing for abbreviated
metacommunication are quite easy to delineate. They mainly consist in adverbials
that typically operate on the whole proposition of the given utterance. The most
concise forms are disjuncts (such as frankly); slightly more voluminous are prepositional phrases/clauses (such as in short), infinitive clauses (such as to tell you in
a few words) and participial clauses (such as bluntly speaking). They usually can
be expanded to a fully fledged metacommunicative utterance (Ill be very frank
with you; I will sum it up in a few words; I dont want to fuss around but say bluntly
what I have in mind). Being briefer than their elaborated counterparts, they are
better integrated into the dialogical flow; the communicative break they cause is
more subtle, and at best syntactically marked off from the main clause providing
the propositional content of what is primarily communicated. While the full forms
of metacommunication were characterized as off-line phenomena (seen from the
perspective of ongoing primary communication), the abbreviated variants keep the
114
Axel Hbler
communication on-line, in a manner of speaking. It is therefore no wonder that

they are unlikely to open up metacommunicative (side-)sequences.
(2) What are the topics that the adverbials cover? Quirk et al.s (1985) classifications of adverbs/adverbial provide a comprehensive overview. Of prime relevance are style disjuncts and with some restrictions conjuncts.
According to Quirk et al., style disjuncts convey the speakers comment on
the style and form of what he is saying, defining in some way under what conditions he is speaking as the authority for the utterance (Quirk et al. 1985: 615).
The metalinguistic character can (often) be brought forth explicitly by transforming the disjunct into a clause in which the adverbial (turning into a process adjunct)
is linked to a verb of speaking with I as subject. (e.g., Frankly, I am tired I tell
you frankly, I am tired).11
In line with Quirk et al. (1985: 615 ff.), we can subclassify style disjuncts. One
group of adverbs that is of interest focuses on the manner in which an utterance is
made; with their help, a speaker can either refer to the style of his/her utterance
(e.g., bluntly, briefly, simply) or include some modal overtones (e.g., truthfully).
Recall that longer variants are also available, such as prepositional phrases, or
(in-)finitive and participial clauses (cf. above). The other group of interest consists
of adverbs (and adverbial expressions) which characterize the make-up of a formulation, whether it has to be taken, say, figuratively or literally. Again, some expressions may be tinged with modal overtones. Thus, an utterance like Hawkins
was not, strictly speaking, a traitor may address not only the issue of whether the
word traitor is a suitable term for Hawkinss behavior but also the issue of whether
or not Hawkins is a traitor. In addition, there are some adverbs of degree (belonging to the category of subjuncts) that can serve the same metalinguistic purpose:
compromisers (such as kind of, sort of) and approximators (such as almost, virtually) provide indications as to the reliability and adequacy of the lexical expression chosen. In uttering, for example, He is kind of a traitor or He virtually
stole the money the speaker may want to let the listener know that the verdictive
terms traitor and steal respectively are not to be taken literally, but come close.
These adverbials can, in general, be linked up again with some conversational
principle, norm or maxim. A speaker using truthfully in his/her utterance echoes
Grices maxim of quality, using briefly his maxim of quantity, while with figuratively he alludes to his maxim of relevance (by flouting it), and with frankly to his
maxim of manner.12 Adverbs like sort of pay tribute to Leechs politeness principle
in as far as they may render the utterance more palpable.
The second set of adverbials that in part can be interpreted in a metacommunicative key are conjuncts/conjunctions. They are text-structural markers that
specify how an utterance fits into the surrounding discourse in which what is to
follow is systematically connected to what has gone before (Halliday and Hasan
1976: 227). Not all of the 7 conjunctive roles (and sub-roles) between which Quirk
et al. (1985: 634 ff.) distinguish (i.e., listing, summative, appositional, resultive, in-
Metapragmatics
115
ferential, contrastive, transitional) are metacommunicative. The ones that do not

qualify for inclusion are those that establish the co-textual link purely in terms of
content, not of form. The ones whose use allow for a pragmatic interpretation as
being metacommunicative are adverbs of enumerative listing (such as first, second, third; to begin with, next, to conclude), summative expressions (such as in
conclusion, to summarize), contrastive adverbs with a reformulatory (such as
more accurately, alias, in other words) or replacive meaning (as in alternatively,
rather; better, worse) and finally, adverbials, such as incidentally or by the way,
which mark discoursal transition.13
The preceding classification of metacommunicative adverbials shows more
openly perhaps than the treatment of full metacommunicative utterances does a
clear-cut topical dividing line between personal and textual thematic foci. With respect to narratives, McNeill (1992: 185 ff.) differentiates in my opinion quite felicitously between two levels outside the narrative proper: the meta-level, which
includes the language clues characterizing and structuring the narrative as text, and
the para-level which relates to the narrator listener axis, comprising elements of
an interpersonal nature (e.g., providing orientation to the listener, comments and
attitudes of the narrator etc.) Applied to communication in general, we could distinguish between a meta-communicative level that would pertain to the structural
aspects of dialogues and turns and the para-communicative level that would account for those non-primary forms of communication which are person-centered
and focus on the interactional aspects involving the speaker and/or hearer.14 More
recent treatments draw similar distinctions, though less categorically. del (2006)
starts out from the notion of metadiscourse embracing various configurations that
enfocus either the text, the writer (or speaker), and the reader (or hearer) and distinguishes in a second step between metatext, which relates to the text or code
proper, and writer-reader interaction. Also Hyland (2005) uses metadiscourse
as cover-term; he defines it as basically interpersonal and then draws a functional
distinction between means that help to guide the reader (or listener) through the
text (interactive function) and those that involve the reader (or listener) in the text
(interactional category). Some such terminological differentiation could help to
disentangle the complexity of the concept at issue.
(3) The question as to who is the target (and who has got the right to freely select the target) is easy to answer for the adverbials with a metacommunicative
function. Where a speaker uses style disjuncts, s/he has necessarily him/herself as
target in that his/her metacommunicative comment refers pro- or retrospectively to
his or her own primary conversational contribution. Whereas the speaker as metacommunicative target is oftentimes quite transparent with such style disjuncts (cf.
frankly speaking), s/he seems to withdraw from being target where conjuncts that
are reflexive of discourse structure come into play. The sociological question concerning the right to select the target, in this case, boils down to the question of
whether any speakers in any communicative situation may take the opportunity to
116
Axel Hbler
comment on him- or herself as communicator in such brief (adverbial) terms. This

is such an elementary (and minimalistic) question that it can either be answered in
strictly categorical moral terms along the line of Habermas or in empirical terms on
the basis of rich data sets.
(4) The final question to face concerns, again, the raison detre, the why or
what for? The purposes that a pragmatic view can ascribe to these reduced forms of
metacommunication cannot be expected to be essentially different from those
identified for the extended forms with the speaker as target. But they can be expected to be somewhat reduced, according to the iconic principle of a proportionality between quantity of form and quantity of meaning. This seems, indeed, to be the
case. In one way or another, they help the speaker optimize the ongoing primarylevel conversation. The style disjuncts serve him/her as (prophylactic) lubricant to
avoid clashes with interactants (and subsequent repairs). The textual conjuncts facilitate the reception of the primary message emitted.
2.
Metapragmatics as the study of implicit metacommunication
Auer (1986: 22) has hailed it as a new research paradigm and Verschueren (1998:
60) holds that it is the central area, the proper domain of metapragmatic studies:
peoples metapragmatic communicative behavior based on their metapragmatic
awareness of situation and context as co-determinants of communication. No specific explicit repertoire is available to meet this meta-function. The approach conceives metalanguage as a dimension of language use and does not relate it to the
object notion of metalanguage (Verschueren 1998: 55); cf. section 1. What we look
for and find in utterances, instead, are indicators that testify to a metapragmatic
awareness of speakers, a kind of self-monitoring activity which constantly calibrates (to use a Silversteinian term) the communicative intention with situational
and contextual conditions. The significance is highlighted by Silverstein, to whom
Verschueren repeatedly refers in his argumentation.
Without a metapragmatic function simultaneously in play with whatever pragmatic
function(s) there may be in a discursive interaction, there is no possibility of interactional coherence, since there is no framework of structure here, interactional text
structure in which indexical origins or centerings are relatable one to another as aggregated contributions to some segmentable, accomplishable event(s). (Silverstein
1993: 3637)
The corresponding type of reflexive language are indexical forms, which change
their value depending on the actual event of speaking (Lucy 1993: 10). The specific segment of reflexive language consists of shifters, contextualization cues
(prosodic patterns, code switching, etc.) or implicit voices (cf. Verschueren 1998:
61).15
Metapragmatics
117
Communicatively speaking, there is, consequently, nothing that could be paraphrased in explicit metacommunicative terms (cf. 1.2). And likewise, no neat
break or boundary between a primary and a secondary level of communication can
be found, and no turn to off-line communication. But at least one metacommunicative feature is preserved: a reader/listener will be lead to attend to the words instead of attending from them, to recall Antons (1998) subtle play with prepositions
(quoted in section 1).
The attending to words, more specifically to indexical words and their metapragmatic dimension, shall be further elaborated and illustrated in two points. The
first point concerns the so-called shifters themselves, e.g., personal pronouns, deictic expressions, or tenses whose meanings shift due to their dependency on the
individual single text and situation.16 The pronoun you, for example, acquires its
specific referential meaning according to the communicative situation. This situation-dependent meaning assignment then is the result of the metapragmatic reasoning, on the speakers as well as the receivers side. More interesting and complex are such metapragmatical assignments in the case of some text-cohesive
devices, i.e., reference, substitution, or ellipsis. Halliday and Hasan (1976) provide
characterizations in which the metapragmatics behind their use and understanding
is (unvoluntarily) well highlighted. All three are forms of presupposition, devices
for identifying something by referring it to something that is already there known
to, or at least recoverable by, the hearer (Halliday and Hasan 1976: 144).17 With
respect to reference, we read that [b]oth exophoric and endophoric reference embody an instruction to retrieve from elsewhere the information necessary for interpreting the passage in question (Halliday and Hasan 1976: 33). More concretely,
a reference item signals supply the appropriate instantial meaning, the referent in this
instance, which is already available (or shortly to become available); and one source of
its availability is the preceding (or following) text. (Halliday and Hasan 1976: 227)
In substitution and ellipsis (as a special case of substitution), the presupposition is

at the level of words and structures. When a substitute is used, it signals that the
actual item required, the particular word or group or clause, is recoverable from the
environment (Halliday and Hasan 1976: 145).
The second point pertains to contextualization cues, whose metapragmatic dimension is as Gumperz (1982: 131) states rarely consciously noted. He gives
the following definition:
[A] contextualization cue is any feature of linguistic form that contributes to the signaling of contextual presuppositions. Such cues may have a number of such linguistic realizations depending on the historically given linguistic repertoire of the participants. The
code, dialect and style switching processes, some [] prosodic phenomena [] as well
as choice among lexical and syntactic options, formulaic expressions, conversational
[] strategies can all have similar contextualizing functions (Gumperz 1982: 131).
118
Axel Hbler
Among his examples, we find one where a graduate student is supposed to interview a black housewife in a low income, inner city neighborhood. After an appointment has been arranged over the phone by someone from the university, the
student goes to the black home and is met by the husband, who opens the door,
smiles and says to him So yre gonna check out ma ol lady, hah? What he utters is
a formulaic phrase identifiable through co-occurrent selections of phonological,
prosodic, morphological and lexical features (Gumperz 1982: 133); it is marked
for ethnicity and signals both expectations about what is to be accomplished and
about the form that replies must take (Gumperz 1982: 134). That such expectations exist comes best to the fore where they are not met by the interactant and
subsequently cause irritation, break-down or the like. In the example, the student
does not meet the husbands expectations by (inappropriately) replying Ah, no,
I only came to get some information. They called from the office. Thereupon the
husband dropped his smile and disappeared without a word while calling his
wife.
Sometimes, the metapragmatic reasoning behind an utterance may not be interactionally oriented, but may rather serve the maintenance and construction of identity (cf. Tajfel 1982). Identity is not only to be understood as a fixed set of properties or operations residing in the individuals cognitive make-up, where language is
regarded as a largely docile medium through which dimensions of a persons
identity [] may receive occasional expression (Wooffitt and Clark 1998: 107); it
is also (and above all) an accomplishment with respect to interactional and inferential concerns generated by the trajectory of verbal exchanges [], [whose] ascription is inextricably tied to the details of talk-in-interaction (Wooffitt and Clark
1998: 107). Take the utterance Hang on, love. The term of address used could be
indicative of an informal (cross-gender) relationship between speaker and addressee. In the actual instance, however, the addressee is not the speakers wife, but
a (female) mediator the couple has turned to in their divorce case. The speaker here
constitutes his relationship with [the addressee] in a way which is at odds with the
mediator-disputant relationship they have thus far collaboratively sustained
(Greatbatch and Dingwall 1998: 128). With this utterance, the speaker is questioning the addressees professional identity.
What the metapragmatic devices under consideration all have in common and
whereby they differ from the reflexive types taken into consideration in the preceding two sections is the mode of reflection. Bally (1965) distinguishes between the
mode pur and the mode vcu; it is the latter one which we have been confronted
with in the present section, whereas we were concerned with the former one previously. Ballys distinction is worth quoting in greater detail.
Si le langage nest pas une cration logique, cest que la vie dont il est lespression na
que faire des ides pures. Si lon me dit que la vie est courte, cet axiome ne mintresse
pas en lui-meme, tant que je ne le sens pas, tant quil nest pas vcu [] (Bally 1965:
15)
Metapragmatics
119
If language is not a logical creation, it is due to the fact that life, whose expression it
represents, does not simply produce pure thoughts: If I am told that life is short, I am not
interested in this idea unless I feel it, unless I experience it.
The following illustration fits well in our context, since the Italian pur(e) is a discourse marker.
Le jugement intellectuel La terre tourne se change en jugement de valeur dans la bouche
de Galile seciant devant ses juges: E pur si muove! Ce nest plus une vrit scientifique, cest laffirmation dune valeur attache cette vrit [] (Bally 1965: 15)
The logical judgement The globe revolves becomes an evaluative judgement in the
mouth of Galileo screaming into the face of his judges E pur si muove! Then it is no longer
a simple scientific truth, rather it is the affirmation of a value attached to this truth.
Uncovering the metapragmatic dimension of language and language use in indexical forms is the objective of the corresponding field of scientific activity, which, in
turn, derives its name from the very objective. Strictly speaking, metapragmatics is
in this sense then the study of the metapragmatic dimension of language and
speech rather than the study of metacommunication. If we were to continue in the
attempt at terminologically distinguishing the various conceptualizations of metapragmatics, the term syncommunicative could well serve this purpose; the prefix
syn- (of Greek origin like the prefixes meta- and para- in the complementary expressions) conveys quite adequately the fact that this reflexive level is not isolatable but inextricably linked to the primary level of communication.18
3.
Beyond interacting
Leaving ongoing conversations and taking a more generic view instead opens up
perspectives to be dealt with now. Observing communication in its general design
and different patterns and describing and systematizing the observations is the perspective that constitutes a first-level step away from conversation; theorizing about
the conditions, presuppositions and implications of such endeavors constitutes a
second-level step. While the second level is reserved for the professional linguist
(cf. 3.2), the first level is open to both the professional and the ordinary conversationalists (cf. 3.1).
3.1.
Metapragmatics as the study of peoples abstracting from interacting
Through their participation in all sorts of conversations, members of a speech community gather in the course of time knowledge about conversation in general, their
genres, patterns, styles, norms, etc. Since this knowledge relates to the pragmatic
dimension of communication, it may be specified as metapragmatic knowledge.
Some similar metapragmatic knowledge may be obtained by the scientist who
120
Axel Hbler
studies the conversations of this speech community as an outsider yet in the role of
a participating observer. In other words, it is hardly possible to clearly distinguish
between the scientific metapragmatic knowledge and the metapragmatic knowledge of the users of a language. The assumption of a continuum between both
stances, as made by Caffi (2006: 87), might therefore seem quite appropriate;19 the
difference consists merely in a smaller or greater distance from the object, viz.: the
use of the language and a lower or higher degree of explicitness.
Folk theories, however, call for the mediating linguist, the linguist, that is, who
elicits folk knowledge or elaborates on their shorthand versions. Gumperz (1982)
identifies as mediating linguists in particular those working in the fields of discourse analysis or ethnography of communication. Folk-linguistics is a separate
discipline (cf. Niedzielski and Preston 2000), but has roots in both. In any case the
task is to come to grips with the varying ways in which linguistic behavior is conceptualized by those engaged in it (Verschueren 1998: 6061). Of the folk-linguistic topics, speakers awareness of and knowledge about pragmatic issues is, of
course, particularly relevant for the current context. Two areas and modes (cf. Preston 1996) shall be considered in greater detail.
(1) A major source of information about peoples metapragmatic knowledge
could be the lexical repertoire they have at their disposal for referring to aspects of
speech; it is not an autonomous language, but a subset of the ordinary language.
Speech act verbs are a favorite research area (where folk taxonomies merge with
the empirical-conceptual approach of speech act theories and the ethnography of
speaking, cf. Caffi 2006: 85). But other lexical expressions relating to other selected aspects of communication could also be common ground where folk- and
empirical-conceptual linguistics meet. Lexical expressions, in general, can be interpreted as a means that a speech community uses for coding cognitive and cultural models, which consist in more or less coherent sets of concepts for structuring
experience (cf. also Gee 1999: 40 ff.). Experience also includes acting by communicating, and this, in turn, calls for adequate conceptualizations of a metapragmatic kind and their storage in words.
One of the most comprehensive collections of English speech act verbs is the
dictionary compiled by Ballmer and Brennenstuhl (1981).20 Its comprehensiveness
derives from two decisions. (a) They apply a much wider definition of speech act
verb by including all those verbs which designate (aspects of) speech activities
(Ballmer and Brennenstuhl 1981: 3), no matter whether they can be used performatively or not; thus lie and persuade, for example, are speech act verbs just as
well as admit or promise. (b) They include literal as well as metaphorical expressions, as, for example, contradict and reject. Of particular interest is the authors ordering of the gathered material into a huge word field and its many subfields,
because these can be understood as identifying the metapragmatic aspects of communication that the English speaking community takes recourse to. The word
fields in their entirety unfold a global metapragmatic conceptualization whose
Metapragmatics
121
main constituents delineate models entering into an integral system of speech activities. Each model has as its linguistic correlate a set of verbs, which again have
substructures. The enaction model, for example, consists of verbs, of which wish,
aim at, or anticipate express VOLITION , alert s. o., indicate, or refer to PUTTING INTO
FOCUS , while expressions of offering, asking, commissioning, ordering, warning
etc. are instances of ENACTION , and verbs like accept, refuse, or obey instances of
REACTIONS ON ENACTIONS . At the most refined level, we find sets of verbs that show
a high degree of synonymy; for the enactment category SURPRISING , for instance,
the following verbs are listed: amaze, astonish, astound, bewilder, disconcert, flabbergast, shock, stagger, startle, strike, surprise. Taken together they delineate, if
we will, how the English language community conceptualizes acts of surprising
somebody by words; seen in a differentiating way, each single item contributes a
special nuance to the overall-concept, in comparison to the others.
Along these lines, Goossens (1987) has made an analysis of some such nearsynonymous verbs. His treatment of the speech act verbs say, tell, talk and speak
can well illustrate how much an in-depth analysis is able to reveal about peoples
metapragmatic models, though it explicitly tries just to answer a typically Fillmorean question, namely, how do linguistic action verbs frame the scene of linguistic (inter)action? (Goossens 1987: 95). Thus, native English speakers assign
different roles to receivers, encoding in tell and say the receptor who plays a more
passive part, and in talk the interactor who plays a more active, participative part;
in the case of speak, the English speaker accounts for the two roles by using distinct prepositions, with where the receptor acts as interactor, to where as receptor
(cf. Goossens 1987: 103). The four verbs reveal, furthermore, that native English
speakers differentiate messages according to the degree to which they have been
condensed. This is reflected in the facts that, for example, say freely combines
with direct enunciations, while tell is considerably restricted in taking them
(Goossens 1987: 103); the differentiation is in principle even reflected where the
distributional pattern is reversed, in other words where more extreme condensations go with tell, but not with say (e.g., tell the truth vs. *say the truth).
Lexical expressions that belong to certain metaphorical concepts offer particularly rich information about peoples cognitive metapragmatic models.21 Metaphors have in general attracted much attention by linguists because they seem to
offer insights into how the human mind operates. As Lakoff and Johnson state in
their pivotal study of 1980: [M]etaphor is pervasive in everyday life, not just in
language but in thought and action. Our ordinary conceptual system, in terms of
which we both think and act, is fundamentally metaphorical in nature (Lakoff and
Johnson 1980: 3); and metaphors that conceptualize communication in one respect
or another have been paid attention to right from the very beginning.
For an illustration, we may turn to two of the most widely known examples,
i.e., the so-called conduit metaphor, identified by Reddy (1979) and differentiated
later by Vanparys (1995) and Semino (2006), and the argument is war concept
122
Axel Hbler
proposed by Lakoff and Johnson (1980) and modified by Grady (1998) and Semino
(2006). The results of these investigations, derived from a close analysis of sets of
metaphorical expressions in general use and subsequently supplemented by corpus
data, can be understood as reconstructions of native English speakers conceptualizations of certain aspects of how they communicate and what their form of communication is like. These conceptualizations are metapragmatic in nature.
As far as communication between communicators in general is concerned, the
dominant meta-pragmatic model people have developed shows three main constituents: the speaker/sender puts an idea-object into a word-container and sends it
(through a conduit) to the receiver, who extracts from the word-container the ideaobject.
Discussion/argumentation represents a specific form of communication/discourse. The corresponding metaphors widely in circulation reveal that English
speakers share a (metapragmatic) model of argumentative discourse that centers
around physical conflict; debatants have two positions, attacking the others
opinion, while defending ones own; we have phases of maneuvering and retreat;
the outcome amounts to truce or victory/surrender.
(2) Another important source for gaining insights into peoples understanding
of communication would consist in information that researchers are able to elicit
directly from members of the speech community; this is a method which is often
applied by discourse analysts, where the aim is to produce ideotypical descriptions [of concrete discourses] that can be dissected into significant components
(Gumperz 1982: 157). But the results are on the whole quite unsatisfying; ordinary
conversationalists are obviously poor descriptors (Gumperz 1982: 157),22 and
this should hold true even more, were they confronted with a more abstract and
complicated task. Alternatively, an experimental way of eliciting metapragmatic
knowledge may be feasible (similar to the method that sociologists, e.g., Goffman,
occasionally deploy, for discovering certain norms which people subscribe to in
relations in public, i.e., by confronting them with transgressions), but has not, to
my knowledge, been pursued. The use of questionnaires seems to be the relatively
more reliable method.
The study by Simon-Vandenbergen (1995) shall be used to further illustrate the
point at stake, even if it is not exactly made for such a purpose. Its aim was to obtain assessments of linguistic behavior from English native speakers. The method
consisted in having natives evaluate selected aspects and forms of communicative
behavior through evaluating corresponding (metaphorical) expressions, which, as
we saw, embody parts and aspects of folk models of behavior. The questionnaire
applied included expressions covering social functions of talk (e.g., conversation,
chat, or prattle), turn-taking (e.g., run on, cut someone short, step in), topic management (e.g., bring up, move unto, meander), and manner of speaking (e.g., rattle,
babble, or air, thunder, or tell someone flat out, wrap up ones meaning). The questions/tasks that were to stimulate statements about peoples metapragmatic views
Metapragmatics
123
had three formats: 1. (social) assessments in terms of positive negative neutral,

2. (aesthetic) assessments in terms of light heavy neither, and 3. the same as
format 1 but supplemented by (a) a question asking why the subject rated the expression the way s/he did and (b) the request to give an example sentence.
While certainly all tasks can be interpreted as throwing light on peoples understanding of communication, it is the supplementary task (a) which should elicit the
most; unfortunately, Simon-Vandenbergen does not provide any cue as to what and
how subjects met this task. Nevertheless, some of the results she obtained allow
conclusions. Speech tempo is a good case in point. This is the picture emerging for
native English conversationalists (cf. Simon-Vandenbergen 1995: 108): Speed is
an important criterion; high and low speeds are often negatively evaluated as too
fast and too slowly respectively, in which case they tend to be talked about in
metaphorical terms.
One could, of course, object and claim that the negative evaluations elicited by
the questionnaire were just triggered by the metaphoricity of the expressions themselves. Any such objection, however, would not be justified. As Simon-Vandenbergen already seems to imply, it is the perceived marked value of the communicative phenomenon in relation to norms and standards in a given community that
triggers metaphorical expressions and assessments, and not the metaphorical expression that triggers a corresponding evaluation. Besides, in the case of speed,
there is even some independent support from social-psychological research for the
claim that the negative evaluations relate, indeed, to speed and are not simply triggered by the metaphoricity of the lexical expression. Smith et al. (1975), for
example, had subjects listen to recordings of speakers performing one and the same
text with different speeds; in reality, it was just one speaker whose natural speech
tempo had been technically manipulated. The subjects had to judge the speakers
as to benevolence, a social criterion which aligns well with Simon-Vandenbergens positive-negative scale. The authors found that [t]he benevolence/rate
plot reveals an inverted U-relationship, with the mean for normal voices [] being
rated more benevolent (Smith et al. 1975: 150), while high- and low-speed
speakers scored significantly low in benevolence.
The insights obtained by metapragmatic research outlined so far can be enriched (and maybe even relativized) by widening the scope so as to also include
possibilities of variation, always applying the same analytical parameters.23 Of the
three main types of variation, i.e., social, developmental, and historical, the social
variant has not yet received due attention. Illustrations will therefore be restricted
to the other two, turning to the lexicological source first.
As Geeraerts and Grondelaers (1995) point out,
if cognitive models are also cultural models, they are also cultural institutions, and as
such, they carry their history along with them: their institutional nature implies their historical continuity. It is only by investigating their historical origins and their gradual
transformation that their contemporary form can be properly understood. (1995: 177)
124
Axel Hbler
In line with this stance, Arnovick (1999) has examined changes in the conceptualization of promising:
The frequency of promises rendered defective at inception or follow-through has created a practical (if not moral) dilemma, a problem with promises. People do not always
trust the promises they hear. As a result, the need to convince the hearer is often assumed by the speaker. To perform a promise that satisfies the hearer, the speaker often
has to emphasize his or her commitment. (Arnovick 1999: 60)
She notes that even a performative use of promise, which is already a very strong
form of promising (cf. Traugott 1997), is nowadays often felt to be insufficient and
to require some further declarations, such as those observed with George Bush:
Read my lips, I guarantee it, Believe you me. A simple I will nowadays works
only in a promissory situation and an extra-linguistic institution for its force (Arnovick 1999: 58), in contrast to OE times when both the use of sculan (expressing a
subjects obligation) and wilan (expressing a subjects intention) were binding on
the speaker. Nowadays, promises usually undergo relativizations and rationalizations: promises merely express honorable intentions; changing circumstances
can render vows unrealistic; insincerity is politically motivated, deception is committed for the sake of the task pursued. Such views are further warranted by references to independent proposals (by Muck 1989 and Rawls 1955) for constructing
theories of promising that incorporate a limit on speaker responsibility (Arnovick 1999: 60).
The historical and culture-bound relativity of metaphorical concepts on the
pragmatics of language can well be illustrated by two investigations of mine. In
Hbler (1998), I showed that the predominant spatial metaphorization relating to
the notion of express, which represents the first constituent within the conduit
paradigm resulted from a change that had taken place around the turn from the 16th
to the 17th centuries. The argument centered around the change in the use of prepositions going with the notion of express; it was a change from the instrumental
with to the spatial in (words). The former conceptualization aligns well with
and foreshadows the tools-paradigm proposed by Reddy (1979) as a possible alternative to the predominance of the conduit metaphor. And in Hbler (ms), I examined data from the 16th and 17th centuries which show that argument was conceived
during that period in less fierce and more playful terms, more specifically in terms
of a tennis match; Gradys (1998) findings can be taken as a diluted late version of
this early concept.
A discoursal source is tapped on by Gotti (2006). He shows for the 16th/17th
century that neologizing (in contrast to the practice nowadays) was still a matter to
talk and write about; it was the time when a scientific language developed in England, by taking over elements from the common code. The following passage can
be read as an instance of the public discourse about it, particularly about using new
words of foreign origin in English.
Metapragmatics
125
But now as touching myne entent in writing this treatise in the English. Though this
cause might seme sufficient to satisfy many men that I am an englysh man, and therefore
may more easely and plainly write in my natyue tonge, rather then in any other: yet vnto
them that know the hardness of the mater, this answer shuld seme vnlykely: considering
that it is more harder to translate into such a tonge, wherein the arte hath not ben written
before, then to write in those tongues that are accustomed, and (as I might say) acquainted with the termes of the science (Recorde 1547, quoted in Gotti 2006: 218)
Recorde was a physician and mathematician. His topic in this passage (probably
from the preface) is metalinguistic and concerns the (un-)availability of English
words for certain concepts. But the discourse itself could be a case for metapragmatics (in the sense discussed in section 1); the passage is metacommunicative in
that the writer raises the problem of effability in English and thus justifies his use
of un-English words in the treatise to follow. In the current context, however, the
quotation can be seen as documenting metapragmatic knowledge about the limits
and chances of informing an uninformed readership about specific and new states
of affairs. Representing the metapragmatic knowledge of a member of an educated
16th-century English speech community, the document could also give grounds for
comparing it with (in-)effability treatments by speakers from different social-cultural backgrounds; but I do not know of any such investigation.
The developmental type of variation in metapragmatic knowledge is, in
contrast, well investigated. For the purpose of illustration, the article by Bernicot
and Laval (1996) has been chosen, because it focuses, again, on promises. Their
subjects are children of three age groups, around 3, around 6 and around 10 years
of age. The childrens metapragmatic knowledge is defined as their capacity to talk
and think about acts of keeping and breaking promises and tested by examining
their verbal comments stimulated by a series of cartoon stories; each story consists
of a promise (by a child or a parent) and the subsequent fulfilment or non-fulfilment; the effect of sadness (as sign of dissatisfaction) and happiness (as sign of satisfaction) respectively, which the (non-)fulfilment has on the person to whom the
promise was made, has then to be determined by the subjects themselves and subsequently justified. The authors summarize their results thus:
Metapragmatic knowledge was found to evolve with age. At the ages of 3 and 6,
childrens metapragmatic knowledge mainly concerns the execution of the action. At
age 6, the listeners desires start being added in cases where [this] preparatory condition
is not satisfied. At age 10, explanations pertaining to action accomplishment completely
disappear, and explanations about the speakers intentions alone or about both the
speakers intentions and the listeners desires appear. (Bernicot and Laval 1996: 120)
The metapragmatic knowledge of adults would at least also comprise the sincerity
condition of promises, i.e., that the speaker intends to accomplish the future action
(cf. Gibbs and Delaney 1987).
126
3.2.
Axel Hbler
Metapragmatics as metatheory of pragmatics
So far, we have dealt with cases where speakers provide the data/material, which
linguists then process, refine (as was the case with metacommunication) or elaborate on (as in the case of speakers theoremes). We now leave the sphere where linguists occupy themselves with other peoples products and turn to their own activity, subjecting it to reflection.24 That portion of such reflections which becomes
manifest constitutes a professional theoretical dialogue (mainly in the written
medium) about pragmatics, with the aim of defining the field, supporting, supplementing, or justifying certain theoretical stances or challenging them. Such theoretical discussions show, in principle, two orientations. They either operate introspectively or involve external considerations, and sometimes these perspectives
merge. Since pragmatics as such is already the most complex language-related discipline, it is easy to understand that the meta-discourse about it is immense and
open-ended; problems abound. The few topics that will subsequently be addressed
can, therefore, do nothing else but give some indications.
The problems already start on the most elementary of levels, including even the
question what pragmatics is all about. What, in other words, is the object of pragmatics? Is it human communication or just the verbal part of it? The answer may
depend on the answer given to an even more elementary question: How to define
the scope of activity for a linguist working in the field of pragmatics? Is it the discipline that defines the scope or is it the object as such? In that particular case,
pragmatics originated in philosophy but was subsequently fully absorbed by linguistics; should it therefore remain a linguistic discipline and restrict its activities
accordingly? According to what? Was it not one of the major figures of linguistics,
de Saussure, who embedded linguistics in semiotics, which consequently would
license any occupation with the non-verbal as well? If, on the other hand, it is the
object which defines for scholars of pragmatics the scope of their activities, then
the non-verbal with its phonetic and kinesic domains would be included anyway
because natural communication, even any single utterance, can only be fully
understood as a trimodal phenomenon. It is not even justifiable to claim that the
verbal part plays the main role because intonation and/or gestures oftentimes modify in an important way what is verbally said; they sometimes even override the
verbally mediated meaning. Even if one were to subscribe to the view that, ideally
speaking, the nonverbal modes should be attended to as well, it could still be argued that such a holistic approach is simply not practicable; several disciplines
should therefore participate and join in a common endeavor. But how much compartmentalizing does the subject matter tolerate? Interdisciplinary research may
be a good answer, but transdisciplinarity may still be unavoidable up to a certain
degree.
Linguistics and communication research did not fall out of the sky; its predecessor was rhetoric, the art of persuasion. Traces of this tradition of thinking can
Metapragmatics
127
still be found in Searle (1969) in that he keeps the link alive that exists between a
speech act and its effect, the perlocution. Though the ensuing research paradigm
widened the scope so as to treat the speech act as part of speech act sequences, it
may cause wonder why perlocution never played a role. Why has research in indirect communication strategies been restricted to what the speaker plots with the
addressee underdetermined; why has research not systematically investigated the
(perlocutionary) effects that certain strategies are likely to cause in receivers? Because social psychology is doing this job? At some US universities, departments of
Communication (not tantamount to mass communication) have been established
whose activities are at the cross-roads of various lines of research, including linguistics, rhetoric and persuasion, and social psychology. Is that a more adequate
discipline for coming to terms with interpersonal communication? Or what do they
lack that the (European) language-biased approach has to offer?
Questions of involving other disciplines do not only represent boundary issues (Caffi); they may also be raised and pursued with pleasure when they help
dignify central views or important findings of ones own discipline. The dignifying effect results from their proving that the views or results are not isolated, but
align well with what is considered to be more fundamental in the sense that it provides the foundations of ones own discipline, its overarching rationale. A good
case in point is the ever growing theoretical horizon associated with pragmatics.
The cognitive turn in the humanities necessitated a cognitive foundation also for
pragmatics. Since the cognitive sciences, however, have (neuro-)biological foundations, to what degree are they to be paid tribute? Is it enough to allude every now
and then to such deep connections? Is it acceptable to do so only when it seems opportune, i.e., when it serves the argument at issue, and neglect them otherwise?
McNeills treatment of (everyday conversational) narratives provides an interesting example. A central part of his production theory, which (to varying degrees)
takes all three modes of realization into consideration (i.e., language, intonation,
and gestures), operates with a model of self-organization that obviously has its
roots in the theory of autopoietic (or self-referential, self-organizing) systems,
which goes back to Maturana (1970a, 1970b). McNeill, however, does not discuss
(metapragmatically!) these roots. Was it because the theory is difficult for students
of interpersonal communication to digest?25 Where are the limits, and how far do
we have to go, in how much detail?
Many of the questions raised (as well as others that could be raised) still await
answers. Whatever the answers are or will be, the fundamental nature of these issues (and even the questions themselves) are rarely free of any ideological bias,
whether one wants to admit it or not. A sensibility for the ideological dimension of
research in pragmatics cannot be taken for granted; it seems, in fact, that it still
needs to be called in. [A] constant monitoring of linguistic rhetoric in view of the
ideological underpinnings of theories and analyses is not a luxury but a prerequisite for the advancement of linguistic pragmatics (Verschueren 1998: 67). How
128
Axel Hbler
subtle the dangers of ideological pitfalls can sometimes be is well highlighted by

Caffi.
[O]ne has to avoid the risk of new forms of hypostasis [], brought about by certain,
apparently harmless, metalinguistic uses. For instance, the use of a definite noun phrase
can be idealizing: the risk of an idealization exists whenever one refers to the speaker,
who could end up by being an hyperuranian subject, or a lunar effet-sujet. (Caffi 2006:
83)
But more tangible issues have, of course, to be subjected to such self-examination

as well. What is the researchers stance to the speech community that s/he wants to
investigate about, even towards the individual informant? Does the stance have
any political implications? And on a methodological level, are the questions asked
(orally or by means of a questionnaire) politically correct?
A subspecies of this kind of metatheoretical problem consists in the question of
whether the methods of tackling pragmatically relevant data is neutral enough to
cope with different yet comparable sets of data. The problem of culturally biased
approaches is addressed here as well. Among the English speech act verbs, for
example, we find none that would have lexicalized any specification as to time and
location (as in Tenejapa Tzeltal, where concepts like talk that has taken place just
a short time ago or talk occurring in a grassy area are lexicalized), but we do find
English speech act verbs of directing, such as appoint, convene or ordain, which
include specifications as to social setting and institution (cf. Verschueren 1987:
131133). A simple speech act classification would not do because it would neglect the specificities. In the face of such shortcomings, Verschueren proposes to
avoid attempts at classifying linguistic action verbs (and the corresponding acts)
and [] concentrate instead on the careful contrastive analysis of the semantic dimensions involved in the lexicalization of linguistic action (Verschueren 1987:
126). The descriptive framework he offers is supposed to allow for accommodating
(by way of comparing) intra-linguistic, cross-linguistic and cross-cultural differences.
A (partial) translation of Verschuerens program into action is Meys analysis
of a (by now familiar) speech act, promise.
[T]he context in which a promise is made is of the utmost importance for its status as a
promise and for its binding effects. Take the case of a young person promising his or her
parents not to smoke before the age of eighteen. In this case, the societal conditions surrounding the execution of such a promise can be exceedingly difficult (peer-group
pressure, work conditions, etc.). (Mey 1993: 170)
But it serves the author to underline a metapragmatic argument: the view that a
philosophical speech act theory la Austin (21990) and Searle (1969) may not suffice for linguistic purposes.
Both Austin and Searle operate on the one sentence one case principle. That is to say,
in order to illustrate their theory, they use examples that are characteristic of what they
Metapragmatics
129
see as the case under discussion, such as, e.g., a particular kind of speech act. I will, in
the following, show the shortcomings of the case approach (Mey 1993: 170).
The example also demonstrates that metapragmatic discourse does not always
have to be the raison detre for an entire (theoretical) text. It often has, rather, a
more limited purpose, i.e., to profile ones own theoretical assumptions against the
backdrop of what is widely held to be the case. This pattern is pervasive and can be
found in almost all scholarly work in whatever discipline; but it is of course only in
pragmatics that such metatheoretical discourse is metapragmatic. The illustration
to follow is taken from a prominent source.
A more radical difference between Grices approach and relevance theory [as proposed
by Sperber and Wilson, A.H.] is this. Grices principle and maxims are norms which
communicators and audience must know in order to communicate adequately. Communicators generally keep to the norms, but may also violate them to achieve particular
effects; and the audience uses its knowledge of the norms in interpreting communicative
behaviour. The principle of relevance, by contrast, is a generalisation about ostensiveinferential communication. Communicators need no more know the principle of relevance to communicate than they need to know the principles of genetics to reproduce
[] (Sperber and Wilson 1986: 162).
The quotation is an excerpt from a short metapragmatic passage of one and a half
pages, where Sperber and Wilson confront their concept of relevance with that of
Grice (e.g. 1975). The metapragmatic discourse in this case is even sparked off by
a metalinguistic problem, i.e., the fact that both parties use one and the same term,
but associate it with different concepts. The passage, however, is not metacommunicative because concepts, methods and the like are the issue, not the communicative form or aspects of it. This is a feature that generally applies to any form in
which pragmatics is theoretically reflected. It is a type of reflection that is at best
accidentally language-reflexive. This sets it apart from all the other phenomena
that have been taken account of in this article and aligns it with the genre metatheory. Metapragmatics may be, in the end, an all too generous term.
Notes
1. On July 31, 2009.
2. This is basically what Caffi (2006) takes account of as metapragmatics 2.
3. Applying the typology of metamessages proposed by Jakobson (1971 [1957] 130133),
we are mainly concerned with type (M/M), i.e., messages about messages. Certain attention will also be given to type (C/M), that is to those metamessages in which so-called
shifters (as the code involved) incorporate a message about a message. Messages about
code (M/C) as well as the meta-message type (C/C), where proper names are the protagonists, fall outside the scope of metacommunication proper. Cf. also Lucy (1993: 16).
4. In section 2, a different concept will be introduced, which conceptualizes metalanguage
as a dimension.
130
Axel Hbler
5. Metalanguage in the sense outlined, however, is part of but not equivalent to reflexive
language as described by Lucy (1993) just as speech reflexive activity/ reflexive language use is not tantamount to metacommunication.
6. Such reduced forms will be tackled separately below.
7. Quite extensive studies along these lines have been conducted earlier by Meyer-Hermann (1980 [unpublished]). Lumping together topic and function, he identified 72
metacommunicative speech act types (on the basis of French and Portuguese material).
The 11 most frequently used types are: requesting the performance of speech acts, acknowledging the performance of speech acts, announcing the performance of speech
acts, describing communicative interaction, guiding discussions, criticizing communicative interaction, restating a communicative function more precisely, eliciting receptive readiness, securing comprehension, evaluating communicative interaction, and anticipating sanctions prophylactically (cf. Welte 1990: 178180). They all have been
included in one way or another in the present outline.
8. Meyer-Hermann (1980), in fact, tackles functions only, the topics being implied.
9. Quoting, in fact, is the phenomenon that Jakobsons meta-message type (M/M) was obviously meant to mainly accommodate.
10. As Wilson (2000: 148) points out, [the] attitudes conveyed by echoic utterances can be
very rich and varied: the speaker may indicate that she agrees or disagrees with the original, is puzzled, angry, amused, intrigued, skeptical etc., or any combination of these.
11. Often the disjunct may have a paraphrase with a verb of saying taking the form of a
finite clause (If I may say so without offence, your writing is immature).
12. Hagemann (1997) conducted a detailed investigation on German diction-characterizing adverbs in Gricean terms.
13. I am not sure about the status of adverbs such as well which can either function as discourse initiators (cf. Quirk et al. 1995) or continuation markers (cf. Halliday and Hasan
1976).
14. The topical distinction as such has been highlighted, under different labels, by various
authors; cf. Bamford and Biondi (2005: XVII), Hagemann (1997: 2728), Geissner
(1981: 206). Watzlawick et al. (1968) and Bateson (1972) focus exclusively on the interpersonal dimension and reserve for this aspect the term metacommunication.
15. In Jakobsons typology (cf. footnote 3), they would represent metamessages of the type
(C/M).
16. Jakobson (1971: 131), who adopted the concept from Jespersen (1922: 123), discusses
shifters at some length.
17. Halliday and Hasans use of the term presupposition differs, of course, from its usual
semantic-pragmatic reading; they use it in the wider sense that something is to be supplied, or understood (Halliday and Hasan 1976: 144).
18. Cf. also analogous term formations such as synsemantic.
19. But caution is advised. Speakers awareness [] does not have to match the linguists
metapragmatic descriptions. Naively confusing these two may contribute to the furthering of a specific kind of folk metapragmatics inherent in western linguistics itself. Assuming that such nave confusion can be avoided, at least some of my further comments
will be based on results obtained in this line of research (Verschueren (1998: 6061).
20. In certain respects complementary, but of a narrower scope, is Schmid (2000) with his
investigation of shell nouns, a subset of which overlaps with notions taken account of
by Ballmer and Brennenstuhl (1981).
Metapragmatics
131
21. For obvious reasons, I simply take both kinds for granted, though it is certainly a matter
worth discussing; cf. Kvecses (2005: 193 ff.) and his discussion of the question
whether cultural models can exist without prior metaphorical understanding, as claimed
by Quinn (1991).
22. When participants report on actual verbal encounters, they tend to do so by mentioning
some item of content, or by referring to what people were getting at (Gumperz 1982:
157).
23. For questioning the parameters, cf. section 3.
24. This overlaps with what Caffi characterizes (under metapragmatics 1) as concern for
the epistemological implications of what researchers assume pragmatics is about (Caffi
2006: 83).
25. The reader may test it, in the original or in a version such as Benseler et al. (eds.) 1980,
which has communication scientists as its target audience.
References
del, Annelie
2006
Metadiscourse in L1 and L2 English. Amsterdam: Benjamins.
Anton, Corey
1998
About talk: The category of talk-reflexive words. Semiotica 121/3,4:
193212.
Arnovick, Leslie K.
1999
Diachronic pragmatics. Seven case studies in English illocutionary development. Amsterdam: Benjamins.
Auer, Peter
1986
Kontextualisierung. Studium Linguistik 19: 2247.
Austin, John L.
1990
How to do things with words. Edited by James O. Urmson and Marina Sbis.
2nd. ed. Oxford: Oxford University Press.
Bakhtin, Michail M.
1981
The dialogic imagination. Edited by Michael Holquist translated by Emerson,
Caryl and Michael Holquist. Austin, TX: University of Texas Press.
Ballmer, Thomas and Waltraud Brennenstuhl
1981
Speech act classification. A study in the lexical analysis of English speech activity verbs. Berlin: Springer.
Bally, Charles
1913 [1965] Le langage et la vie. Genve: Ed. Atar, [reprinted Genve: Librairie Droz].
Bamford, Julia and Marina Biondi
2005
Introduction. In: Julia Bamford and Marina Biondi, Dialogue within discourse
communities. Metadiscursive perspectives on academic genres. Tbingen:
Niemeyer. IXXIV.
Bateson, Gregory
1972 [1955] Steps to an ecology of mind. New York: Balantine.
Benseler, Frank, Peter M. Hejl and Wolfram K. Kck (eds.)
1980
Autopoiesis, communication and society. The theory of autopoietic system in
the social sciences. Frankfurt, New York: Campus.
132
Axel Hbler
Bernicot, Josie and Virginie Laval

1996
Promises in French children: Comprehension and metapragmatic knowledge.
2007
Metapragmatics in use. Amsterdam: Benjamins.
Caffi, Claudia
2006
Metapragmatics. Encyclopedia of Language and Linguistics. Amsterdam:
Elsevier. 8288.
Cameron, Deborah
1995
Verbal hygiene. London: Routledge.
Cameron, Deborah
2000
Good to talk? Living and working in a communication culture. London: Sage.
Coupland, Nikolas
2004
Stylised deception. In: Adam Jaworski, Nikolas Coupland and Dariusz Galasinski (eds.), Metalanguage. Social and ideological perspectives, 249274.
Coupland, Nikolas and Adam Jaworski
2004
Sociolinguistic perspectives on metalanguage: Reflexivity, evaluation and
ideology. In: Adam Jaworski, Nikolas Coupland and Dariusz Galasinski (eds.),
Metalanguage. Social and ideological perspectives, 1551. Berlin: Mouton de
Gruyter.
Fairclough, Norman (ed.)
1992
Critical language awareness. London: Longman.
Gee, James P.
1999
Discourse analysis. London, New York: Routledge.
Geeraerts, Dirk and Stefan Grondelaers
1995
Looking back at anger: Cultural traditions and metaphorical patterns. In: John
R. Taylor and Robert E. MacLaury (eds.), Language and the cognitive construal of the world, 153179. Berlin: Mouton de Gruyter.
Geissner, Hellmut
1981
Sprechwissenschaft. Theorie der mndlichen Kommunikation. Knigstein im
Taunus: Scriptor.
Gibbs, Raymond W. and Suzanne M. Delaney
1987
Pragmatic factors in making and understanding promises. Discourse Processes
10: 107126.
Goffman, Erving
1972
Relations in public: Microstudies of the public order. New York: Harper and
Row.
Goossens, L.
1987
Dealing with linguistic action verbs in depth. In: Jef Verschueren (ed.), Linguistic action: Some empirical-conceptual studies. Norwood, NJ: Ablex.
95106.
Gotti, Maurizio
2006
Metalinguistic considerations in 17th-century specialized discourse. In: Anna
Duszak and Urszula Okulska (eds.), Bridges and barriers in metalinguistic
discourse, 217231. Frankfurt: Peter Lang.
Metapragmatics
133
Grady, Joseph
1998
The Conduit metaphor revisited: A reassessment of metaphors for communication. In: Jean-Pierre Koenig (ed.), Discourse and cognition: Bridging the
gap. Stanford, CA: CSLI Publications. 205218.
Greatbatch, David and Robert Dingwall
1998
Talk and identity in divorce mediation. In: Charles Antaki and Sue Widdicombe (eds.), Identities in talk, 121132. London: Sage.
Graf, Claudia
2007
Echo und Zitat. In: Elke Brendel, Jrg Maibauer and Markus Steinbach (eds.),
Zitat und Bedeutung. Linguistische Berichte, Sonderheft 15: 89109.
Grice, H. Paul
1975
semantics. Vol. 3: Speech acts. New York: Academic Press.
Gumperz, John J.
1982
Discourse strategies. Cambridge: Cambridge University Press.
Habermas, Jrgen
1971
Vorbereitende Bemerkungen zu einer Theorie der kommunikativen Kompetenz. In: Jrgen Habermas and Niklas Luhmann, Theorie der Gesellschaft oder
Sozialtechnologie, 101141. Frankfurt: Suhrkamp.
Hagemann, Jrg
1997
Reflexiver Sprachgebrauch: Diktumscharakterisierung aus Gricescher Sicht.
Halliday, Michael A.K. and Ruquaia Hasan
1976
Hbler, Axel
1998
The expressivity of grammar. Grammatical devices expressing emotion across
time. Berlin: Mouton de Gruyter.
Hbler, Axel
o.J.
The conceptualization of argument in terms other than war. A socio-cultural
case study. (Ms)
Hbler, Axel and Wolfram Bublitz
2007
Introducing metapragmatics in use. In: Wolfram Bublitz and Axel Hbler
(eds.), Metapragmatics in use, 126. Amsterdam: Benjamins.
Hyland, Ken
2005
Metadiscourse. Exploring interaction in writing. London, New York: Continuum.
Jakobson, Roman
1971 [1957] Shifters, verbal categories, and the Russian verb. In: Selected writings II:
Word and language, 130147. The Hague: Mouton.
Jespersen, Otto
1922
Kvesces, Zoltan
2005
Metaphor in culture. Universality and variation. Cambridge: Cambridge University Press.
1980
Metaphors we live by. Chicago: The University of Chicago Press.
134
Axel Hbler
Leder, Drew
1990
The absent body. Chicago: The University of Chicago Press.
Lucy, John A.
1993
Reflexive language and the human disciplines. In: John A. Lucy (ed.), Reflexive language. Reported speech and metapragmatics. Cambridge: Cambridge
University Press. 932.
1970a
Biology of cognition. Report 9. Biological Computer Laboratory, Department
of Electrical Engineering. University of Illinois, Ill.
1970b
Neurophysiology of cognition. In: Paul L. Garvin (ed.), Cognition: A multiple
view, 323. New York: Spartan.
McNeill, David
1992
Hand and mind. What gestures reveal about thought. Chicago, London: The
Mey, Jacob L.
1993
Pragmatics. An introduction. Oxford, Cambridge: Blackwell.
1978
Aspekte der Analyse metakommunikativer Interaktionen. In: Reinhard MeyerHermann (ed.), Sprechen Handeln Interaktion, 103142. Tbingen: Niemeyer.
1980
Studien zur Funktion von Metakommunikation am Beispiel portugiesischer
und franzsischer Alltagssprache. Habilitationsschrift Universitt Bielefeld.
(Ms.)
Muck, Bruce
1989
Promising rules and intentions. PhD Diss. University of Wisconsin.
Niedzielski, Nancy A. and Dennis R. Preston
2000
Folk linguistics. Berlin: Mouton de Gruyter.
Paul, Ingwer
1999
Praktische Sprachreflexion. Tbingen: Niemeyer.
Preston, Dennis R.
1996
Whaddayaknow? The modes of folk linguistic awareness. Language Awareness 5: 4077.
Quinn, Naomi
1991
The cultural basis of metaphor. In: James Fernandez (ed.), Beyond metaphor:
The theory of tropes in anthropology. Stanford, CA: Stanford University Press.
5693.
Quirk, Randoph, Sidney Greenbaum, Geoffrey Leech and Jan Svartvik
1985
A comprehensive grammar of the English language. London, New York: Longman.
Rawls, John
1955
Two concepts of rules. Philosophical Review 64/1: 332.
Reddy, Michael
1979
The conduit metaphor: a case of frame conflict in our language about language.
In: Andrew Ortony (ed.), Metaphor and thought, 164201. Cambridge: Cambridge University Press.
Metapragmatics
135
Schmid, Hans-Jrg
2000
English abstract nouns as conceptual shells. Berlin: de Gruyter.
Searle, John R.
1969
Semino, Elena
2006
A corpus-based study of metaphors for speech activity in British English. In:
Anatol Stefanowitsch and Stefan T. Gries (eds.), Corpus-based approaches to
metaphor and metonymy, 3662. Berlin: Mouton de Gruyter.
1993
Reflexive language, 3358. Cambridge: Cambridge University Press.
Simon-Vandenbergen, Anne-Marie
1995
Assessing linguistic behaviour. A study of value judgements. In: Louis
Goossens, Paul Pauwels, Brygida Rudzka-Ostyn, Anne-Marie Simon-Vandenbergen and Johan Vanparys (eds.), By word of mouth, 71124. Amsterdam:
Benjamins.
Smith, Bruce L., Bruce L. Brown, William J. Strong and Alvin C. Rencher
1975
Effects of speech rate on personality perception. Language and Speech 18:
145152.
1986
Relevance. Communication and cognition. Oxford: Blackwell.
Tajfel, Henri
1982
Social identity and intergroup relations. Cambridge: Cambridge University
Press.
Traugott, Elizabeth
1997
Subjectivication and the development of epistemic meaning: The case of
promise and threaten. In: Toril Swan and Olaf J. Westvik (eds.), Modality in
Germanic languages: Historical and comparative perspectives, 185210. Berlin: de Gruyter.
Ungeheuer, Gerold
1970
Kommunikative und extrakommunikative Betrachtungsweisen in der Phonetik. In: Proceedings of the Sixth International Congress of Phonetic Sciences,
7386. (Prag 1967). Prague.
Vanparys, Johan
1995
A survey of metalinguistic metaphors. In: Louis Goossens, Paul Pauwels,
Brygida Rudzka-Ostyn, Anne-Marie Simon-Vandenbergen and Johan Vanparys (eds.), By word of mouth, 134. Amsterdam: Benjamins.
Verschueren, Jef
1987
Metapragmatics and universals of linguistic action. In: Jef Verschueren (ed.),
Linguistic action. Some empirical-conceptual studie,125147. Norwood, NJ:
Ablex.
Verschueren, Jef
1998
Notes on the role of metapragmatic awareness in language use. In: Adam
Jaworski, Nikolas Coupland and Dariusz Galasinski (eds.), Metalanguage.
Social and ideological perspectives, 5373. Berlin: Mouton de Gruyter.
Watzlawick, Paul, Janet H. Beavin and Don D. Jackson
1968
Pragmatics of human communication: A study of interactional patterns, pathologies, and paradoxes. London: Faber and Faber.
136
Axel Hbler
Welte, Werner
1990
Alltagssprachliche Metakommunikation im Englischen und Deutschen. Frankfurt: Lang.
Wilson, Deirdre
2000
Metarepresentation in linguistic communication. In: UCLWPL 11: 127161.
Wooffitt, Robin and Clark Colin
1998
Mobilizing discourse and social identities in knowledge talk. In: Charles
Antaki and Sue Widdicombe (eds.), Identities in talk, 107120. London: Sage.
Part II
Theoretical foundations
5.
The rise of pragmatics: a historiographic

overview
Wataru Koyama
1.
Introduction
This article investigates the theoretical, philosophical, social institutional, and interpersonal genealogies of several key concepts of contemporary linguistic pragmatics, such as presupposition, implicature, speech act, and performativity, as well
as perhaps more fundamental ones, viz., context, indexicality, function, referential
and non-referential practice, metapragmatics, and reflexivity. More specifically,
after briefly dealing with the genealogy of the concepts context and indexicality in Section 2, the article discusses the genealogies of presupposition and
effect (presupposing and creative functions) in Section 3, function (multifunctionality) in Sections 4, 5, and 6, and performativity, metapragmatics, and
reflexivity in Section 7.
In so doing, this article tries to articulate a historiographic overview of how linguistic pragmatics was established and developed before it became an institutionalized international discipline. It specifically focuses on two critical developments since the late nineteenth century that led to the rise of pragmatics in the
1970s: i.e., the linguistic turn, to be followed by the pragmatic turn. The
former, which took place from the 1880s to the 1930s, gave rise to structural-functionalist linguistics. This historical transformation was closely linked to the neoKantian movement, marking the empirical turn of Kantian philosophy, and the
phenomenological movement, focusing on the synchronic here and now of communicative events (see Sections 5 and 6). On the other hand, the linguistic turn also
led to the emergence of analytic philosophy, with its focus on the propositional,
logical-semantic aspects of language. Structural linguistics, interacting with analytic philosophy, took the formalist turn in the 1940s, turning away from context
and (communicative) function and moving to concentrate on the (morpho)phonological, (morpho)syntactic, and semantic aspects of language. Gradually, however,
especially in the 1950s and 60s, there was increasing awareness of the importance
of context, often discussed under the rubrics of indexicality and pragmatics, even
in logico-analytic philosophy (see Section 2). This led to the pragmatic turn and the
institutional establishment of linguistic pragmatics in the 70s, linguistics and philosophy moving from the analytically-oriented study of de-contextualizable regularities of language toward the empirical study of the contextualized use of language. Thus, since the 1970s, the theoretical notions articulated or suggested by
the older traditions of structural-functional linguistics, such as context, indexical-
140
Wataru Koyama
ity, function, and metalanguage, have been re-discovered and further explored. In
addition, the pragmatic turn has allowed the convergence of pragmatics with the
empirical social sciences, which are traditionally concerned with actions and interpretations in context. As is described in Section 3, this convergence characterizes
the current state of linguistic pragmatics, where the earlier emphasis on regularities
in language use is increasingly supplemented with the focus on the pragmatically
achieved (trans)formation of group identities and power-relations of language
users in specific sociocultural contexts.
In such a way, this article tries to present a historiographic account showing
that the true home of linguistic pragmatics lies in the Continental tradition (cf.
Huang 2007: 45), which tries to provide a contextual, pragmatic view of language
and communication in our life-worlds, not in the Anglo-American tradition,
which sees pragmatics as a narrowly delimited component supplementary to the
ahistorical, universalizing studies of syntax and semantics, abstractable from referential regularities. As this article will demonstrate, pragmatics came out of the
Continental tradition in the late nineteenth century and, after the formalist lapse at
the mid-twentieth century, the discipline has been moving back from the AngloAmerican to the Continental tradition, further exploring referential and non-referential practices, i.e., what we do, with or without words, in our life-worlds (see
Section 5 for details).
2.
Genealogy of the concepts context and indexicality
Let us begin our historiographic reconstruction with one of the most basic notions
of pragmatics, namely, context, which is closely related to the phenomenological
notion of life-world and which includes not only the tokens of linguistic types
but also communicative-event participants and their pragmatic acts (cf. Mey
[1993] 2001: 206235). It would not be too far-fetched to say that pragmatic
means contextual(ized), as the discipline itself is differentiated by its focus on
context, from structural linguistics and semantic logic, which are mainly concerned with de-contextualized regularities abstractable from referential practices.
Clearly, the historiography of pragmatics must be written around this theme: how
context has been discovered as a theoretical topic, and theoretically dealt with (cf.
Nerlich and Clarke 1996).
Of course, all empirical studies of language or communication have been
at least implicitly concerned with context, so that we might argue that modern
theorization on context(uality) started with the emergence of modern empiricism,
as suggested by John Lockes (16321704) treatise on human understanding
and communication (cf. Locke [1689] 1975), which anticipated Thomas Reids
(171096) pragmatic commonsense philosophy and thus paved the way for the
development of modern semiotics, pragmati(ci)sm, and pragmatics.1
141
Yet, from the linguistic (structural) point of view, which focuses on de-contextualizable linguistic categories, it is, of course, personal and demonstrative pronouns, moods, modalities, tenses, and other deictic categories (deixis) or indexical
types that most clearly show contextuality, often called context-dependency (cf.
Fillmore 1975; Levinson 1983: 5496); hence, the specifically linguistic origins
of todays linguistic pragmatics might be found in empirical investigations and
theorizations of the deictic categories. Thus, the roots of linguistic pragmatics
may be sought, as Nerlich and Clarke (1996) have done, in the structural-functionalist traditions: i.e., the (Austro-)German tradition, which led to Karl Bhlers
(18791963) theory of indexical expressions (Zeigwrter), the deictic field (Zeigfeld; vs. Symbolfeld), and Organon model, developed in 193334; the French
one, culminating in Benvenistes (190276) theory of pronouns, subjectivity, and
smantique, or what would become known as nunciation (Benveniste, 1966); and
the British tradition, which eventually led to Lyonss (1977) semantics (in the
broad sense, including linguistic pragmatics) and Hallidays text linguistics, the
latter synthesizing the Firthian theory (London School) and the Praguean theory
of referential (informational) communication and multi-functionalism (see Section 6).2
We shall follow these functionalist movements in Europe, anticipating the
worldwide rise of pragmatics in the latter half of the twentieth century, later in this
article (see Sections 46). Here, we should take a brief look at the development on
the other side of the Atlantic. That is, in the United States, the Praguean functionalist theory became combined with the Peircean theory of semiotic pragmati(ci)sm
by Jakobson (18961982), who articulated a contextually-anchored, that is, speech
event-centered account of grammatical categories, including what Bally called
actualisateurs or Jespersen (18601943) named shifters (Jespersen 1922),
which Jakobson (1957) characterized as denotational-indexicals. As is well
known, the notion of indexicality came from Peirce (see note 1), who defined it as
the kind of relationship between signs (e.g., tokens of linguistic types) and their
objects (e.g., contextual referents) that is pragmatically motivated by the empirical
principle of contiguity in the context of signification (communication). Note that
indexicality is a general semiotic notion, characterizing not only deixis, but
also pointing gestures, gazes, spatial arrangements (e.g., proximity), and any other
modes of communicative practices based on the principle of contiguity in the contexts of communicative events (cf. Peirce [18931910] 1955). Clearly, indexicality
is a key to the proper theorization of context and contextualized communication, as
duly recognized by the micro-sociologist (ethnomethodologist) Garfinkel (see
Section 3) and, later, the social pragmatician Mey ([1993] 2001: 5260).
The significance of indexicality for pragmatics had been also acknowledged, in
a different disciplinary context, in Bar-Hillel (1954), which would lead to Richard
Montagues (193071) attempt at a formal pragmatics. As this indicates, Bar-Hillel
was an analytic philosopher generally working in the logical syntax framework
142
Wataru Koyama
of Rudolf Carnap (18911970), closely associated with the behavioristic pragmatist Charles Morris (190179), i.e., two figures involved in the logical-empiricist
project of Unified Science (Einheitswissenschaft), or the Unity of Science Movement, led by Otto Neurath (18821945; a member of the Vienna Circle), in the
early-to-mid-twentieth century (cf. Carnap [1934] 1937; Morris 1938). In particular, Morris, partially following Peirces semiotic pragmati(ci)sm (see immediately
above), had articulated, in 1937, the famous semiotic triad of syntax, semantics,
and pragmatics,3 to be adopted by Carnap. Coming out of such a tradition, Bar-Hillels work on indexicality, pragmatics (as wastebasket of syntax and semantics),
and implying (what Grice would later call implicature) in the 1940s and 1950s
(cf. Bar-Hillel 1946; Chapman 2005: 92) anticipated, in retrospect, not only Montague and Grice, but also the pragmatic turn in analytic philosophy in the 1960s
and 70s, which saw Kripkes (1972) theory of rigid designation, causal theory
of reference on proper names (cf. Putnam 1975), speakers reference (true reference; Donnellan 1966, 1978; Kripke 1979), and Kaplans (1978) theory of
Dthat on demonstratives. With these developments, by which analytic philosophy became increasingly focused on indexicality and pragmatics, the discipline
turned post-analytic and came to merge with neo-pragmatism in the 1970s. Also,
whereas the Carnapian logical syntax tradition was associated with neo-Bloomfieldian structural linguistics and its derivatives like Chomskyan generativism, the
newer, post-analytic developments became associated with Generative Semantics
and the emerging discipline of linguistic pragmatics in the 1970s (cf. Donnellan
1978; Davis 1991).
Thus, indexicality, along with its associated concepts, has been a key element
for the constitution of pragmatics and the proper theorization of context. Further,
the Peircean notion of indexicality and Bhlers ideas of deictic field (Zeigfeld) and deictic center (origo) became theoretically combined in the linguisticanthropological tradition, derived from not only Boas (18581942) and Sapir
(18841939), but also Peirce, Jakobson, Benveniste, and Kuryowicz (1895
1978), and led by a disciple of Jakobsons: namely, Silverstein (b. 1945). As documented by Pressmans (1994) contribution to the historiography of pragmatics
(also, Caton 1987), Silverstein (1976a) demonstrated, in his famous Jakobsonian
work on case-marking and the Noun Phrase (NP) hierarchy, that grammatical categories are universally, systematically anchored onto the deictic center, located at
the center of the context of communicative interaction (i.e., Jakobsons (1957)
speech event). Further, Silverstein (1976b) made clear that this pragmatic, indexicality-based theory of linguistic structure was part of the general linguistic-anthropological theory of sociocultural communication, covering not only linguistic
structure, but also referential and non-referential communicative practices, the
latter pertaining to social indexicality, or the indexing of the group identities and
power-relations of communicative-event participants. Thus, indexicality has been
shown to be a key concept in the pragmatic study of contextuality, across the struc-
143
tural, referential, and non-referential domains of language and communication (see

Section 6, below, for more on indexicality; cf. Hanks 1990; Duranti and Goodwin
1992; Agha 2007).
3.
Genealogy of the concepts (contextual) presupposition and effect
Having briefly seen how the notions of indexicality and deictic center have
been made use of to account for the communicative event-centered constitution of
context and linguistic structure, we now move to explore historical developments
in the theoretical understanding about pragmatic processes that take place in, and
(re-)create, context. Here, it appears that the intuitive notions of situational (contextual) appropriateness and effectiveness (or efficacy) of actions have led to,
and still lie behind, more elaborate theorizations of this theme, in terms of, for
example, presupposition and perlocution(ary effect), respectively.
Let us first deal with theoretical notions related to contextual appropriateness.
As noted immediately above, the concept of presupposition, be it semantic or
pragmatic (cf. Karttunen 1974), is a theoretical elaboration of the intuitive notion
appropriateness. The concept was developed by Frege (18481925; see Section 5) and Russell (18721970; a leading figure in the logicist tradition of British
analytic philosophy), and was later picked up by Peter Strawson (19192006; a
major figure in the Oxford Ordinary Language School). Subsequently, the
schools theoretical framework and concepts were discovered and integrated into
linguistics in the late 1960s and 70s by (post-Chomskyan) Generative Semanticists: viz., George Lakoff, McCawley, Robin Lakoff, Ross, Sadock, et al. (cf. Lakoff 1989; Harris 1993; Koyama 2000). As these were immediate predecessors of
linguistic pragmatics (which became established in the 1970s; cf. Mey [1993]
2001: 1927),4 presupposition became a main theme in the new discipline, especially in the 1970s and early 1980s (cf. Gazdar 1979: 89128; Levinson 1983:
167225). Generally speaking, since the 1970s, the studies of presupposition have
been characterized by the increasing focus on pragmatic, rather than semantic, presupposition and the increasing awareness of the centrality of pragmatic context in
matters of presupposition in general.
Similarly, the notion of felicity condition, apparently concerning contextual
appropriateness and derived from the works of Austin (191160) and Searle (b.
1932), became a prominent theme in the new discipline. (Austin was the leading
figure of the Oxford philosophy of language, and Searle was an American-born
student of Austin and Strawson at Oxford in the 1950s). As Austin must have been
at least dimly aware, the theme of felicity condition, which concerns the contextual conditions that must obtain in order for a (speech) act to occur validly, was
actually an old theme in law (see note 7), and, we should also observe, in other disciplines concerned with (judicial or other) rituals, such as sociology and anthropol-
144
Wataru Koyama
ogy, which had the long tradition of investigating the social or cultural (contextual)
appropriateness of rituals and other actions (see Section 7). Thus, in the 1970s and
early 80s, felicity condition, an old theme recently re-thematized by Austins
disciple Searle, became a topic not only in the new discipline of pragmatics, but
also in the social sciences (see immediately below), thereby facilitating the interaction and partial merger between pragmatics, micro-sociology, and linguistic
anthropology. In other words, it appears that this theme may have helped the new
discipline to move in the direction of sociocultural: that is, toward the microsociology of conversation analysis (cf. Levinson 1983: 284370) and, beyond
that, linguistic anthropology, social semiotics, social pragmatics, critical pragmatics, and critical discourse analysis, which have been gaining prominence in
pragmatics since the late 1970s and especially 80s and 90s (cf. Mey [1993]
2001). Note that, in the 1970s and early 80s, when felicity condition was hotly
discussed, the micro-sociologist Goffman (192282) wrote an important article on
it (Goffman 1983), and a variety of concepts coming out of his uvre, such as
face (cf. Brown and Levinson [1978] 1987), frame (cf. Tannen 1986), and
footing (cf. Irvine 1996), began to exercise great influence among pragmaticians, discourse analysts, and linguistic anthropologists (cf. Goffman 1967, 1974,
1981, 1983). On the other hand, in the same period, the more individual-psychological aspects of felicity condition, such as the speakers intention and sincerity,
articulated by Searle, became critically examined from cross-cultural perspectives
by linguistic anthropologists (cf. Rosaldo 1982; Du Bois 1987; Duranti 1993),
working in the tradition of ethnography of speaking/communication (see Section 6, below, for details; cf. Gumperz and Hymes [1972] 1986; Bauman and
Sherzer [1974] 1989). This move anticipated the increasing appreciation of crosscultural differences in pragmatics especially since the 1980s (cf. Mey [1993] 2001:
262288).
Indeed, the social anthropological concern with situational (contextual) appropriateness of actions goes back to a founder of British social anthropology,
Malinowski (18841942), who advanced the notions of context of situation
and context of culture,5 adopted in the 1930s by his close associate, J. R. Firth
(18901960), i.e., the leading figure of the London school of functional linguistics.
The modern anthropological idea, advanced by Malinowski, Radcliff-Brown
(18811955), and Boas (18581942) in the early twentieth century, that any action,
be it linguistic or non-linguistic, should be understood in relation to the situational
or cultural context in which the action takes place, was intertwined with the anthropological doctrine of cultural relativism, which urged that actions be seen and
understood from natives point of view, from within, or emically in the terminology of Pike (1967: 3772). Beyond the disciplinary matrix of twentieth-century social/cultural anthropology, the idea of the cultural relativity of actions, especially linguistic actions, was also shared by Wittgenstein (18891951) and Peter
Winch (192697), the former sometimes considered a precursor of the Ordinary
145
Language School, including Austin, Grice, and Searle, and subsequently, linguistic
pragmatics (cf. Levinson 1983: 227).
Importantly, the (situational or cultural) contextualist doctrine, advanced by
modern anthropology and later adopted by much of linguistic pragmatics, not only
emphasized the role of context in determining the meanings of actions, but also
tended to take context to be an independent macro-social variable determining the
meanings of micro-social actions, the latter considered as dependent variables. In
other words, the doctrine failed to pay sufficient attention to how actions create
and change the contexts (situations and cultures) in which they take place. This
static view of culture was perhaps most pronounced in Radcliff-Browns Durkheimian functionalism, where the meanings of actions were explained by the
(macro-)social functions they served. In this view, which came close to macrosocial determinism, the meanings of micro-social actions (what we do, with or
without words) were equated with the functions they served in the macro-social
order (contextual reality) presupposed to exist independently of actions.
To be sure, in Malinowskis functionalism, which was less Durkheimian and
more Freudian than Radliff-Browns (cf. Kuklick 1991), it was not the macro-social order, but the psychological needs of socially contextualized individuals that
were taken as presupposed variables, teleologically explaining the meanings
(functions) of their actions; all the same, the meanings of actions were functionally explained by the conformity (fit, or appropriate correspondence) of the actions with their presupposed ends, be they psychological or (macro)sociological.
The functionalist doctrine was espoused by much of linguistic functionalism
and the nascent discipline of linguistic pragmatics in the 1970s. For instance, in
these disciplines, actions (utterances, or speech acts) were often explained or theorized by appealing to the notions such as (social or psychosocial) maxims and
conditions, which were, as in Malinowskian functionalism, taken to be independent variables explaining the meanings (implicatures) or efficacy of utterances and
other practices. This functionalist way of seeing speech and other social actions
had been dominant in the post-World War Two behavioral sciences, which became
challenged in the social sciences in the 1960s and 70s, when the idea that society
and culture are contingently (re-)created in and by daily actions and, thus, can be
changed by actions, gained momentum, especially in the anti-establishment,
youth, peace, and (second-wave) feminist movements. In the social sciences, even
before the 1960s, the ethnomethodologist Harold Garfinkel (b. 1917) had begun
frontally to attack the functionalist doctrine, attributed to Durkheim (18581917),
and explore how micro-social, contextualized actions create, and indeed lie at the
foundation of, social organization (cf. Heritage 1984). Interacting with the microsociological studies carried out by Goffman (mentioned above, in this Section),
ethnomethodology subsequently saw a new development called conversation
analysis, explored by Sacks, Jefferson, Schegloff, and their followers, and influencing pragmatics (cf. Levinson 1983: 284370). Meanwhile, since the 1970s,
146
Wataru Koyama
second-wave feminism, critical of the functionalist view of society, has given rise
to womens studies, gender studies, and later, queer studies, a series of developments in which feminist linguistics/pragmatics have been involved (cf. Lakoff
1973; Cameron and Kulick 2003), and in which the term performativity, taken
from Austin, has played a crucial role (cf. Butler 1990). As this indicates, the
speech act theories of Oxonian philosophy and early pragmatics were not exclusively concerned with the question of contextual appropriateness. Rather, they investigated conditions, maxims, and other presupposed contextual variables, at
least partially so as to explain the efficacy, or the creative (performative) function,
of actions, the intuitive appreciation of which was, indeed, the point of departure
for them. And, it was this insight of performativity that attracted theoretical interests of many literary critics, historians, social scientists, and feminists, to pragmatics (see Section 7 for more on performativity). Thus, interacting with them,
pragmaticians moved to construct pragmatic theories focusing on the creative,
rather than presupposing, function of language use, especially as related to powerrelations and group identities of language users. As a result, not only feminist pragmatics, but also critical linguistics/pragmatics, associated with the Frankfurt
School, the New Left, and neo-Marxism, came into being in the late 1970s (cf. Mey
1979, 1985, [1993] 2001), in critical opposition to the functionalist view of society.
Eventually, the 1980s and 90s saw the rise of critical discourse analysis in Western
and Central Europe. Even sociolinguistics, in the wake of criticisms against the
functionalist doctrine (cf. Williams 1992), has begun to move in the direction of
social constructivism, investigating how social realities and systems are dynamically created by discursive interactions (cf. Coupland, Sarangi, and Candlin 2001).
4.
Genealogy of the idea (communicative) functions

(multi-functionality)
Thus, appropriateness and effectiveness, that is, the presupposing and creative
functions of interactional practices, constitute one dimension of pragmatic processes in context, around which the history of pragmatics can be described. Crosscutting this is another dimension of pragmatic processes in context: viz., the dimension of what one might call the functional (indexical) orientations of
communication, such as the referential (descriptive), expressive (emotional), and
appeal functions (cf. Bhler [1934] 1990). Jakobsons (1960) six-functions
model, one of the most solid models of the multi-functionality of communication,
identified the emotive, conative, referential, metalingual (metalinguistic), poetic,
and phatic functions, respectively characterized by the six factors constituting
the event of communicative practice, i.e., addresser, addressee, referent, code,
message, and channel (contact; see Section 6 for more details). Indeed, the term
function has been usually used in this sense in linguistic functionalism and prag-
147
matics, and the history of pragmatics has been often written by tracing the history
of studies dealing with these functions in linguistics, philosophy, psychology, law,
and the social sciences (cf. Nerlich and Clarke 1996). Although Austins theory of
speech act seems to have arisen more or less independently of these functionalist
studies, which had blossomed from the 1880s to the 1930s (see Section 5), it is
clear that they anticipated the basic insight of Wittgenstein and the Ordinary Language School (including Austin, Warnock, Urmson, Grice, Strawson), to be absorbed by Generative Semantics and linguistic pragmatics from the late 1960s onward, that language includes many functions other than the referential function or
propositional meaning, which appears to have been the exclusive concern of logic,
structural linguistics, and generative syntax.
That is, as Nerlich and Clarke (1996) have shown, the social, psychological,
linguistic, and philosophical studies of the non-referential functions (multi-functionality) of language and communication, were vigorously pursued especially
from the 1880s to the 1930s. This was the period of the linguistic turn (cf. Rorty
1967; Apel 1994), before the rise of formalist structural linguistics, which concentrated on the formal and semantic, but not pragmatic or indexical aspects of language, in the 1940s. To be sure, the functionalist traditions were kept alive in the
works of Jakobson, Benveniste, Kuryowicz, Pragueans, and (neo-)Firthians, to
name but a few important figures and schools, but the dominance of formalists
(such as neo-Bloomfieldians and Chomskyans) in structural and generative linguistics, associated with the philosophical project of logical syntax, in the United
States and beyond, continued until the 1970s (see Section 2). As a result, the earlier
functionalist theories and their achievements became largely forgotten, so that,
when Generative Semanticists re-discovered the importance of context (pragmatics) and started to rebel against the orthodox Chomskyan doctrine of autonomous syntax in the late 1960s, they made appeal to the philosophical theories of
Austin, Grice, Strawson, and Searle, but not to the older functionalist traditions (cf.
Harris 1993). Linguistic pragmatics, which arose in the 1970s in the wake of Generative Semantics, was no less strongly influenced by the philosophical theories,
but it also actively interacted with the offspring of the older functionalist traditions,
i.e., a variety of functionalist approaches to grammar, text, and discourse, which
saw a renaissance in the 1970s (cf. Bolinger 1968; Comrie [1981] 1989; Halliday
and Hasan 1976; Leech 1983: 4678). Thus, the importance of the notions advanced by the older traditions, especially deictic field (Zeigfeld), indexicality,
and metapragmatics (the latter two derived in part from Jakobsons (1957, 1960)
code-and-message model and metalingual function, respectively; cf. Silverstein 1976b), has become more fully appreciated, and a number of attempts have
been made to explore and elaborate on the implications of these important notions,
especially in the recent literature of pragmatics (see Sections 2 and 7; cf. Mey
[1993] 2001; Verschueren 1999; Bublitz and Hbler 2007).
148
5.
Wataru Koyama
The linguistic turn and the philosophy of language: functionalism,

phenomenology, and neo-Kantianism
The older traditions of functionalism, as noted above, blossomed from the 1880s to
the 1930s, the period when philosophy took the linguistic turn. Here, focusing on
the disciplinary development of philosophy, which was closely linked to the rise of
semantics and pragmatics, we note that the late nineteenth century saw the returnto-Kant movement, as can be seen in the Kant-inspired pragmatic philosophy of
Peirce in the United States and, especially, the epistemology of neo-Kantians in
Germany, whose opponents included the logico-mathematician Frege (cf. Coffa
1991). This suggests that logico-analytic philosophy, other kinds of the philosophy
of language, neo-Kantianism, and phenomenology were, directly or indirectly, all
involved in the linguistic turn, and interacted with each other, as well as with the
older traditions of functionalism (cf. Richardson 1998; Friedman 2000). Let us
briefly explore historical connections among them, taking Frege as our starting
point.
Needless to say, Frege (18481925) was one of the founders of modern predicate logic and semantics, and his distinction between sense (Sinn) and reference
(Bedeutung) anticipated the later division of semantics and pragmatics, articulated
by Morris and Bar-Hillel (see Section 2, above). No less important, Freges differentiation between proposition and force (Kraft), anticipated by Condillac among
others,6 was a precursor of the distinctions between constatives and performatives,
and between locutionary, illocutionary, and perlocutionary forces, advanced by
Austin, who was a translator of Freges Die Grundlagen der Arithmetik (The
Foundations of Arithmetic). As is well known, these distinctions underlay the formula F(p) (where p and F stand for proposition and illocutionary force, respectively), articulated by Austins student, Searle (cf. Nerlich and Clarke 1996:
22, 76, 202204). Interestingly, the later Searle (1983) advanced a phenomenologically-oriented theory of speech, as suggested by the title Intentionality (a notion
derived from Husserl (18591938), Brentano (18381917), and, ultimately, the
Scholastic concept of intentio); and Searles master, Austin, intended his philosophy of language to be a kind of linguistic phenomenology.7
Thus, there appears a link between Freges logic(ism) and Husserlian phenomenology, although Husserl, who started his career as a philosopher of mathematics
and logic, moved away from logic towards transcendental phenomenology and,
eventually, the phenomenology of the life-world (Lebenswelt). Subsequently, phenomenology inspired various kinds of the philosophy of language: e.g., not only
Austins and Searles, but also the French dialectician Merleau-Pontys (1960),
who exchanged his views with Oxonians, viz., Ryle, Austin, and Urmson, at a colloquium in 1958. Further, phenomenology also influenced the critical social theory
of Habermas (b. 1929; cf. Mey [1993] 2001: 315), who has adopted the notion of
Lebenswelt, and, earlier, the sociology of Alfred Schutz (18991959), who carried
149
it from Vienna to New York. In North America, phenomenological sociology was

then carried out by Berger and Luckmann (cf. Mey [1993] 2001: 115) and by ethnomethodologists, including Garfinkel (cf. Giddens [1976] 1993; Bourdieu 1980).
As linguistic pragmatics has been variously influenced by ethnomethodology, conversation analysis, and Critical Theory, it can be said that the phenomenological
orientation has entered linguistic pragmatics not only from the Ordinary Language
School (see immediately above), but also from phenomenological sociology.
Hence, the phenomenological strand is fairly strong in linguistic pragmatics. After
all, the discipline was established in reaction against formalist structuralism in linguistics, which was, unlike the phenomenologically-oriented functional theories of
Bhler, Jakobson, and the Prague School (cf. Holenstein 1977), non- or even antiphenomenological.
Thus, although analytic logic and phenomenology may appear polar opposites,
there is a linkage between them, as indicated by the careers of Husserl and Austin,
as well as by the history of linguistic pragmatics, which has been, since the late
1970s, moving from de-contextualized logic and maxims towards contextualized
events, i.e., social interactions in historical and cultural life-worlds.
Further, this strongly suggests a certain affinity between linguistic pragmatics
and the philosophical movement of neo-Kantianism. Here, observe that, just like
pragmatics, neo-Kantianism seems located between logic and phenomenology, as
witnessed by the fact that Peirce, who belonged to the Return-to-Kant Movement,
invented a semiotically-based predicate logic, distinct from Freges arithmetic
one, and called his pragmati(ci)sm phenomenology and phaneroscopy.8 Also,
one may note that a neo-Kantian philosopher par excellence, namely Cassirer
(18741945), who belonged to the Marburg School of neo-Kantianism and emigrated to New York, contributed an important paper (Cassirer 1945) on the status
of structural linguistics to the New York-based journal Word, run by Martinet
(190899), Jakobson, and other structural-functionalists, who were opposed to
native-born neo-Bloomfieldian formalists in the United States. This points to the
linkage between neo-Kantianism and functionalist linguistics, and the latter was,
as we have seen, intertwined with phenomenology, especially in the works of Jakobson and Bhler.9
As a matter of fact, Cassirers career, characterized by the application of
(neo)Kantian philosophy, in particular epistemology, onto empirical investigations
into myths, languages, and the history of ideas, was part of the historical transformation, or empirical turn, of Kantian philosophy, which moved from pure and
practical (praktisch in Kants terminology) reason, i.e., critical-philosophical,
transcendental principles of epistemology and praxiology, towards more empirical,
historical, anthropological, pragmatic (pragmatisch) studies of knowledge and actions, at the turn of the twentieth century. Crucially, we should see that, in general,
when knowledge and actions are studied empirically, it becomes clear that they are
mediated by language and communication. Here, recall from Section 2, that Locke
150
Wataru Koyama
(16321704), one of the founders of modern empiricism, felt compelled to write a

treatise on human understanding, in particular, language, communication, and semiotics; and Thomas Reid (171096), deeply influenced by Locke (as well as by
Hume, a Scotsman like Reid), was a precursor of modern pragmatics and, perhaps,
phenomenology (cf. Mulligan 1987: 3334). Subsequently, Peirce (18391914),
influenced by Locke, Reid, and Kant (17241804), developed a semiotic philosophy of communication, which he called pragmati(ci)sm, based on Kants term
pragmatisch (see immediately above). The term was adopted by the behavioristic
pragmatist Morris, who coined the term pragmatics in his famous trichotomy of
communication (see Section 2). We still operate with Morriss terminology, i.e.,
syntax, semantics, and pragmatics; and this fact suggests the intermingling of not
only the terminological but also the conceptual genealogies of todays linguistic
pragmatics and the neo-Kantian movement in philosophy.
Thus, the empirical turn of (neo)Kantianism seems to lie at the roots of linguistic pragmatics. Here, notice that the heyday of neo-Kantianism in philosophy
coincided with the linguistic turn, i.e., the period in which the older traditions of
functionalism blossomed, from the 1880s to the 1930s (cf. Nerlich and Clarke
1996). This cannot be by chance. By the linguistic turn, judgment, a concept integral to Kantian philosophy, became replaced with proposition in post-Fregean
logic, and, more generally, with language in twentieth-century philosophy, including logico-analytic philosophy and the Ordinary Language School;10 at the
same time, (neo)Kantianism turned empirical, giving rise to pragmati(ci)sm and
the empirical science of language, i.e., linguistic functionalism, studying the synchronic, living state of language in action, context, and communication. This
movement, replacing the nineteenth-century historical comparative linguistics
(over)systematized by Neo-Grammarians, was what was called (functionalist)
structuralism (cf. Jakobson 1973), which would become overshadowed by formalist structuralism, the latter turning away from context and pragmatic functions to
formal structure and semantics in the 1940s and 50s.
We shall soon come back to the synchronic turn in the studies of language,
corresponding to the linguistic turn in philosophy: i.e., the rise of structuralism as
part of linguistic functionalism, between the 1880s and the 1930s. But, first, we
must observe that the empirical turn of (neo)Kantian philosophy meant the emergence of the empirical sciences of not just language, but also the mind, society, and
culture: i.e., modern twentieth-century psychology, sociology, and anthropology.
Note that William James (18421910), Wundt (18321920), Max Weber (1864
1920), Simmel (18581918), Durkheim, Mauss, Boas, Malinowski, and RadcliffBrown, who are usually counted as founders of these disciplines, worked in this
period, from the 1880s to the 1930s, and some of them were explicitly associated
with pragmatism and philosophical neo-Kantianism, while others directly or indirectly communicated with them. Further, whether or not they were pragmatists
or neo-Kantians in the narrow sense, their empirical sciences and theories gen-
151
erally shared the epistemological framework of neo-Kantianism, such as the methodologically-based distinction between nomothetic sciences, investigating laws
and other regularities, and idiographic (or historical) sciences, exploring unique
and singular, sociohistorically contextualized phenomena (cf. Stocking 1996). Together with the related opposition between collectives and individuals, this distinction underlay, among others, the Saussurean notions of langue and parole, the
proto-Weberian Methodenstreit concerning the status of economics as a socio-historical (idiographic) or ahistorical (nomothetic, classical Newtonian-Laplacean)
science, and the Durkheimian understanding of sociology in opposition to individualist psychology.
Thus, (neo-)Kantianism was the philosophical foundation for the rise of the
human sciences (cf. Foucault 1966), including linguistics and, later, pragmatics.
The linguistic turn in philosophy and the (functionalist) structuralist turn in linguistics were part of the (broadly conceived) neo-Kantian movement, which also
included the rise of phenomenology. That is, in the studies of language, the linguistic turn at the beginning of the twentieth century consisted of the move towards
the synchronic, living state of language in action (i.e., pragmatics); and the rise
of structuralism should be understood as part of the functionalist, phenomenological movement, oriented towards actions and other happenings taking place hic et
nunc, or the dectic center (origo) of such pragmatic occurrences (see Section 2). As
Jakobson was never tired of pointing out (cf. Jakobson 1973; Holenstein 1977),
structuralism, i.e., functionalist (vs. formalist) structuralism, was not opposed to
phenomenology; they constituted a unified whole (holistic Gestalt), as it were. The
philosophy underlying structuralism was the neo-Kantian one, represented by the
idea of the primacy of (human) (ap)perception, advanced against nineteenthcentury positivism and mechanical empiricism, which had culminated in the NeoGrammarians comparative (historical) linguistics in the late nineteenth century,
and which had earlier served to suppress the proto-linguistic turn, achieved by
Bernhardi and Humboldt in Germany, Condillac and the Idologues in France, and
Reid and the followers of his commonsense philosophy such as Dugald Stewart
(17531828) at the turn of the eighteenth to the nineteenth century (see notes 1 and
2, above; cf. Nerlich and Clarke 1996). The primacy of human perception, or subjective apperception, clearly underlies the structuralist notions of phoneme and
sound pattern, advanced in contradistinction to mechanical physical sounds,
studied by the empiricist, positivist science of nineteenth-century phonetics.
Further, phonemes were considered part of not just subjective (individual), but
also intersubjective, sociocultural reality (see Section 6). The insiders of a community sharing (by degrees) a linguistic structure, i.e., members of a linguistic
community, (ap)perceive the (more or less) same phonemes, but not phones;
they have phonemes as part of their psychological reality (cf. Sapir 1933), that
is, intersubjective phenomenological reality, a pattern of (ap)perception not shared
by the outsiders of the (Durkheimian) socio-cultural, linguistic-structural commu-
152
Wataru Koyama
nity. Thus understood, it becomes less mysterious why the phenomenologist Merleau-Ponty (1945, 1960) appreciated both Gestalt psychology and Saussures linguistics, celebrating the latters discovery of synchrony and parole, and why
Saussure was closely associated with Bral (18321915), a founder (precursor) of
modern twentieth-century semantics and pragmatics, in Paris (cf. Aarsleff 1982;
Nerlich 1992). For all of them, including even Saussure, language was, au fond, an
intersubjective, phenomenological reality, pragmatically unfolding in front of our
eyes, that is, emerging at the synchronic hic et nunc of communication. Thus, the
historical and theoretical roots of modern structural linguistics and linguistic pragmatics converge in the rise of structuralism, i.e., functional structuralism, in the
period from the 1880s to the 1930s, that is, the days of the older traditions of linguistic functionalism, to be eclipsed when structural linguistics became dominated
by neo-Bloomfieldian, Chomskyan, and other kinds of dogmatic formalism after
the achronic (vs. synchronic) turn in the 1940s.
6.
Intersubjectivity and interaction: functionalism, Ordinary Language

philosophy, and the social turn in pragmatics
Intersubjectivity involves not only (1) the speaker (addresser) and (2) the hearer
(addressee), but also (3) the channel (contact) between them, (4) the sign (message) that is sent from the former to the latter, (5) the (intersubjectively) shared
code(s) for interpretations of the sign, and (6) the (intersubjectively) understood
referent(s) of the sign, located in the micro- or macro-context of the interaction between the speech- or communicative-event participants, such as the speaker and
addressee. This is, of course, the contextual, pragmatic matrix of communication,
including language, which was articulated in the Praguean functionalist theory of
Jakobson (1960), mentioned in Section 4. As part of the code, linguistic structure is
embedded in this functional matrix of communication. In other words, structure
emerges through communicative practices; there is no structure (langue) without
pragmatics (parole). According to this Jakobsonian theory, as noted earlier, there
are six basic functions of communication, corresponding to the six aspects of the
contextual matrix: viz., (1) emotive, (2) conative, (3) phatic, (4) poetic, (5) metalingual (metalinguistic), and (6) referential functions.11 Although linguistic structure, as part of the code, is primarily used to interpret the sign (token) to identify
(i.e., intersubjectively construct) its referent(s) and thus primarily involved in the
metalinguistic, poetic, and referential functions (respectively corresponding to the
code, sign, and referents), it inevitably interacts with the other functions as well,
because it exists only as part of the code pragmatically used for intersubjective interpretations of the sign indexically exchanged in the context of communication.
Thus, language, including linguistic structure, i.e., (morpho)phonology, (morpho)syntax, and semantics, should be understood, theorized, and analyzed by tak-
153
ing into proper consideration the six functions, as well as the six aspects, of communicative practices (cf. Silverstein 1976a, 1976b, 1993).
But, of course, the contextual, functional matrix of communication, articulated
in the six-functions model, is not limited to, or centered around, linguistic structure.
For example, the code includes not just linguistic structure, but also maxims, interpretive rules of thumb, passing theories, metapragmatic frames, and models
of interpretation vis--vis (not necessarily linguistic) signs, including gestures, facial expressions, and other paralinguistic phenomena (cf. Bolinger 1968), all of
which occur in specific, sociohistorical contexts of communicative practices (see
Section 7; cf. Mey [1993] 2001: 206235). And such contextualized practices
index12 not just referents (in the referential function) and the semantic and pragmatic codes being used (in the metalinguistic function). They also index the social identities of, and power-relations among, communicative-event participants
(e.g., in the emotive and conative functions). That is, contextualized practices
index not just referents, but also social identities and power-relations, which can be
read in the emergent texts of sociocultural practices of communication. Such
texts are not to be understood as static objects, but as the dynamically emergent,
socioculturally interpretable organizations of referential and non-referential (interactional) practices (mainly corresponding to the poetic, phatic, and metapragmatic functions), that is, organizations studied in conversation analysis (cf.
adjacency pairs; Levinson 1983: 284370), discourse analysis (cf. frames of interaction; Goffman 1974), linguistic anthropology (cf. Silverstein and Urban 1996;
Blommaert 2005), and, of course, pragmatics (cf. Mey [1993] 2001).
Thus, in a sense, the Praguean/Jakobsonian functional theory provided a general framework in which todays empirical approaches to language, listed immediately above, operate (also see Section 2). As this indicates, the rise of structuralism
in the late nineteenth and early twentieth centuries served to prepare the proper theorization of the functional, contextual matrix of language, which would emerge in
the 1970s after the formalist lapse from the 1940s to 60s. This lapse, however,
made obscure the pragmatic, functionalist foundation of structural linguistics. Indeed, by the late 1960s, when Generative Semanticists, who were immediate predecessors of linguistic pragmaticians, started to rebel against formalist structuralism (Chomskyan generativism) and delve into the realm of pragmatics, North
American linguistics had become more closely linked to the philosophy of language, especially logico-analytic philosophy, than to the early twentieth-century
functionalist traditions of structuralism, so that, in trying to find theoretical tools to
deal with pragmatic phenomena, Generative Semanticists discovered and adopted
the theories of speech act, implicature, and presupposition, articulated by Ordinary
Language philosophers like Austin, Grice, and Strawson (see Section 3, above; cf.
Levinson 1983). The publication of Austins (1962; second edition, 1975) William
James Lectures, originally delivered in 1955 at Harvard, which is located in the
same city as MIT (Cambridge, Mass.), i.e., the home of generativism, made the
154
Wataru Koyama
theories of the Ordinary Language philosophers more visible to generativists interested in pragmatic matters (also cf. Grice 1967; Chapman 2005: 100). And Searle
(1969) provided a desideratum, i.e., the desired systematization of speech act theory, making it amenable to the linguists formal, syntactic treatment of pragmatic
phenomena.13 As Generative Semantics yielded to linguistic pragmatics in the
1970s, the early years of the new discipline were marked by the preoccupation with
the issues articulated by the Ordinary Language philosophers, and many of the culture-specific assumptions of their philosophy, i.e., post-Lockean modern British
philosophy, concerning human communication, seem to have been directly borrowed into linguistic pragmatics.
Here, recall that Locke, a founder of modern British empiricism, discussed,
like Saussure, the arbitrary nature of human language (communication), but, unlike the latter, Locke meant by the term the volitional, intentional, willful character of the individuals use of language, whereas Saussure meant the conventional
(non-volitional), extensionally unmotivated (intensional) yet socially existing (i.e.,
Durkheimian) association of the signifiant and the signifi. As this indicates, British and Anglo-American philosophy, especially the empiricist tradition thereof
(see Section 2), has tended to theorize communication (more generally, human
practice) in terms of volitional, intentional actions or, in sociological terms of Max
Weber, the purposive function (cf. Silverstein 1976b; Apel 1994: 63). Ordinary
Language philosophy, an empiricist philosophy of language and communication
concerned with what we do with words and what is meant, is no exception. It
sees language and communication as purposive activities, as explicitly articulated
by G. E. M. Anscombe, G. H. von Wright, and Grice, who tried to explain the social, interactional fact of communication in terms of what we intend to do or what
we intend to mean, that is, individuals conscious intention (cf. Chapman 2005).
Notwithstanding, as Weber and other social scientists have shown, actions have
consequences (perlocutionary effects) which are unintended by agents (subjects),
and which may be more important, in terms of historical consequences (i.e., pragmatic effects), than actors mere intentions (cf. Silverstein 1976b). Further, any
practice at least partially consists of unintentional, non-conscious elements; obviously, we are not conscious of everything we do. For example, we, as ordinary
language users, are usually more aware of surface segments (continuous, segmentable units), such as phonemes, words, and expressions, than of allophones and
discontinuous syntactic units (cf. Lucy 1992); yet, the latter two are often part of
dialectal/sociolectal varieties and thus index their users social identities, pertaining to class, gender, status, etc., and power-relations among them, as can be readily
seen in the sociolinguistic literature, Labovian or otherwise. Thus, there are nonpurposive pragmatic functions in discourse, which are too important to be left out
in any theory of pragmatics.
In addition, cross-cultural comparisons suggest that many societies place much
less value on speakers intentions in the evaluation of (sociocultural) meanings
155
of their linguistic practices than the culture of Ordinary Language philosophers

does, a point poignantly raised and substantiated by linguistic anthropologists,
who have been critical of the intentionalist and individualist bias of the modern
Standard Average European speech act theories ever since the 1970s and early
80s.14 Subsequently, partly as a result of this cross-cultural (anthropological)
critique, linguistic pragmatics has come to appreciate and focus on cross-cultural
differences in speech acts and, more generally, linguistic practices and their
metapragmatic understandings, the latter now discussed under the rubric of language ideology.15
No less important, the individualist assumption of the speech act theories was
undermined by empirical studies of discourse, which showed that the meaning(s)
of an action (speech act) cannot be determined out of context, i.e., without considering a series of interactions which contains the action (cf. Hancher 1979; Levinson 1983). This means that, to the extent that the meaning(s) of an action is
determinable, it is determined by the contextualized interaction of speech-event
participants, not solely by the individual speakers intention. Contributions of conversation analysis (cf. Levinson 1983: 284370) and discourse analysis (cf. Mey
[1993] 2001: 191), focusing their attention on interaction rather than on solitary individual actors minds, have been important in this interactional turn (social
turn) of pragmatics, which has characterized the discipline since the 1980s or so
(cf. Thomas 1996). Pragmatics today is much more interaction- or discourse-centric than in the 1970s, when the new discipline was more intention- and individualcentric under the influence of Ordinary Language philosophy and modern British
empiricism.
7.
Genealogy of performative utterances and other multi-functional

co-incident happenings in the speech event
Let us finally move to the question of performativity, another central theme in linguistic pragmatics. Clearly, this is a complex notion, only a part of which is accounted for by the creative function (efficacy), discussed in Section 3. That is, in
addition to the creative function, performativity involves self-referentiality (cf.
Recanati 1979),16 which is itself a composite notion, made up of reflexitivity and
(transparent) referentiality. The latter means the (relatively) transparent relationship between the referential function (what is said) and the non-referential, interactional function (what is done), as observed when we use a referentially explicit
metapragmatic utterance, such as When I said that, I meant it to be a compliment
(cf. Caffi 1984; Mey [1993] 2001: 173205). On the other hand, reflexivity as
such is just the mutual indexing of some semiotic elements, linguistic or non-linguistic, co-occurring in the event of communication. For example, a pointing gesture and a token of the deictic expression this book, occurring together, mutually
156
Wataru Koyama
index each other and function reflexively, as the gesture and the linguistic utterance
coincide in the same context (deictic, indexical field; see Section 2). Mutual, reflexive indexing also obtains when a defiant smile, or a serious demeanor, occurs
with an utterance Im sorry, or when a judge, properly dressed or completely undressed, utters I declare you guilty. Clearly, these are examples of what Gregory
Bateson ([1955] 1972) called metacommunication (cf. Ruesch and Bateson
1951; Caffi 1984) and what Gumperz (1982) called contextualization cue (cf.
Auer and Luzio 1992; Duranti and Goodwin 1992), notions coming out of the anthropological, psychiatric, and ethological traditions, independently of Austins
discussion of performatives.
Austins (1962) notion of performativity, following the discussions by the phenomenologist Adolf Reinach (see note 7) and others in the study of law, did not
clearly separate these two dimensions of performativity or explore each dimension in depth. Originally, linguistic pragmatics was under the influence of Austins
way of formulating the problem, but it has gradually moved to see that the problem
of performativity should be dealt with by separating out different principles underlying the problem, such as the principle of metapragmatics, which is becoming a
major topic in our discipline (cf. Ducrot 1984; Verschueren 1999; Mey [1993]
2001). Like the cognate notion metacommunication, the concept of metapragmatics came out of the anthropological literature, in particular Silverstein
(1976b), who coined the term to designate the pragmatic (vs. semantic) aspect of
the metalingual function,17 articulated by his teacher, Jakobson (1957, 1960),
who, in turn, derived it from the discussion of metalanguage (vs. object language) by the mathematician David Hilbert (18621943), Kurt Gdel (190678),
Tarski and other logicians, e.g., Russell, the early Wittgenstein, and Carnap (cf.
Apel 1994). The metapragmatic function obtains when a sign, be it an utterance, a
gesture, or whatever, indexes another sign (speech, non-linguistic action, etc.) in
context. When the two signs co-occur (coincide) in the same context, they are said
to show reflexively calibrated metapragmatics (cf. Lucy 1993; Mey [1993]
2001). According to Silverstein (1993), the reflexive metapragmatic function
serves to create texts, be they referential or non-referential (see Section 6 on
text). For instance, the reflexive co-indexing of referential signs, such as (thematic) anaphor and its (rhematic) antecedent, in the same unit of discourse creates
referential textuality, as studied in Praguean functional structuralism, text linguistics, and discourse analysis (cf. Halliday and Hasan 1976). Similarly, the reflexive co-indexing of non-referential signs, such as the two units of an adjacency
pair (e.g., request-compliance or -rejection; cf. Levinson 1983: 284370), in the
same phase of discourse creates interactional, non-referential textuality.
Clearly, the idea of textualization came from, among other places, Bhlers
notion of Sprachwerk language work and, in particular, Jakobsons poetic function (see Section 6), i.e., the communicative function that is focused on messages
(i.e., signs) occurring in the context of interaction and thus gives textual organiz-
157
ation to them, as can be most clearly seen in parallelisms and repetitions abundantly observed in poetry and other rhetorically elaborate verbal arts (e.g., folklore
and myths; cf. Bauman 2004). Importantly, basic units of poetic text are often nonreferential segments, such as meters and rhymes, in contradistinction to referential
units such as rhemes, themes, and syntactic segments like phrases, clauses, sentences. Thus, poetic texts are, at least partially, non-referential texts, i.e., texts
of non-referential interaction. And hence, poetic texts tend to become endowed
with non-referential meanings, such as emotional (emotive) and persuasive
(conative) forces (functions). The same characteristics are found in rituals,
which also typically show the abundance of parallelisms and repetitions in the linguistic and non-linguistic, interactional modes (cf. Caton 1990). To be sure, in ordinary conversations too, we also find such interactional poetic parallelisms and
repetitions especially when they contain what Goffman (1967) called interaction
rituals (cf. greetings and other adjacency pairs, mentioned above). Yet, the characteristic is most salient in rituals, in which, we should recall, Austin (1962) discovered (explicit) performatives: e.g., christening rituals, legal rituals, etc. Performative utterances are the part of rituals characterized by the use of little ritual
one-liners, i.e., performative sentences, which are, thus, socially recognized formulaic types, recursively co-occurring with the non-linguistic elements of ritual
settings (cf. Volosinov [1929] 1973; Silverstein 1993).
Thus, performatives show reflexivity, i.e., the reflexive metapragmatic
function, which is observed more broadly in rituals, poetry, metacommunication,
contextualization cues, etc. On the other hand, performatives involve another
principle, that is, referentiality, or relatively pronounced degrees of transparency between the referential function (what is said) and the non-referential, interactional function (what is done), as noted above. Clearly, it is this principle that
has been the principal focus of pragmatic discussions on performativity and
speech act, most clearly observable in Searles theory of indirect speech act and
Grices theory of implicature. These two notions were formulated to account
for non-explicit or implicit relationships between the referential and non-referential functions of utterances, and they have been main themes of linguistic pragmatics since the 1970s. Much of linguistic pragmatics still operates with these
concepts. For example, Brown and Levinsons ([1978] 1987) politeness theory
and, inter alia, Sperber and Wilsons Relevance Theory seem squarely located in
this tradition.
Clearly, the theme of referential explicitness vs. implicitness goes back to the
old tradition of rhetoric, where literal expressions and tropes were understood in
such terms (cf. Ricur [1975] 1977). When Austin discovered non-referential
functions in explicit performatives, he did not depart from this age-old tradition, as
he focused on non-referential functions only insofar as they appear together with
the referential function, as in performatives, where the two kinds of function
happen to coincide. Just as in ancient rhetoric, Austin and his followers moved
158
Wataru Koyama
from the referential function, as studied in logic and formalist structuralism, to

non-referential functions. That is, Austins was a reference-centric approach to
pragmatic functions, the severe limit of which is indicated by the comparison with
multi-functionalist approaches to pragmatics found in, e.g., Jakobsons theory (see
Section 6), its developments in linguistic anthropology (see Section 2 and immediately above), and the theory of pragmatic act, advanced in social pragmatics (cf.
Mey [1993] 2001: 206235; also see Section 3, above). The rise of social pragmatics and linguistic anthropology in linguistic pragmatics since the 1990s seems
to suggest that the discipline is not only recovering but also further elaborating the
multi-functionalist, social interactionist insights of the older traditions of linguistic
functionalism, which blossomed from the 1880s to the 1930s (see Section 2).
Notes
1. More specifically, Reids philosophy influenced Benjamin Smart (17861872), Adolphe
Garnier (180164), and Henry Sidgwick (18381900), and developed into Charles
Peirces (18391914) semiotic pragmati(ci)sm, Charles Morriss (190179) behaviorist
pragmatism, Sidgwicks student George Stouts (18601944) empiricist and behavioristic philosophy, the latters student G. E. Moores (18731958) commonsense empiricism, and Grices (191388) theory of non-natural meaning and implicature in the
nineteenth and twentieth centuries (cf. Nerlich and Clarke 1996: 2021, 105106; Chapman 2005: 22, 25, 7172).
2. The German tradition ran from Bernhardi (17691820) and Wilhelm von Humboldt
(17671835) to Philipp Wegener (18481916), the later Karl Brugmann (18491919),
Anton Marty (18471914), and Bhler, whereas the French one ran from the Idologues
to Bral, (18321915), Meillet (18661936), Mauss (18721950), Bally (18651947),
Guillaume (18831960), and Benveniste, to mention only a few of the major figures.
3. This was preceded by the medieval trivium (artes sermocinales), i.e., grammar, logic,
and rhetoric. Note that Morriss predecessor Peirce was thoroughly familiar with Scholastic philosophy.
4. The Journal of Pragmatics was launched in 1977 (cf. Haberland and Mey 1977). Subsequently, major textbooks of pragmatics became published in the early 1980s (cf. Leech
1983; Levinson 1983), the International Pragmatics Association (IPrA) was founded in
1986, and Pragmatics was started in 1991, while the second wave of major, and more
Continental-European (vs. Anglo-American), textbooks appeared in the 1990s (cf. Mey
[1993] 2001; Verschueren 1999).
5. These ideas had been suggested and, in part, advanced by Philipp Wegener (18481916).
6. For instance, before Condillac (171480), Grammaire gnrale et raisonne had made
the distinction between dictum and modus, roughly corresponding to proposition and
force, respectively.
7. Also, note that Adolf Reinach (18831917), whose theory of social acts was one of the
precursors of Austins theory of speech acts and performativity, was a phenomenologist. Reinach was a student of Husserls and, with Johannes Daubert (18771947), belonged to the Munich School of phenomenology. The Scottish commonsense philos-
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
159
ophy of Thomas Reid (see note 1), which underlay Ordinary Language philosophy and
much of twentieth-century British empiricism, also exercised influence over the early
phenomenologists, such as Brentano (18381917), one of the predecessors of Reinach.
(Reinach was under the influence of the (individualistic) functionalist Anton Marty
(18471914), a student of Brentanos.) Further, it should be underlined that Reid, Reinach, and Austin all used the philosophy of law as a common source of reference for their
philosophies of language (see Section 7, below; cf. Mulligan 1987: 3334; Nerlich and
Clarke 1996: 12, 104, 211215).
Peirce and Husserl were both influenced by Bolzano (17811848), who had been, in
turn, influenced by Johann Lambert (172877) and Locke; the latter two also directly influenced Peirce (cf. Nerlich and Clarke 1996: 21).
Bhler was also closely associated with the Vienna Circle, including Carnap. As this
suggests, and as Friedman (2000) has shown, logical positivism (of the Vienna Circle)
and phenomenology shared common themes (e.g., intersubjectivity, construction of the
world, etc.) and common roots (e.g., neo-Kantianism). For instance, the early Carnap
made appeal to Gestalt psychology and epistemological holism, which were intimately
associated with structuralism, phenomenology, and especially, the neo-Kantian notion
of constitution.
Generally speaking, the linguistic turn in philosophy was made up of (1) the Russellian critique of ordinary language by means of logico-mathematical analysis, (2) the
Carnapian critique of metaphysical dogmas (later, epistemology), such as realism and
phenomenalist positivism, by means of logic, and (3) the Austinian critique of philosophy, including logic, by means of the commonsense analysis of ordinary language.
Austins philosophy of language was anticipated by Scottish commonsense philosophy,
especially Reid, who influenced Peirce, Moore, and Grice, among others (see note 1,
above).
See Section 7, below, for the metalinguistic (more specifically, metapragmatic) and
poetic functions.
See Section 2 for indexicality.
Recall Rosss and other Generative Semanticists performative hypothesis (cf. Sadock 1974; Levinson 1983: 247263, 271273).
See Section 3, above; cf. Silverstein 1976b; Rosaldo 1982; Du Bois 1987; Duranti 1993.
Cf. Schieffelin, Woolard, and Kroskrity [1992] 1998; Verschueren 1999; Mey [1993]
2001: 173205, 262288; Jaworski, Coupland, and Dariusz 2004.
As Nerlich and Clarke (1996) have pointed out, the property of self-reference (soirfrence) had been discovered, in 1929, by Erwin Koschmieder (18951977), who
started to discuss Koinzidenzfall the case of coincidence in the 1940s.
See Section 6, above.
References
Aarsleff, Hans
1982
From Locke to Saussure: Essays on the Study of Language and Intellectual
History. Minneapolis, MN: University of Minnesota Press.
Agha, Asif
2007
Language and Social Relations. Cambridge: Cambridge University Press.
160
Wataru Koyama
Apel, Karl-Otto
1994
Karl-Otto Apel: Selected Essays, Volume 1: Towards a Transcendental Semiotics. Edited by Eduardo Mendieta. Atlantic Heights, NJ: Humanities Press.
1992
Austin, John L.
1962
How to Do Things with Words: The William James Lectures Delivered at Harvard University in 1955. Edited by James O. Urmson. Oxford: Clarendon
Press.
Bar-Hillel, Yeoshua
1946
Analysis of correct language. Mind 55: 328340.
Bar-Hillel, Yeoshua
1954
Bateson, Gregory
[1955] 1972 A theory of play and fantasy. In: Gregory Bateson, Steps to an Ecology of
Mind, 177193. New York: Ballantine Books.
Bauman, Richard
2004
A World of Others Words: Cross-Cultural Perspectives on Intertextuality. Oxford/Malden, MA: Blackwell.
Bauman, Richard and Joel Sherzer (eds.)
[1974] 1989 Explorations in the Ethnography of Speaking, 2nd edition. Cambridge:
Benveniste, Emile
1966
Problmes de linguistique gnrale, I. Paris: Gallimard.
Blommaert, Jan
2005
Discourse: A Critical Introduction. Cambridge: Cambridge University Press.
Bolinger, Dwight
1968
Aspects of Language. New York: Harcourt Brace Jovanovich.
Bourdieu, Pierre
1980
Le sens pratique. Paris: Minuit.
[1978] 1987 Politeness: Some Universals in Language Usage. Cambridge: Cambridge
University Press.
2007
Metapragmatics in Use. (Pragmatics and Beyond, n. s. 165.) Amsterdam: Benjamins.
Bhler, Karl
[1934] 1990 Theory of Language: The Representational Function of Language. Amsterdam: Benjamins. [Sprachtheorie: Die Darstellungsfunktion der Sprache.
Jena: Fischer.]
Butler, Judith
1990
Gender Trouble: Feminism and the Subversion of Identity. London: Routledge.
Caffi, Claudia (ed.)
1984
Special Issue on Metapragmatics. Journal of Pragmatics 8(4).
Cameron, Deborah and Don Kulick
2003
Language and Sexuality. Cambridge: Cambridge University Press.
161
Carnap, Rudolf
[1934] 1937 The Logical Syntax of Language. London: Kegan Paul. [Logische Syntax
der Sprache. Vienna: Springer.]
Cassirer, Ernst A.
1945
Structuralism in modern linguistics. Word 1: 99120.
Caton, Steven C.
1987
Contributions of Roman Jakobson. Annual Review of Anthropology 16: 223
260.
Caton, Steven C.
1990
Peaks of Yemen I Summon: Poetry as Cultural Practice in a North Yemeni
Tribe. Berkeley/Los Angeles/London: University of California Press.
Chapman, Siobhan
2005
Paul Grice, Philosopher and Linguist. Basingstoke/New York: Palgrave Macmillan.
Coffa, J. Alberto
1991
The Semantic Tradition from Kant to Carnap: To the Vienna Station. Cambridge: Cambridge University Press.
Comrie, Bernard
[1981] 1989 Language Universals and Linguistic Typology: Syntax and Morphology,
2nd edition. Chicago: University of Chicago Press.
Coupland, Nikolas, Srikant Sarangi and Christopher N. Candlin (eds.)
2001
Sociolinguistics and Social Theory. Harlow: Pearson Education.
Davis, Steven (ed.)
1991
Pragmatics: A Reader. Oxford/New York: Oxford University Press.
Donnellan, Keith S.
1966
Donnellan, Keith S.
1978
Speaker reference, descriptions and anaphora. In: Peter Cole (ed.), Syntax and
Semantics, Volume 9: Pragmatics, 4768. New York: Academic Press.
Du Bois, John W.
1987
Meaning without intention: Lessons from divination. IPrA Papers in Pragmatics 1: 80122.
Ducrot, Oswald
1984
Langage, mtalangage et performatifs. In: Le dire et le dit, 117148. Paris:
Minuit.
Duranti, Alessandro
1993
Intentions, self, and responsibility. In: Jane H. Hill and Judith T. Irvine (eds.),
Responsibility and Evidence in Oral Discourse, 2447. Cambridge: Cambridge University Press.
1992
Rethinking Context: Language as an Interactive Phenomenon. Cambridge:
1975
Santa Cruz Lectures on Deixis, 1971. Bloomington, IN: Indiana University
Linguistics Club.
Foucault, Michel
1966
Les mots et les choses: une archologie des sciences humaines. Paris: Gaillmard.
162
Wataru Koyama
Friedman, Michael
2000
A Parting of the Ways: Carnap, Cassirer, and Heidegger. Chicago/La Salle, IL:
Open Court.
Gazdar, Gerald
1979
Giddens, Anthony
[1976] 1993 New Rules of Sociological Method: A Positive Critique of Interpretative
Sociologies, 2nd edition. Stanford: Stanford University Press.
Goffman, Erving
1967
Interaction Ritual: Essays on Face-to-Face Behavior. New York: Anchor
Books.
Goffman, Erving
1974
Frame Analysis: An Essay on the Organization of Experience. New York:
Harper and Row.
Goffman, Erving
1981
Forms of Talk. Philadelphia: University of Pennsylvania Press.
Goffman, Erving
1983
Felicitys condition. American Journal of Sociology 1: 153.
Grice, H. Paul
1967
Logic and conversation. Unpublished manuscript of the William James Lectures, Harvard University. Published in Studied in the Way of Words. Cambridge, MA: Harvard University Press [1989].
Gumperz, John J.
1982
[1972] 1986 Directions in Sociolinguistics: The Ethnography of Communication. Oxford: Basil Blackwell.
Haberland, Hartmut and Jacob L. Mey
1977
Editorial: Pragmatics and linguistics. Journal of Pragmatics 1(1): 116.
1976
Hancher, Michael
1979
The classification of cooperative illocutionary acts. Language in Society 8:
114.
Hanks, William F.
1990
Referential Practice: Language and Lived Space among the Maya. Chicago:
Harris, Randy Allen
1993
The Linguistic Wars. Oxford: Oxford University Press.
Heritage, John
1984
Holenstein, Elmar
1977
Jakobsons contribution to phenomenology. In: Daniel Armstrong and Cornelis H. van Schooneveld (eds.), Roman Jakobson: Echoes of his Scholarship,
145162. Lisse: The Peter de Ridder Press.
163
Huang, Yan
2007
Pragmatics. Oxford/New York: Oxford University Press.
Irvine, Judith T.
1996
Shadow conversations: The indeterminacy of participant roles. In: Michael
Silverstein and Greg Urban (eds.), Natural Histories of Discourse, 131159.
Chicago: University of Chicago Press.
Jakobson, Roman
1957
Shifters, verbal categories, and the Russian verb. Cambridge, MA: Harvard
University Russian Language Project.
Jakobson, Roman
1960
Jakobson, Roman
1973
Main Trends in the Science of Language. London: George Allen and Unwin.
Jaworski, Adam, Nikolas Coupland and Galasinski Dariusz (eds.)
2004
Metalanguage: Social and Ideological Perspectives. Berlin: Mouton de
Gruyter.
Jespersen, Otto
1922
Kaplan, David
1978
Dthat. In: Peter Cole (ed.), Syntax and Semantics, Volume 9: Pragmatics,
221243. New York: Academic Press.
Karttunen, Lauri
1974
Presupposition and linguistic context. Theoretical Linguistics 1: 181194.
Koyama, Wataru
2000
How to be a singular scientist of words, and worlds, and other (possibly) wonderful things. Journal of Pragmatics 32: 651686.
Kripke, Saul
1972
Naming and Necessity. Cambridge, MA: Harvard University Press.
Kripke, Saul
1979
Speakers reference and semantic reference. In: Peter A. French, Theodore E.
Uehling Jr. and Howard K. Wettstein (eds.), Contemporary Perspectives in the
Philosophy of Language, 627. Minneapolis, MN: University of Minnesota
Press.
Kuklick, Henrika
1991
The Savage Within: The Social History of British Anthropology, 18851945.
Lakoff, Robin T.
1973
Language and womans place. Language in Society 2(1): 4580.
Lakoff, Robin T.
1989
The way we were. Journal of Pragmatics 13: 939988.
Leech, Geoffrey N.
1983
1983
Locke, John
[1689] 1975 Essay on Human Understanding. Edited by Peter H. Nidditch. Oxford:
164
Wataru Koyama
Lucy, John A.
1992
Language Diversity and Thought: A Reformulation of the Linguistic Relativity
Hypothesis. Cambridge: Cambridge University Press.
Lucy, John A. (ed.)
1993
Reflexive Language: Reported Speech and Metapragmatics. Cambridge University Press: Cambridge.
Lyons, John
1977
Semantics (2 volumes). Cambridge: Cambridge University Press.
1945
Phnomnologie de la perception. Paris: Gallimard.
1960
Signes. Paris: Gallimard.
Mey, Jacob L.
1979
Zur kritischen Sprachtheorie. In: Jacob L. Mey (ed.), Pragmalinguistics:
Theory and Practice, 411434. The Hague: Mouton.
Mey, Jacob L.
1985
Whose Language?: A Study in Linguistic Pragmatics. Amsterdam: Benjamins.
Mey, Jacob L.
[1993] 2001 Pragmatics: An Introduction, 2nd edition. Oxford/Malden, MA: Blackwell.
Morris, Charles W.
1938
Foundations of the theory of signs. In: Otto Neurath, Rudolf Carnap, and
Charles W. Morris (eds.), International Encyclopedia of Unified Science,
Volume 1, No. 2, 77137. Chicago: University of Chicago Press.
Mulligan, Kevin (ed.)
1987
Speech Act and Sachverhalt: Reinach and the Foundations of Realist Phenomenology. Dordrecht: Martinus Nijhoff.
Nerlich, Brigitte
1992
Semantic Theories in Europe: From Etymology to Contextuality. Amsterdam:
Benjamins.
1996
Peirce, Charles S.
[18931910] 1955 Logic as semiotic: The theory of signs. In: Justus Buchler (ed.),
Philosophical Writings of Peirce, 98119. New York: Dover.
Pike, Kenneth Lee
1967
Language in Relation to a Unified Theory of the Structure of Human Behavior,
2nd, rev. edition. The Hague: Mouton.
Pressman, Jon F.
1994
Pragmatics in the late twentieth century: Countering recent historiographic
neglect. Pragmatics 4(4): 461489.
Putnam, Hilary
1975
Philosophical Papers, Volume 2: Mind, Language, and Reality. Cambridge:
Recanati, Franois
1979
La transparence et lnonciation: pour introduire la pragmatique. Paris:
Seuil.
165
Richardson, Alan
1998
Carnaps Construction of the World: The Aufbau and the Emergence of Logical Empiricism. Cambridge: Cambridge University Press.
Ricur, Paul
[1975] 1977 The Rule of Metaphor: Multi-Disciplinary Studies of the Creation of
Meaning in Language. Toronto: University of Toronto Press.
Rorty, Richard M. (ed.)
1967
The Linguistic Turn: Recent Essays in Philosophical Method. Chicago: University of Chicago Press.
1982
Ruesch, Jrgen and Gregory Bateson
1951
Communication: The Social Matrix of Psychiatry. New York: Norton.
Sadock, Jerrold M.
1974
Toward a Linguistic Theory of Speech Acts. New York: Academic Press.
Sapir, Edward
1933
La ralit psychologique des phonmes. Journal de Psychologie Normale et
Pathologique 30: 247265.
Schieffelin, Bambi B., Kathryn Woolard and Paul V. Kroskrity (eds.)
1998
Language Ideologies: Practice and Theory. New York/Oxford: Oxford University Press. [Originally, 1992, Special Issue on Language Ideologies,
Pragmatics 2(3).]
Searle, John R.
1969
Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge
University Press.
Searle, John R.
1983
Intentionality: An Essay in the Philosophy of Mind. Cambridge: Cambridge
University Press.
1976a
Hierarchy of features and ergativity. In: Robert M. W. Dixon (ed.), Grammatical Categories in Australian Languages, 112171. Canberra: Australian Institute of Aboriginal Studies.
1976b
Shifters, linguistic categories, and cultural description. In: Keith H. Basso and
Henry A. Selby (eds.), Meaning in Anthropology, 1155. Albuquerque, NM:
University of New Mexico Press.
1993
Reflexive Language: Reported Speech and Metapragmatics, 3358. Cambridge: Cambridge University Press.
Silverstein, Michael and Greg Urban (eds.)
1996
Natural Histories of Discourse. Chicago: University of Chicago Press.
Stocking, Jr., George W. (ed.)
1996
Volksgeist as Method and Ethic: Essays on Boasian Ethnography and the
German Anthropological Tradition. (History of Anthropology, Volume 8.)
Madison, WI: University of Wisconsin Press.
166
Wataru Koyama
Tannen, Deborah (ed.)

1986
Framing in Discourse. Oxford/New York: Oxford University Press.
Thomas, Jenny
1996
Verschueren, Jef
1999
Volosinov, Valentin N.
[1929] 1973 Marxism and the Philosophy of Language. Cambridge, MA: Harvard University Press.
Williams, Glyn
1992
Sociolinguistics: A Sociological Critique. London: Routledge.
6.

Winfried Nth
1.
Semiotics as a general framework of pragmatics
Levinson (1983: 15) gives the standard account of how semiotics, the general
study of signs, is considered to be fundamental to linguistic pragmatics. The author of the Cambridge Textbook on Pragmatics begins with Charles W. Morris
(19011979), the founder of a general semiotics first outlined in 1938. Morriss
science of signs (1938: 12) is both broader and narrower than the scope of modern linguistic pragmatics.1 It is broader insofar as it is not restricted to the study of
verbal communication but deals with signs in all their forms and manifestations,
whether in animals or men, whether normal or pathological, whether linguistic
or nonlinguistic, whether personal or social (1964: 1), acoustic, visual, olfactory,
gustatory, or tactile. It is also broader since pragmatics, as conceived by Morris,
comprises branches of language studies which today fall into the domain of other
disciplines of linguistics, such as psycholinguistics, sociolinguistics, neurolinguistics, and much besides (Levinson 1983: 2). It is narrower insofar as Morris
aims at a science of signs on a biological basis and specifically within the framework of the science of behavior (1946: 80) indebted to the positivist paradigm of
psychological and social behaviorism. As Levinson (1983: 24) points out, the
scope of pragmatics became still narrower in the frameworks of logical positivism
and analytical philosophy of ordinary language: Carnap (1942) restricts pragmatics to the logical study of language use, for Bar-Hillel (1954) pragmatics is the
study of language use involving indexical words, and Montague (1970: 68) conceives pragmatics as a contextual theory of truth conditions founded on intensional
logic.
Sign behavior, according to Morris, involves three main factors: that which
acts as a sign [the sign vehicle], that which the sign refers to [the designatum], and
that effect on some interpreter in virtue of which the thing in question is a sign to
that interpreter [the interpretant] (1938: 3). Based on this triad, Morris (1938:
67) defines semiotics as a field of study of the following three domains corresponding to three well-known branches of modern linguistics: syntax (or syntactics), the study of the relation between sign vehicles, semantics, the study of the relations between sign vehicles and their designata, and pragmatics, the study of the
relation between sign vehicles and their interpreters (cf. Posner 1985).
20th century semiotics has developed in many directions, and some of the
schools of semiotic research have largely ignored the pragmatic dimension of sign
processes. This is especially true of structuralist semiotics in the tradition of Fer-
168
Winfried Nth
dinand de Saussure (18571913) and Louis Hjelmslev (18991965), whose focus

is entirely on language as a system and not on language use. It is no wonder, therefore, that Jacob L. Mey, once Hjelmslevs PhD student in Copenhagen, in his Introduction to Pragmatics, disregards semiotics entirely as a framework of linguistic pragmatics and that Dan Sperber and Deirdre Wilson, while reducing
semiotics to a theory of codes, dismiss its relevance to pragmatics altogether
(Sperber and Wilson 1986: 7, 2).
The structuralist and behaviorist heritage of semiotics has been an important
obstacle to a more fruitful interdisciplinary dialogue between general semiotics
and linguistic pragmatics. For a long time, it has impeded the spread of the seminal
ideas which the semiotic writings of Charles S. Peirce (18391914) offer to linguistics in general (Nth 2002a) and pragmatics in particular. Only relatively recently has the relevance of these ideas to pragmatics become acknowledged,2 but
so far mostly only by philosophers of language and hardly by linguists (e. g., Levinson 1983: 1, fn. 1).
When Morris conceived the triadic subdivision of semiotics into syntactics, semantics, and pragmatics, he was clearly influenced by Peirce. Peirce had also proposed a triadic framework for the study of signs, but the three branches of semiotics which Peirce set up describe a much wider field of study. Peirce was
inspired by the medieval canon of the three liberal arts, the trivium of grammar,
logic, and rhetoric.3 In contrast to the Saussurean tradition, in which linguistics is a
branch of semiology (cf. Nth 2000: 7273), Peirce considers linguistics and semiotics as two different branches of the system of sciences. Whereas linguistics, in
Peirces detailed classification of the sciences of 1902 (CP4 1.203283), is a
branch of the empirical or special sciences, semiotic(s), the quasi-necessary, or
formal, doctrine of signs, belongs to the normative and formal sciences (CP 2.227,
c. 1897). In contrast to the empirical sciences, which study what is in the actual
world, semiotics, as a normative science, aims at inquiring into what must be the
character of all signs and what would be true of signs in all cases (CP 2.227).
The normative approach is particularly evident in logic since to decide whether a
statement or conclusion is true or false is a normative decision. This is why, to
Peirce, semiotics was only another name for logic (CP 2.227, c. 1897).
The three branches of semiotics, in Peirces redefinition of the trivium of the
liberal arts, are the following: (1.) speculative (or formal) grammar, (2.) logic
(proper), and (3.) pure, formal or speculative rhetoric.5 Speculative grammar, the
precursor of Morriss syntactics, is the study of signs as such; it inquires into the
general conditions of signs being signs (CP 1.444, c. 1896), and it is a general
theory of the nature and meanings of signs, whether they be icons, indices, or symbols (CP 1.191, 1903). Of the branches of modern linguistics, not only syntax, but
also phonology and morphology belong to this branch of semiotics. Logic proper,
the second branch of semiotics, which Peirce also calls critic, is the precursor of
Morriss semantics and certainly also comprises linguistic semantics. According to
169
Peirce, it deals with the relations of signs to the objects that they represent, being
the science of what is quasi-necessarily true of signs [] in order that they []
may be true; [ it] is the formal science of the conditions of the truth of representations (CP 2.229, c. 1897).
Speculative rhetoric, the precursor of Morriss pragmatics and with it of linguistic pragmatics, deals with the relations of signs to their producers and interpreters; it studies the effects created by signs in the participants of the process of
semiosis. Speculative rhetoric is the study of the necessary conditions of the
transmission of meaning by signs from mind to mind, and from one state of mind to
another (CP 1.444, c. 1896); it deals with the formal conditions of the force of
symbols, or their power of appealing to a mind (CP 8.342, 1908). Speculative
rhetoric is broader than Morriss pragmatics in two respects. First, to study how
signs are transmitted from one state of mind to another means that this branch of
semiotics does not only study signs in communication but also in thought not communicated to any other interpreter. Thinking has pragmatic dimensions since it always proceeds in the form of a dialogue a dialogue between different phases of
the ego (CP 4.6, 1906). Second, speculative rhetoric is also conceived as a rhetoric of fine art, a rhetoric of practical persuasion, and a rhetoric of science (EP
26: 329, 1904). Peirce had the vision that the rhetoric of science, which he also
called methodeutics (Santaella 1999), was destined to grow into a colossal doctrine which may be expected to lead to most important philosophical conclusions
(CP 3.454, 1896).
2.
Semiotic foundations of communication
Linguistic pragmatics studies how language is used in communication, especially,

how verbal messages are exchanged between speakers and hearers in dialogue.
Merely incidentally transmitted utterances, monologues, and other forms of language use without an addressee are of no concern to the field of study (cf. Levinson
1983: 16).
Communication is a central issue of semiotics; its study is also called semiotics
of communication (Santaella and Nth 2004). The general study of signs and sign
processes may thus provide a theoretical framework for linguistic pragmatics.
Among its topics of relevance to linguistic pragmatics are: (1) models and functions of communication, (2) the instrumental view of language in communication,
(3) the problem of intentionality, (4) the nature of cooperation in conversation, and
(5) the dialogical nature of the sign and its relevance to the study of communication.
170
2.1.
Winfried Nth
Models and functions of communication
Most theories of pragmatics take a very rudimentary model of communication as

the point of departure of their analyses of speech acts. With reference to Grice
(1957), Levinson considers the triad S-U-R, sender (speaker, writer), utterance,
and receiver (hearer, reader), to be fundamental: Communication consists of the
sender intending to cause the receiver to think or do something, just by getting
the receiver to recognize that the sender is trying to cause that thought or action (Levinson 1983: 16). The utterance, in this triad, is a linguistic token; U has
meaning, which becomes mutual knowledge to S and R in the process of communication, but initially, U may have two kinds of meaning, possibly in conflict, a
speaker meaning and a sentence meaning (Levinson 1983: 16).
Leech, in his Principles of Pragmatics, proposes a flow diagram of communication according to which an S uttering U to H is engaged in three kinds of semiotic
transaction: (1) an interpersonal transaction, or discourse, (2) an ideational transaction or message transmission, and (3) a textual transaction or text (Leech 1983:
59). As to its three constituents, S-U-H, and its notion of message transmission, this
model is certainly inspired by the technical model of the flow of information from
S to H by Shannon and Weaver (1949). This model is inadequate for the study of
human communication, since it is linear, unidirectional, and assigns a merely passive role to the recipient of the message, making the sender the only active agent in
communication. It disregards Hs autonomy in the process of interpreting Ss message as well as the important elements of circularity by which S and H are connected
in communication through feedback loops by which H contributes to the production
of Ss message from the moment of its articulation on (cf. Nth 2000: 24446).
A much cited communication model in linguistic pragmatics is the semiotic
model of the six factors and functions of communication which Roman Jakobson
proposed in extension of Bhlers triadic organon model of language (Bhler
1933). Jakobson defines the factors involved in a process of communication as follows: An addresser sends a message to an addressee. This message refers to a context (the referent of the verbal message) seizable by the addressee. Addresser
and addressee have a common code, such as the vocabulary and grammar of the
language in which they communicate, and finally, there is a contact which serves as
a physical channel and psychological connection between the addresser and the
addressee, enabling both of them to enter and stay in communication (Jakobson
1960: 353).
Each of these six factors is the determinant of a specific communicative function; one and the same message may have several functions, but one is usually predominant. Messages which focus on the referent (context), such as informative, descriptive, or narrative texts, have a predominantly referential function. The
emotive or expressive function is predominant when the focus is on the speakers
attitude in relation to the meaning of the message. The use of interjections, a tremb-
171
ling voice, or the frequent use of the pronouns I or me are examples of signs of an
utterance in which the emotive function of language is predominant. The conative
function is oriented toward the addressee; its most typical grammatical expressions
are the vocative and the imperative. The phatic function predominates in messages
which, without necessarily conveying any particular meaning at all, serve to establish, to prolong, or to discontinue communication, to check whether the channel
works (Hello, do you hear me?), to attract the attention of the interlocutor or to
confirm his continued attention (Are you listening?) (1960: 355). The metalinguistic function predominates when language is used to refer to language and verbal communication; metalanguage is language about language, e.g., as in grammar
books or in the discussion of questions of terminology. Finally, the poetic function,
according to Jakobson, predominates when the focus of verbal communication is
on the qualities of language irrespective of its message, e.g., on the sound pattern,
rhythm, or meter of a text.
Although Peirce considers the model of the sign to be more fundamental than
the model of communication, it is not true that the founder of modern semiotics had
little to say about the topic of communication, as Habermas (1995) believes. Elements of a Peircean theory of communication have meanwhile been reconstructed.7 In his manuscript on Pragmatism of 1907, Peirce defines communication
of a speaker with a hearer as follows: Signs mostly function between two minds,
or theatres of consciousness, of which one is the agent that utters the sign (whether
acoustically, optically, or otherwise) while the other is the patient mind that interprets the sign (EP 2: 403). In this description, the characterization of communication as a theater, a staged scenario of acting agents, is more than a mere metaphor. Actors on stage do not act autonomously, but they represent actions and act
on behalf of other agents. But who should be the agents in communication if not S
and H? This question is the topic of sections 2.2 and 2.3.
2.2.
Are verbal signs instruments of those who communicate?
Most approaches to linguistic pragmatics are based on the assumption that verbal
messages are instruments by which senders seek to influence receivers. Language
use is regarded as a means to some end. Searle (1969: 48), e.g., writes: The sentence uttered by a speaker who knows its meaning then provides a conventional
means of achieving the intention to produce certain illocutionary effects in the
hearer. Leech (1983: 36), too, postulates that speaking is a means to an end and
hence an instrument for a purpose: spoken or written messages are transferred or
transmitted from a speaker to a hearer, and the speakers purpose in this transaction is to convey a particular illocutionary force to the hearer (Leech 1983:
5860).
The view that verbal signs are used as instruments has been defended since antiquity. It is central to one of the earliest treatises on the philosophy of language,
172
Winfried Nth
Platos Cratylus. In this dialogue on the relation between words and the objects for
which they stand, Socrates compares the practical utility of the craftsmans tools
with the utility which words have for a speaker. A tool, says Socrates, is naturally
adequate and appropriate to the task for which it serves. To the weaver, the right
and natural (391E) instrument is the shuttle, to the blacksmith, the anvil and the
hammer serve best to shape the iron, and to the shoemaker, the awl is the natural
tool for piercing leather. As any practical action requires its proper tool, so does the
act of speaking. The speakers appropriate instruments for communicating their
ideas are their words (388a-389c). Practical instruments have a practical utility,
whereas words serve a semiotic utility. As the weavers shuttle is useful for separating the warp from the woof, so is the word useful to the speaker who wants to
communicate or to organize his or her ideas.
In the history of semiotics, the instrumental theory of the sign has lived on in
many variants (cf. Nth 2009). In the context of pragmatics, Wittgensteins (1953)
revival of the topic is of special interest. Instrument, or tool, are appropriate
characterizations of verbal signs in Wittgensteins philosophy of language since
these terms illustrate that the proper study of verbal signs is the study of their use.
What language and instruments have in common is that they are both used and
have functions: Think of the tools in a box: there is the hammer, pliers, a saw,
screwdriver, a glue-pot, nails and screws. The functions of words are as diverse
as the functions of these objects ( 11). Whereas Socrates focuses on the utility of
verbal signs in communication, Wittgensteins focus is on their meaning and the
way meaning is revealed in the use of signs. Not only is language in general an instrument, but also its concepts are instruments ( 569). Instruments as well as
language have their sense in their employment: Look at a sentence as an instrument and its sense as its employment ( 421).
Although the view that language serves as an instrument in communication is a
proverbial commonplace, Peirce contests it. A tool is only a means, i.e., an efficient
cause, but not an end, i.e., not a final cause, in the craftsmans process of producing
an effect. Words used in communication, by contrast, are more than mere tools
since they have purposes of their own, which is not their speakers purpose. What a
word means is not determined by its user. The truth conditions of a proposition are
not determined by the utterer of the sentence which it contains either; they are determined by the reality to which it refers and by the consequences the utterance
has. Whether an argument is valid or not is determined by the logic of language and
not by the person who argues, and whether a speech act is felicitous or not depends
on whether it fulfills certain felicity conditions that cannot be ignored by the utterer. Hence, to use words means to be constrained by restrictions determined by
language and the consequences of its use.
Instead of saying that it is the speaker who produces the utterance in accordance with the rules of syntax, semantics, and pragmatics, we can also say that these
rules determine the speaker to speak the way he or she speaks, but the pragmatist
173
account of language and meaning goes a step further. According to Peirces pragmatic maxim (see 3.1), it is not the rules of language and logic that determine our
utterances, but the anticipated feedback which the success or failure of our utterances gives us whenever we speak. The anticipation of the effects of our words
makes us utter them in this or that way. Nor can it be argued that the utterers
thoughts in which his or her intention to speak is conceived use the ensuing utterance as instruments to put the speakers intentions into effect. Utterances which express thoughts are interpretants of these thoughts, which make the interpretants the
final causes of the thoughts. Since the utterance is itself a sign with the purpose of
creating further interpretants in the hearers mind it cannot be merely instrumentally (or efficiently) caused by the thoughts in which they were conceived.
Nothing could be less in accordance with Peirces semiotic premises than a
dualist assumption of internal thoughts becoming merely externalized as spoken or
written words in communication. Sperber and Wilson (1986: 1), for example, make
this assumption when they reflect on the relationship between their own thoughts
and their expression in writing as follows: what we put down on paper are little
dark marks []. As for our thoughts, they remain where they always were, inside
our brains. Peirce, by contrast, argues that signs exert their semiotic agency inside
as well as outside of human brains, and perhaps even more so outside than inside.
He underlines the latter view by saying that it is even much more true that the
thoughts of a living writer are in any printed copy of his book than that they are in
his brain (CP 7.364, 1902). Hence, thought-signs do not remain contained in the
container of a human brain just as written words do not remain fixed in their meaning once they are written down. Words are symbols which live and grow outside of
human brains in the semiotic effects which they have, and this growth takes place
both in thought and in communication. What they have in common is their dialogical nature:
Before the sign was uttered, it already was virtually present to the consciousness of the
utterer, in the form of thought. But [] a thought is itself a sign, and should itself have
an utterer (namely, the ego of a previous moment) to whose consciousness it must have
been already virtually present, and so back. Likewise, after a sign has been interpreted,
it will virtually remain in the consciousness of its interpreter, where it will be a sign []
and, as a sign should, in its turn, have an interpreter, and so on forward. (EP 2: 403,
1907; see below, 2.4 for the argument of the infinite regress.)
In sum, if signs have purposes, or final causes, and exert a semiotic agency of their
own, they cannot be the mere instruments, that is, instrumental (or efficient)
causes, of other signs; and since both thoughts and utterances are signs, utterances
cannot be the instruments of the thoughts in which they were first conceived.
Although the sign is not an instrument, instruments are necessary for the dissemination of signs. We need a pen and paper to write down words, and our voices
are necessary instruments by which our words reach their hearers, but these words
themselves are not our instruments. Not words are our semiotic instruments, but
174
Winfried Nth
our voices, tongues and lips, pens, paper, typewriters, telephones, or computers
are. The instruments by which thought-signs exert their semiotic agency are the
neurons, the brain lobes, and the perceptual organs, eyes and ears, by means
of which thoughts are processed. This is probably Peirces most radical thesis:
our brain is not the only cause of our utterances; it is merely the efficient cause
of the thoughts and of their expression. The ideas on which this approach to semiosis is based have parallels in current quasi-biological positions of the philosophy
of mind known as memetics and teleosemantics; the biolinguistic assumption is
here that words in culture spread like genes spread in biological species (cf. Nth
2009).
In order to deconstruct the autonomy of the speaking subject, Peirce, for the
sake of the argument, may exaggerate here and there when he speaks of the purpose of the sign which is not the purpose of the user. Elsewhere, he acknowledges
two purposes, the purpose of the sign and the purpose of the sign user, in a mutual
feedback control circuit. In an evolutionary perspective, he describes the loop in
which sign users using signs and signs using sign users interact in terms of a process of co-evolution:
Man makes the word, and the word means nothing which the man has not made it mean,
and that only to some man. But since man can think only by means of words or other external symbols, these might turn round and say: You mean nothing which we have not
taught you, and then only so far as you address some word as the interpretant of your
thought. In fact, therefore, men and words reciprocally educate each other. (CP 5.313,
1868)
There is a certain affinity between this Peircean account of the relationship between the signs and their users on the one hand and Saussures theories of the immutability of the verbal sign and the determination of each signifier by the system
of which it is part on the other. As Stetter (1989: 159) has pointed out in his Saussurean interpretation of the above quote: we are not the masters of our signs who
determine the interpretation of terms but it is rather determined by presuppositions
inherent in the system of language, but Peirces object of the sign is of course
more than the system of which the sign is a part.
The co-evolution of signs and their users can only take place if purposes are ascribed to both. But this does not mean that the sign user is the autonomous agent
external to the signs as which the intentionalists describe him (see 2.3). Peirce
solves the apparent paradox of the conflict between the purposes of the sign user
and the one of the sign with the argument that sign users are themselves objects of
signs (see 2.5) so that their intentions are, in turn, purposes of signs. By no means
do sign users pursue nonsemiotic purposes, for they themselves are and act as signs
in the flow of thought-signs and public signs which constitute them as signs. This is
the line of argument of the famous passage in which Peirce defines the human
being as a sign:
175
There is no element whatever of mans consciousness which has not something corresponding to it in the word; and the reason is obvious. It is that the word or sign which
man uses is the man himself. For, as the fact that every thought is a sign, taken in conjunction with the fact that life is a train of thought, proves that man is a sign; so, that
every thought is an external sign, proves that man is an external sign. That is to say, the
man and the external sign are identical, in the same sense in which the words homo and
man are identical. Thus my language is the sum total of myself; for the man is the
thought. (CP 2.314, 1868)
If the sign user is a sign, his or her intentions are signs, too, and in the speakers utterances the underlying intentions have to be interpreted as objects of signs. As the
object determines the sign so does the sign determine its interpretants. Words only
stand for the objects they do, and signify the qualities they do, because they will
determine, in the mind of the auditor, corresponding signs (CP 2.92, 1902). As
any object of a sign determines the sign which it represents so do the speakers intentions, since they are objects of signs. This is how the speakers intention participates in the creation of an interpretant.
2.3.
Intentionality of communication and the purpose of verbal signs
The classics of linguistic pragmatics agree in considering intentionality as a prerequisite of verbal communication. Levinson (1983: 16) summarizes as follows:
Communication consists of the sender intending to cause the receiver to think
or do something, just by getting the receiver to recognize that the sender is trying to cause that thought or action. So communication is a complex kind of intention that is achieved or satisfied just by being recognized.
Searle (1983: 16466) postulates two intentions. First, there must be an intention to represent on the speakers side, but since it is possible to represent without
the purpose to address or to influence anybody, a second intention is required: the
intention to communicate. Searle (1983: 166) explains that it is impossible to inform a hearer that it is raining without intending that this message represents
truly or falsely the weather condition expressed. Not only the speakers intention
but also the hearers awareness of this intention is a prerequisite of communication. Furthermore, the belief produced in the hearers mind as the result of the
message must have its cause in the speakers intention (cf. Sperber and Wilson
1986: 23).
In structural semiotics, intentionality is the criterion by which communication
is distinguished from signification. The distinction has its roots in the writings of
Buyssens (1943).8 All signs which humans can interpret, including the signs of
physical nature, have signification, but not all of them are intended; only those
signs constitute acts of communication which have an addresser whose intention it
is to convey a message to an addressee. In the tradition of analytical language philosophy, Grice (1957) interprets the difference in question in terms of the dichot-
176
Winfried Nth
omy of natural (i.e., unintentional) and nonnatural (i.e., intended) meanings.

The latter term is unfortunate, though. To call the signs of natural language nonnatural, as Grice does, fails to do justice to the continuity of semiosis in the evolution of life from nature to culture (cf. Nth 2008).
Prieto (1966: 3338) anticipates Searles theory of the two communicative intentions when he distinguishes between two indications which signs have in communication, a notificative and a significative indication. Significative indication is
the one by which a sign indicates whatever it means (i.e., its signification). Notificative indication, by contrast, indicates that the sign is being communicated with
the intention of a sender. The content of a notificative indication is so to speak: Attention! This sign is intended to convey a message to you! (cf. Hervey 1982: 71).
If signs can only be intended or not, intentionality cannot be a matter of degree.
However, intentions are volitions, and wishes and desires may vary in strength.
Addressers may be divided as to their intentions, they may have doubts and hesitate whether they should communicate or not, or they may merely pretend to communicate and act as if they communicated (cf. Parret 1994). Such factors of intentionality are rarely taken into consideration in linguistic pragmatics.
Psychoanalysis, by contrast, has shown how verbal and nonverbal messages
may also communicate hidden intentions which the utterer may want to conceal
but fails to do, e.g., when he or she, impatient with the hearer, loses control over his
or her voice, sounding angry despite the intention to appear calm. In such a scenario, the speakers loss of control over his or her voice is the sign of a conflict between two intentions, one determined by the unconscious mind, the other by the
conscious self. The conflict of intentions is thus the symptom of a divided self in
which a conscious mind aims at concealing what the unconscious mind intends to
express.
Linguistic pragmatics also disregards what the communication theorists
Watzlawick, Beavin, and Jackson (1967) say about the ubiquity of communication
in everyday life. In their study Pragmatics of Human Communication, the authors
postulate the metacommunicative axiom that it is impossible not to communicate,
be it verbally or nonverbally: Behavior has no opposite. There is no such thing as
nonbehavior []. One cannot not behave, and any form of behavior in social interaction is of semiotic relevance since one cannot not communicate (Watzlawick et al. 1967: 4849). Even silence and doing nothing are messages.
Peirces semiotic theory of intentionality (cf. Kappner 2004) and his theory of
the purpose of signs in sign processes are rather incompatible with the views of intentionality held in mainstream linguistic pragmatics.9 Peirce defines the sign as a
medium (MS10 339: 283r, 1906), but not as one which mediates between an addresser and an addressee; instead, the sign logically mediates between its object,
which it represents, and its interpretant, which is the interpretative effect it creates.
The object of the sign is the verbal or nonverbal knowledge, feeling, behavior, or
habit which the sign user associates with the sign. The knowledge of its object is a
177
prerequisite for the interpretation of the sign. While the object of the sign includes
aspects of what linguists define as meaning (see 3.3), the interpretant includes the
illocutionary force and the perlocutionary effect of a speech act, as defined by Austin (1962). In Peirces definition, the interpretant is the cognitive, emotional, or behavioral effect which results from the interpretation of the sign.
According to Peirce, the sign is a semiotic agent with its own intention (cf. 2.2).
Signs continue to exert their agency in the absence of those who first produced
them. This topic is also discussed by Derrida (1988: 8), who uses the metaphor of
the machine for the agency which the written sign exerts irrespective of its producer: To write is to produce a mark that will constitute a sort of machine which is
productive in turn, and which my future disappearance will not, in principle, hinder
in its functioning, offering things and itself to be read and to be rewritten. Peirces
view of the agency of the sign goes beyond the mere assumption of its interpretability in the absence of its writer. Signs have a life of their own, which can be
studied in processes of language change and in the way texts are interpreted in new
ways as time passes and signs change. In 1904, Peirce writes about the purpose of
the symbol: The symbol, by the very definition of it, has an interpretant in view.
Its very meaning is intended. Indeed, a purpose is precisely the interpretant of a
symbol (EP 2: 308).
With purpose, Peirce thus does not mean the sign users intention, but the
signs intention to represent its object and to create and interpretant, i.e., to be interpreted in another sign (MS 1476, 1904). Purpose is thus a semiotic teleology
inherent in the sign. Not only uttered or written signs have purposes, but also
thought-signs. Their purpose is to act in a mental dialogue in which one thoughtsign is translated or interpreted in a subsequent one (CP 5.284, 1868).
On the one hand, the sign represents its object to the mind of an interpreter, on
the other, the object represented by the sign is in a sense the cause, or determinant,
of the sign even if the sign represents its object falsely (CP 6.347, 1909; cf. Parmentier 1985). To justify the autonomy of the sign by its being determined sounds
contradictory, but it describes the process of an autopoietic self-determination of
the sign. Since the object of the sign is not external, but internal to the sign, to say
that the sign is determined by its object means that it is determined by a dynamics
inherent in itself and not by an agency outside of the sign (cf. Colapietro 1995).
The way in which a sign is determined by its object depends on its modes of relation to its object and its interpretants. Rhematic signs, such as words in isolation,
are determined to represent their objects as such, without affirming, questioning, or
negating their existence. Dicentic signs, such as affirmations, are determined by
the truth of what they represent, and arguments are determined by the laws of logic
and reasoning in general. Symbols are determined by the habits of their users, by
which they have become symbolic signs. Indices are determined by their objects
because they are related to them by causal, temporal, or spatial relations, whereas
icons represent their objects by means of qualities of their own. All these factors
178
Winfried Nth
determine the sign independently of their use. For example, a rhematic symbolic
legisign such as the word apple, uttered by a speaker, is determined by the habits of
the speakers of English to represent the fruit of the apple tree by means of this
word; the word represents its object as a general kind without affirming the existence of any particular fruit. The sign user may try to escape from his or her determination by the object of this sign and use this word to refer to a pear, but in the
long run, words cannot be used in this way because the speech acts will fail. A sign
user may also lie or argue against the laws of logic, but in the long run, to ignore the
truth and the laws of reasoning will not pay. Furthermore, verbal signs are determined by biolinguistic and evolutionary constraints, which have restricted the
speakers possibilities of articulation and perception (cf. Deacon, 1997: 116).
2.4.
Conversational cooperation, commonage, and the common ground
The model of communication underlying the classical theories of linguistic pragmatics is one of rational discourse.11 The aim of dialogic communication is guided
by the ideal of mutual consensus; the speakers purpose must be achieved. Communication is modeled as being cooperative; its goal is not only comprehension but
also the full acceptance of the intended message on the part of the hearer; speech
acts can only succeed or fail, and successful speech acts fulfill felicity conditions;
the speakers intentions are described as being sincere, and the messages are always relevant. Such accounts of verbal exchange between speakers and hearers
model communication according to the ideals of a desirable symmetry between the
agents interacting in dialogue. The underlying assumption is that a full congruence
between the speakers intended meaning and the hearers understanding of it is not
only possible but also desirable.
To reach consensus is indeed an ideal of dialogic interaction in rational discourse. The very concept of communication implies, in its root, the notion of making something common to those who communicate. Morris (1946: 195) expresses
this normative goal of communicative interaction in a definition according to
which communication is the use of signs to establish a commonage of signification. The assumption that commonage is an essential characteristic of communication is not undisputed. In everyday language, e.g., it is contradicted by the
military metaphors by which speakers represent their arguments as positions
which need be defended (Lakoff and Johnson 1980: 7). In the sociological theory
of communication, Luhmann (1984) starts from the basic assumption that the differences between the utterers and the interpreters frames of mind instead of their
common ground, is the source of all dialogic interaction, whereas Habermas
(1981) founds his Theory of Communicative Action on the assumption that consensus is constitutive of rational communicative interaction.
As a prerequisite for, and means of, reaching the normative ideal of dialogic
commonage, linguistic pragmatics recognizes such principles as Grices maxims
179
of conversational cooperation and its derivative, the principle of relevance advocated by Sperber and Wilson. A precursor in the study of these normative principles is Peirce (Pietarinen 2004, Short 2007: 21314). Although by no means restricted to the study of rational discourse (Oehler 1995: 26768), Peirces
speculative rhetoric describes conversational cooperation as follows:
Honest people, when not joking, intend to make the meaning of their words determinate,
so that there shall be no latitude of interpretation at all. That is to say, the character of
their meaning consists in the implications and non-implications of their words; and they
intend to fix what is implied and what is not implied. They believe that they succeed in
doing so, and if their chat is about the theory of numbers, perhaps they may. (CP 5.447,
1905)
However, immediately after this description of rational conversational cooperation, Peirce also points out that commonage can only serve as a regulative idea, an
ideal which can rarely be reached in ordinary conversation. The sentence after the
above passage goes on to say: But the further their topics are from such presciss12,
or abstract, subjects, the less possibility is there of such precision of speech. Insofar as the implication is not determinate, it is usually left vague (CP 5.447,
1905). Furthermore, Peirce also argues that, despite all efforts of cooperation, dialogues also evince a fundamental divergence of interest between the participants.
From this perspective, Peirce calls the participants of dialogue opponents (MS
515: 25, s. d.; cf. Hilpinen 1995: 293) and describes dialogic interaction in terms of
a metaphorical scenario of war: The utterer is essentially a defender of his own
proposition and wishes to interpret it so that it will be defensible. The interpreter,
not being so interested, and being unable to interpret it fully without considering to
what extreme it may reach, is relatively in a hostile attitude, and looks for the interpretation least defensible (MS 9: 34, c.1903).
Relevance, so much discussed in linguistic pragmatics since Grice, is another
principle about which Peirce has written (cf. Pietarinen 2004). Peirce discusses it
as follows: If the utterer says Fine day! he does not dream of any possibility of
the interpreters thinking of any mere desire for a fine day that a Finn at the North
Cape might have entertained on April 19, 1776. He means, of course, to refer to the
actual weather, then and there, where he and the interpreter have it near the surface
of their common consciousness (MS 318: 3233, 1907). Whereas Grice would
describe the elliptic utterance Fine day as one which requires filling the gap of
what remains unsaid by means of nonconventional conversational implicatures,
Peirce would consider most conversational implicatures unnecessary since communication, studied from the semiotic perspective, is not only verbal communication. What linguistic pragmatics describes as gaps between what is said and what
is left out may be a gap in the verbal message, but interpreters of a dialogue do not
only interpret words; they also interpret nonverbal signs and the situational circumstances of the utterance. What appears a gap in the verbal message may not be
a gap in the multi-channel message of the dialogic exchange. Furthermore, not
180
Winfried Nth
everything that is left open and undetermined in conversation needs to be filled by

implicatures either. It is wrong to assume that the speakers message is always
clearly determined and that the interpreters only task is to fill the gaps in order to
restore the intentions which seem unclear. Vagueness and the impossibility of resolving it may be the speakers purpose and the interpreters source of creative
thinking (see 3.3).
In Peirces speculative rhetoric, commonage is not only a normative goal; it is
also its prerequisite of the dialogue. As the prerequisite of communication, it belongs to the domain of the object of the sign; as its goal, it pertains to the interpretant created by the sign. The object of the sign can only be known if both the utterer
and the interpreter have collateral experience of the Real or Dynamical Object of a
sign. In order to know what a sign represents, we need to have previous acquaintance with what the sign denotes (CP 8.179, s. d.). The sign itself cannot express its
dynamical object, it can only indicate and leave the interpreter to find out by collateral experience what it represents (EP 2: 198, 1909). There is nothing nonconventional in this process of interpretation, as Grice assumes with his implicatures. Interpretation, according to Peirce is always a natural process of abductive,
inductive, and deductive inferencing.
An aspect of the object of the sign, i.e., that which must be presupposed for
communication to be successful, is the common ground of experience and knowledge which the utterer and the interpreter share. Both Grice and Searle have described it as a prerequisite of conversation. Pietarinen (2004: 302305) shows in
which respect Grices and Searles theories of the preconditions of successful communication were anticipated by Peirce. In 1908 Peirce formulates it as follows:
No man can communicate the smallest item of information to his brother-man unless they have [] common familiar knowledge [] such that each knower knows
that every other familiarly knows it, and familiarly knows that every other one of
the knowers has a familiar knowledge of all this (MS 614: 12, in Pietarinen
2006: 438). To make clear that no circulus vitiosus is involved in this endless reciprocity, Peirce underlined that this infinite regress is a logical and not a psychological one: Of course, two endless series of knowings are involved; but knowing
is not an action but a habit, which may remain passive for an indefinite time (MS
614: 12, in Pietarinen 2006: 438).
Insofar as commonage is the goal of communication, it pertains to the interpretant of the sign. Peirce calls the meaning effects common to the speaker and the
hearer the cominterpretant of the sign. At the same time, he recognizes that there
may be effects of meaning which are specific to the utterer and the interpreter, the
former being the intentional, the latter the effectual interpretant:
181
There is the Intentional Interpretant, which is a determination of the mind of the utterer;
the Effectual Interpretant, which is a determination of the mind of the interpreter; and
the Communicational Interpretant, or say the Cominterpretant, which is a determination
of that mind into which the minds of utterer and interpreter have to be fused in order that
any communication should take place. This mind may be called commens. It consists of
all that is, and must be, well understood between utterer and interpreter at the outset, in
order that the sign in question should fulfil its function. (EP 2: 478, 1906)
The commind, as Peirce calls the fusion of the two minds in one and the same
commens elsewhere, is a normative ideal. In practice, the ideal of a commind can
never be fully reached since mutual understanding is only possible in a fragmentary way. In 1907 (MS 318), Peirce points out that no interpreter can be said to have
access to the utterers mind. Knowledge about what a speaker means can only be
obtained in fragments, and such fragments are mere copies of a scrap torn out of
anothers life. By such scraps we can only supplement the ideas of [our own]
life (MS 318: 194, 1907; cf. Johansen 1993: 198202). As interpreters we must
match those fragments found in the signs of the utterance with our own discourse
universe and find out where they can be inserted or recopied in our own panorama of universal life (MS 318: 194, 1907). Interpretation is thus a semiotic
patchwork put together by abductive, inductive and deductive reasoning.
2.5.
Dialogue as the prototype of verbal communication, and the dialogical

nature of the sign
Peirce not only considered the dialogue as a wonderfully perfect kind of signfunctioning (EP 2: 391, c. 1906), he also had derived his first model of the sign together with the first version of his three categories, the I, the IT, and the THOU
from the three elementary constituents of a dialogue (W 113: xxvii, 174, 1865).14
Fisch explains how the model of the sign is derived from the model of the dialogue:
Peirce began where most of us begin, with a model, which, taken by itself, would
suggest too narrow a definition; namely the model of conversation between two
competent speakers of the same natural language say, English. [] It goes without saying that words are signs; and it goes almost without saying that phrases,
clauses, sentences, speeches, and extended conversations are signs (Fisch 1986:
357).
In Peirces fully developed semiotics, a sign, or representamen, is something
which stands to somebody for something in some respect or capacity. It addresses
somebody, that is, creates in the mind of that person an equivalent sign, or perhaps
a more developed sign. That sign which it creates I call the interpretant of the first
sign. The sign stands for something, its object (CP 2.228, c.1897). Sign, object,
and interpretant must be understood as following each other in time. The object
comes first insofar as it precedes the sign; it is that which must be known before the
sign can be understood. The sign comes next since it represents the object to an in-
182
Winfried Nth
terpretant, and the interpretant follows last since it is created by the sign. This temporal sequence makes clear how the three correlates of the sign correspond to the
three correlates of a dialogue, addresser (or sender), message, addressee (or receiver): the sign (or Representamen) is the message; it addresses the interpretant as
its receiver, and the sender of this message is its object, which determines the sign
(cf. Bergman 2003: 11, Pietarinen 2006: 426).
Despite the close association of the interpreter and the interpretant in the process of interpretation, the interpretant must not be confused with the mind of the
addressee of a message. In its full definition, the interpretant is any significant effect of the sign (CP 2.303, 1902), be it a feeling, action, or thought, and since significant effects of signs are not restricted to individuals and to the present moment
of sign interpretation, the interpretant can also be a memory, a habit, a fashion, an
innovation, a revolution, a war, or a peace treaty. In all cases, the interpretant is
itself a sign.
The object of the sign corresponds to its sender in an equally impersonal understanding (cf. 2.2 on the utterers intention as an object of the sign). Ransdell (1977)
explains why Peirce equates the two instances of conversation which linguists are
used to distinguishing as the referent of the sign on the one hand and the sender of a
message on the other: to the degree that the reality, the phenomena which we perceive, experience, or think about, are interpretable phenomena, they are objects of
signs. Reality is a sign producer since it is the cause of the signs which we have of
it; after all, we interpret reality by means of its signs, which in turn create interpretants. Both the sender of a message who, although not a creation of this message,
is nevertheless the source of its distribution, and the signs object precede the
moment and are the causes of the production of the sign.
Not only the sender but also the sign itself has an agency in the process of semiosis, as shown above (2.23). Peirce calls it the action of the sign (e.g.,
CP 5.47273, 1905), but again, this agency, which is due to the purpose of the sign,
is neither restricted to the purpose of a human being nor must it be the effect of a
verbal sign. There are even signs which have no sender at all. These are evidently
the natural signs (cf. Nth 2008). Since there can be no isolated sign, Peirce attributes a Quasi-utterer to (natural) signs without a real utterer (CP 4.551, 1906).
There are also signs without an interpreter but a Quasi-interpreter (CP 4.551,
1906), for example, if nobody takes the trouble to study the record (EP 2: 404,
1907). According to Peirces dialogic semiotics, it is essential that the utterer or
quasi-utterer as well as the interpreter or quasi-interpreter are not instances exterior to, but can be found within the sign, in which they must nevertheless be distinct although they are, so to say, welded in the sign (CP 4.551, 1906). A summary of the resulting simultaneity of the presence and absence of the utterer in the
dialogical sign is the following:
183
The action of a sign generally takes place between two parties, the utterer and the interpreter. They need not be persons; for a chameleon and many kinds of insects and even
plants make their livings by uttering signs, and lying signs at that. Who is the utterer of
signs of weather, which are not remarkably veracious, always? However, every sign certainly conveys something of the general nature of thought, if not from a mind, yet from
some repository of ideas, or significant forms, and if not to a person, yet to something
capable of some how catching on. (MS 318, S. 17, 1907)
In the dialogic process of conversation, the continuous alternation between the role
of the utterer and the one of the hearer involves a permanent transformation of
signs into interpretants and interpretants into the objects of new signs creating new
interpretants. This continuous process of alternations and transformations is a
source of creativity in semiosis. Since not only the utterers but also the objects of
the signs participate in the dialogue in the sense that they determine its interpretants, not only the voices of two participants but a plurality of voices is audible.
Dialogue is a Bakhtinian polyphonic chorus with a plurality of participants speaking the voices of many discourses (cf. Nth 2009).
3.
Pragmatism and Peirces semiotics of meaning and reference
Peirces semiotics is also relevant to issues of linguistic pragmatics such as the relationship between and the respective scopes of semantics and pragmatics, the difference between speaker, hearer, and utterance meaning, the nature of reference
and indexicality in language, and the theory of speech acts. Even the concept of
pragmatics has its origin in Peirces writings.
3.1.
Pragmatics, pragmatism, and pragmaticism
The term pragmatics, as introduced by Morris (see 1.), was clearly inspired by the
philosophy of pragmatism, whose main representatives in the first decades of the
20th century were William James and John Dewey, and whose founder was Peirce
(cf. Murphy 1990). In a letter of 1912, Peirce explained why and how he coined the
term in the 1870s (Houser 1998: xxi-xxii): When I gave the doctrine of pragmatism the name it bears [] I derived [ it] from , behavior, in order
that it should be understood that the doctrine is that the only real significance of a
general term lies in the general behavior which it implies (in: Eisele 1987: 95).
Morriss change of the concept of pragmatism to pragmatics made the term more
compatible with the names of the neighboring fields of pragmatics in linguistics
syntactics and semantics as well as with the term linguistics itself.
One of the main aims of pragmatism is to overcome the dualism between
thought and action (Colapietro 1992: 43031). Peirce formulated the essence of
pragmatism, as he conceived it, in several versions of his pragmatic maxim for-
184
Winfried Nth
mulated in several versions since 1878. In 1902, the maxim stated the following:
Consider what effects, that might conceivably have practical bearings, we conceive the object of our conception to have. Then, our conception of these effects is
the whole of our conception of the object (CP 5.9, c. 1905). Peirce gives the following account of how this maxim should be applied: Pragmatism is the principle
that every theoretical judgment expressible in a sentence in the indicative mood is
a confused form of thought whose only meaning, if it has any, lies in its tendency to
enforce a corresponding practical maxim expressible as a conditional sentence
having its apodosis in the imperative mood (CP 5.18, 1903). Colapietro (1992:
433) gives an example of how this principle can be followed in practice: To learn
most clearly the meaning of an assertion e.g., of Sugar is soluble requires us
to translate the assertion into a maxim such as If you place this substance into
water or similar liquids it will dissolve.
Disappointed with how his word coinage became abused in the merciless way
that words have to expect when they fall into literary clutches (CP 5.414, 1905) in
a trend that culminated in the deterioration of pragmatism into behaviorism
(Nadin 1993: 223), Peirce, in 1905, began to replace the term pragmatism with the
neologism pragmaticism, a word ugly enough to be safe from kidnappers, as its
author remarked (CP 5.414, 1905).
Parret (1983: 3) believes that there are three main lessons which the philosophy
of pragmatism has taught to linguist pragmatics: Meaning is context bound,
rationality is discourse bound, and signifying should be seen as understanding.
What linguistic pragmatics has learned from the tradition of pragmatism is certainly the importance of the study of language with respect to its consequences for
human action and behavior.
A common ground of Peirces semiotics and the classics of linguistic pragmatics is the normative approach to communication. Among the other insights
which Peirce has contributed to linguistic pragmatics is the pragmatic principle of
expressibility often attributed to Searle. Searles (1969: 220) definition is: For
any meaning X and any speaker S whenever S means (intends to convey, wishes to
communicate in an utterance, etc.) X then it is possible that there is some expression E such that E is an exact expression of or formulation of X. Peirce reduces it to the form: thought and expression are really one (CP 1.349, 1903).
3.2.
Reference and indexicality
Reference and deixis are key issues of linguistic pragmatics; in Peirces semiotics,
in which the term reference is hardly used (cf. Nth 2006), these topics are studied
as indexical signs. Searle (1969: 2629, 7296) studies reference as a speech act.
Expressions may be used to refer or they may not be used to refer. A man came is
used to refer to a man who came; John is a man is not used referentially, but
predicatively (Searle 1969: 27). Necessary conditions of a referring speech act are:
185
whatever is referred to must exist (1969: 77); There must exist one and only one
object to which the speakers utterance of the expression applies, and: The hearer
must be given sufficient means to identify the object from the speakers utterance
of the expression (Searle 1969: 82).
Existence, singularity, and referential selectivity are criteria of the genuine
index in Peirces semiotics. In fact, all of Searles conditions of reference are conditions of the genuinely indexical sign: A genuine Index and its Object must be
existent individuals (whether things or facts) (CP 2.283, c. 1902), and indexical
signs are directions of what to do to find the object (CP 2.289, c. 1893). In
contrast to Searle, Peirce also studies to what degree nonverbal signs can be genuine indices and to what degree language may be insufficient for the purpose of referring to an individual object. Furthermore, he also studies another kind of index,
which he calls degenerate index. Such indices do not refer directly to individual
existing object; instead, they refer to symbols or mental objects (cf. Nth 2000:
187). Their prototype is the relative pronoun, which is a sign referring to another
sign.
Since not all expressions are used to refer, according to Searle, not all of them
have a referent. In Peirces semiotics, too, only genuine indices have a referent in
the sense of referring indexically to an individual existent object, but although not
all signs have a referent, all signs have an object.15 Peirces semiotics is utterly incompatible with the dualistic theories of referential semantics which split verbal
reference into two mutually independent domains, one being external to the human
mind, the other being mental and hence internal to it. According to Peirce, not reference and sense are the correlates of the sign, but the object and the interpretant,
and all signs have an object as well as an interpretant. All words are signs which
represent an object and determine an interpretant, even the conjunction and. Savan
(1994: 189) describes its object as our experience of how of two or more entities
are combined to form a group, a set, or otherwise a whole, but since the object of
the sign need not be a material thing, we should add that the idea of a symmetrical
conjunction is among the objects of this sign.
A sign without an object would be a contradiction in terms in Peirces semiotics. In contrast to the referent of logical semantics, which is defined as an individual or class of existents, Peirces object of the sign may also be a feeling, an
experience, a cognition, a thought, an imagination, or even a fictional event. Unlike
the sense or meaning of a word, the interpretant is not necessarily a mental concept, a thought, or an idea corresponding to the sign; it can again be a feeling, an
action, or even be a material thing, provided it is a result of a sign. For example, the
cake baked according to a cooking recipe is among the interpretants of the recipe
practically interpreted by the cook. The object of the sign may also be something
believed formerly to have existed or expected to exist or something of a
general nature (CP 2.232, 1910). Only rarely is it a thing. It is rather that information, knowledge, or experience which an observer of the sign must have in
186
Winfried Nth
order to be able to interpret the sign. Peirce had adopted the term object from 13th
century scholastic terminology where objectum meant a creation of the mind in its
reaction with a more or less real something [] upon which cognition is directed
(MS 693A: 33; Pape 1996b: 115).
The index is defined in opposition to the symbol and to the icon as a class of
signs comprising natural but also conventional signs. An index is really affected
(CP 2.248, c. 1903) by, or physically connected with, its object; the sign and its
object make an organic pair, but the interpreting mind has nothing to do with
this connection except remarking it, after it is established (CP 2.299, c. 1895).
Among the features of Peirces index are that it directs the interpreters attention
towards the object and that it asserts nothing, but shows the object (cf. Goudge
1965: 5354). Examples of verbal and nonverbal indexical signs are: a weathercock, a yardstick, a photograph, a rap on the door, a pointing finger, any other gesture that indicates a present emotion, a cry for help, deictic words, proper names,
possessive, relative, personal, and selective pronouns.
All verbal utterances contain explicit or implicit indices referring to, or identifying, the situational circumstances, the time, place, or persons in whose context
they were uttered. The adequate interpretation of these indices is necessary to distinguish successfully between the relevant and irrelevant interpretation of a sign
(cf. 2.4):
If, for example, a man remarks, Why, it is raining! it is only by some such circumstances as that he is now standing here looking out at a window as he speaks, which
would serve as an Index (not, however, as a Symbol) that he is speaking of this place at
this time, whereby we can be assured that he cannot be speaking of the weather on the
satellite of Procyon, fifty centuries ago. (CP 4.544, 1906)
Morris did not adopt the term index, but his category of identifiors corresponds to
Peirces index (1946: 154, 362). But in contrast to Peirce, Morris restricted the
class of identifiors to spatio-temporal deixis. Identifiors signify locations in space
and time (locata) and direct behavior toward a certain region of the environment.
The identifior has a genuine, though minimal, sign status; it is a preparatorystimulus influencing the orientation of behavior with respect to the location of
something other than itself. Morris distinguished three kinds of identifiors: indicators, which are non-language signals, descriptors, which describe a spatial or
temporal location, and namors, which are language symbols, and hence, substitute signs synonymous with other identifiers.
3.3.
Meaning in dialogue and its vagueness
Several meanings are involved in the utterance of a speaker addressed to a hearer.

Searle (1969: 4548) distinguishes between intended and conventional meanings,
Leech (1983: 6) between pragmatic and semantic meanings, and Levinson (1983:
187
1617) between speaker meanings and utterance or sentence meanings. The various kinds of meaning distinguished by these authors partly overlap, but some of the
criteria according to which these distinctions are made differ significantly.
Leechs semantic meaning roughly corresponds to Levinsons sentence
meaning and Searles conventional meaning. All three kinds of meaning are opposed to what is studied in linguistic pragmatics, i.e., how meaning is intended by a
speaker and understood by a hearer. Leech defines semantic meaning as a meaning defined purely as a property of expressions in a given language, in abstraction
from particular situations, speakers, or hearers (1983: 6). Can such a meaning
abstracted from the circumstances of its use be called conventional, as Searle
suggests when he opposes conventional to intended meaning? The idea that
intended meaning is opposed to conventional meaning is rejected by Strawson
(1971), who points out that conventional meaning is not only semantic meaning
but pertains to pragmatic meaning, too, since speech acts, like words and sentences, are also determined by conventions.
The study of pragmatic meaning, according to Leech (1983: 14), begins with
the study of utterances: In fact we can correctly describe pragmatics as dealing
with utterance meaning and semantics as dealing with sentence meaning. With
this definition, Leechs dichotomy of semantic vs. pragmatic meaning is no longer
compatible with Levinsons dichotomy of utterance meaning and speaker meaning, which suggests that the study of speaker meaning is the domain of pragmatics
and not the study of utterance meaning.
Furthermore, Levinsons distinction between speaker meaning and utterance
meaning raises the question of how speaker meaning can differ from utterance
meaning at all. Is not the utterance meaning the meaning conveyed by the utterer,
i.e., speaker? Can an utterer, by means of an utterance, convey any other meaning
than his own or her own meaning?
The distinctions drawn by the three authors thus do not only overlap, they also
contradict each other. If speaker meaning is the utterers and thus also the intended
meaning, one can no longer oppose it to utterance meaning (as Levinson does).
Speaker meaning can only be intended meaning since whatever a speaker utters
must be intended, but if utterance meaning is speaker meaning and speaker meaning is intended meaning, the congruence of theses meanings makes Leechs distinction between semantic and pragmatic meanings invalid. On the other hand, if
speaker and utterance meanings are intended meanings, and if utterances are determined by conventions, too, Searles distinction between intended and conventional
meanings is invalid.
If there is speaker meaning there should also be hearer meaning. Grice (1989)
takes it into consideration but the dichotomies set up by Leech, Searle and Levinson neglect it. Sbis and Fabbri (1980: 314) criticize this neglect for reducing the
interpretive work of the addressee [] to the mere recognition of a set of coherent
intentions on the part of the agent/speaker. The neglect of the category of a hearer
188
Winfried Nth
meaning in linguistic pragmatics is due to its normative approach to the study of

conversation. Like Chomsky, who pursues a normative approach to syntax when
he postulates the idealized competent speaker as one who commits no mistakes in
uttering sentences, Searle (1969: 20) explicitly adopts a normative approach to the
description of the speakers meanings in conversation when he declares that cases
where the speaker does not say exactly what he means the principal cases of
which are nonliteralness, vagueness, ambiguity, and incompleteness are not theoretically essential to linguistic communication.
Speakers can mean more than they say (Leech 1983: 9). According to Searle
(1969: 45) the difference between what a speaker means and what he or she says is
a difference between the meanings intended by the speaker and those conventionally inherent in his or her utterances. Meaning is more than a matter of intention; it
is also at least sometimes a matter of convention, says Searle (1969). The theory
of implicatures is based on the assumption of such differences between meaning
and saying.
Peirces theory of meaning is pragmatic throughout (cf. Hilpinen 1995: 297).
Meaning is the idea which the sign attaches to its object (CP 5.6, 1905), but
meanings are not a priori inherent in signs; they can only reveal themselves in processes of semiosis (Pape 1996a: 308). Words only have meaning in so far as we
are able to make use of it in communicating our knowledge to others and in getting
at the knowledge that these others seek to communicate to us (CP 8.176, 1903).
Thus, no semantic or conventional meaning can be found in Peirces assumptions. In accordance with the pragmatic maxim, the meaning of an utterance reveals itself in its interpretations as well as in the actions and habits which are its
consequences. In one of his later definitions, Peirce concludes that the meaning of
any [assertion] is the meaning of the composite of all the propositions which that
[assertion] would under all circumstances empower the interpreter to scribe (MS
280, ca. 1905; cf. Pietarinen 2005: 1769). Since utterances have no inherent meaning independent of their use in speech acts, Peirce does not distinguish between utterance and speaker meaning. Furthermore, he distinguishes between many more
kinds of meaning. Some aspects of meaning are associated with the object of the
sign insofar as the sign presupposes collateral knowledge of its object and a common ground of knowledge shared by the speaker and the hearer. Other meanings
pertain to the interpretant, the effect of meaning created by the sign (cf. Bergman
2000: 234 for both). In the theory of the interpretant discussed above, we find the
intentional interpretant as a counterpart to speaker meaning and the effectual interpretant as the counterpart to Grices hearer meaning, but Peirce distinguishes
other kinds of interpretants (cf. Johansen 2002: 48). His immediate interpretant,
the meaning which the sign indicates before it is interpreted, comes close to
Searles conventional meaning, but the interpretant is not always the meaning of a
sign in the linguistic sense. For example Peirce distinguishes between an emotional, and energetic, and a logical interpretant. The first is an emotional, the sec-
189
ond a bodily reaction to a sign. Only the third comes close to the kind of meaning
which linguists usually describe since logical meaning is the meaning of a concept
but since not all words are concepts, words also have meanings which are not logical interpretants.
Peirce also acknowledges the possibility of differences between saying and
meaning so much discussed in linguistic pragmatics, but his account of how meaning is conveyed in the dialogue of an utterer and an interpreter is based on other
premises. Meaning is not determined by rules, conventions, and speakers intentions, as Searle (1969: 4345) sees it, but by habits, common grounds, and the collateral experience of the object of the sign. Meaning is not only vague by exception, but it is essentially vague (Nth and Santaella 2009). If meaning is fuzzy in
language because of the generality and vagueness of words, it is even vaguer if the
feelings, desires, and inner conflicts are considered which speakers and hearers
have when they communicate. Emotions, e.g., are never certain and clear; they are
necessarily vague, merely expressible and interpretable in the form of allusions,
conjectures, surmises, guesses, intuitions, and communicative negotiations. Therefore, no communication of one person to another can be entirely definite, i.e., nonvague. [] Wherever degree or any other possibility of continuous variation subsists, absolute precision is impossible. Much else must be vague, because no mans
interpretation of words is based on exactly the same experience as any other mans.
Even in our most intellectual conceptions, the more we strive to be precise, the
more unattainable precision seems (CP 5.506, c. 1905).
Vagueness and generality are the two major causes of indeterminacy in conversation, but this indeterminacy leads not only to misunderstandings or even failures
of communication. Instead, indeterminate signs are also a source of creativity to
both utterers and interpreters. The interpreter can and must exert his or her own
imagination in determining what the vague or general sign means. He or she can
specify what is left undetermined by associating it with meanings from his or her
own textual universe. The utterer of a vague sign has the advantage that he or she
can suggest meanings without being held responsible for them. Hence vagueness is
not only a risk that endangers communication; it also offers a useful potential of
creativity to those who communicate. Peirce develops these ideas of the advantages and risks of taking and giving freedom of semantic choice with an example of
a generality and one of referential vagueness:
A sign [] is objectively general in so far as it extends to the interpreter the privilege of
carrying its determination further. Example: Man is mortal. To the question, What
man? the reply is that the proposition explicitly leaves it to you to apply its assertion to
what man or men you will. A sign that is objectively indeterminate in any respect is objectively vague in so far as it reserves further determination to be made in some other
conceivable sign, or at least does not appoint the interpreter as its deputy in this office.
Example: A man whom I could mention seems to be a little conceited. The suggestion
here is that the man in view is the person addressed; but the utterer does not authorize
190
Winfried Nth
such an interpretation or any other application of what she says. She can still say, if she
likes, that she does not mean the person addressed. Every utterance naturally leaves the
right of further exposition in the utterer; and therefore, in so far as a sign is indeterminate, it is vague, unless it is expressly or by a well-understood convention rendered general. (CP 5.447, 1905)
Elsewhere, Peirce gives indefinite pronouns as the example of the risks and
chances which generality and vagueness bring to the semantic choices made by the
utterer and the interpreter: Some means that the speaker is to select an instance,
while Every or Any means that a second person is to perform the selection. Of
course, it is easier to satisfy the conditions of a statement if one can select ones
own examples, except for this, that he who undertakes to find an example guarantees that there is one, while if he leaves the selection to another, and there is none,
his statement is not broken down (CP 2.523, fn 1, 1893).
4.
Semiotics of speech acts: Peircean and Greimasian contributions
The theory of speech acts is the branch of linguistic pragmatics to which semiotics
has both contributed as a precursor and to which it has offered alternative approaches. The semiotic precursor of speech act theory is again Peirce; the alternative approach to be presented in the following is an approach based on the semiotics of A. J. Greimas. Only passing reference can be made to the semiotic
elements of a pragmatic theory of speech acts which may be found in the typology
of discourse of Charles Morris (1946: 203205; cf. Nth 2000: 94) and to the rather
different elements of a semiotics of some speech acts developed in the framework
of M. A. K. Hallidays systemic semiotics by Ventola (1987).
4.1.
Peirces contributions to speech act theory
Peirces contributions to speech act theory have been the focus of interest of several studies.16 In wider contexts they have also been discussed by Hilpinen (1995,
1998) and Pietarinen (2004, 2006). Actually, Peirce has even anticipated key terms
and issues of speech act theory, among them Austins thesis that by speaking we
do things and that this doing is an act. In a fragment of 1908, Peirce describes
the speech act of taking an oath in these terms: Taking an oath is not mainly an
event of the nature of a [] Vorstellung, or representing. It is not mere saying, but
is doing. The law, I believe, calls it an act (CP 5.546).
In anticipation of the distinction between an illocutionary act and its propositional content of an utterance later drawn by Searle (1969: 2329), Peirce defines a
proposition as the logical content of a sentence, irrespective of the particular
mood in which it is uttered (cf. Brock 1981a: 281). Sentences differ from utterances as types differ from tokens:
191
A sentence, in the sense here used, is a single object. Every time it is copied or pronounced, a new sentence is made. But a proposition is not a single thing and cannot
properly be said to have any existence. Its mode of being consists in its possibility. A
proposition which might be expressed has all the being that belongs to propositions although nobody ever expresses it or thinks it. It is the same proposition every time it is
thought, spoken, or written, whether in English, German, Spanish, Taglog, or how. A
proposition consists in a meaning, whether adopted or not, and however expressed. (MS
599: 56, ca. 1902)
Among the speech acts which Peirce describes as moods of propositions are assertion, judgment, question, denial, doubt, command, teaching, taking an oath, laying
a wager, lying, poetic language, and forms of politeness.17 Expressing a proposition in one of these moods involves an act, an exertion of energy, and is liable to
real consequences, or effects (CP 5.547, ca. 1908). The exertion of energy associated with these moods is evidently equivalent to Searles illocutionary force of
speech acts. Examples and illustrations of the consequences which such speech
acts have for their utterers and interpreter are: An assertion is an act by which a
person makes himself responsible for the truth of a proposition. Nobody ever asserted that the moon is made of green cheese; yet this is a familiar proposition
(MS 599: 5, ca. 1902). The person who took an oath has only made an assertion if
the denotation of the proposition is a fact (CP 4.500, ca. 1903):
[An oath] would be followed by very real effects, in case the substance of what is asserted should be proved untrue. This ingredient, the assuming of responsibility, which is
so prominent in solemn assertion, must be present in every genuine assertion. For
clearly, every assertion involves an effort to make the intended interpreter believe what
is asserted, to which end a reason for believing it must be furnished. But if a lie would
not endanger the esteem in which the utterer was held, nor otherwise be apt to entail
such real effects as he would avoid, the interpreter would have no reason to believe the
assertion. (CP 5.546, 1908)
By contrast, a conversation solely determined by the conventions of politeness has

no consequences as to its truth: Nobody takes any positive stock in those conventional utterances, such as I am perfectly delighted to see you, upon whose falsehood no punishment at all is visited (CP 5.546, 1908). Among the speech acts
whose consequences depend on the truth of the proposition expressed are not only
assertions and oaths, but also laying a wager. Peirce describes the consequences of
these speech acts as follows:
What is the difference between making an assertion and laying a wager? Both are acts
whereby the agent deliberately subjects himself to evil consequences if a certain proposition is not true. Only when he offers to bet he hopes the other man will make himself
responsible in the same way for the truth of the contrary proposition; while when he
makes an assertion he always (or almost always) wishes the man to whom he makes it to
be led to do what he does. Accordingly in our vernacular I will bet so and so, is the
phrase expressive of a private opinion which one does not expect others to share, while
You bet is a form of assertion intended to cause another to follow suit. (CP 5.31, 1902)
192
Winfried Nth
Peirce considers not only the conditions of publicly expressed speech acts but also
such mental acts, as a judgment or a doubt. A judgment is a mental act by which
one makes a resolution to adhere to a proposition as true, with all its logical consequences (CP 5.31, 1902). Among the mental acts expressing propositions which
are not asserted is the doubt: I may state it to myself and worry as to whether I
shall embrace it or reject it, being dissatisfied with the idea of doing either
(CP 5.31, 1902). Peirce also specifies several of those conditions of speech acts
which Searle later called preparatory conditions (cf. Brock 1981a: 282). For
example, Peirces formulation of what Searle calls the preparatory condition of a
speech act is: The only way to get a sufficient understanding of a proposition is to
imagine one person communicates in some language to a second. The utterer does
not, properly speaking, communicate it to the interpreter unless the latter is at the
time ignorant (or oblivious) of its truth (MS 284: 41, 1905; Brock 1981a).
The distinction between the illocutionary and the perlocutionary effects of a
speech act has almost no counterpart in Peirces theory of speech acts. A consequence of the pragmatic maxim is that speech acts do not only sometimes have perlocutionary effects; they have always such effects. The study of these effects is the
study of the interpretant, which is a necessary constituent of all signs (see 2.4).
4.2.
Speech acts as narrative scenarios: The Greimasian approach
Elements of the structural semiotics of A. J. Greimas18 are the foundation of a theory of speech acts developed by Sbis and Cooren.19 Sbis (1983: 100101) criticizes Searles speech act theory for being based on a causal theory of action (cf.
Davidson 1980) which results in a too strictly speaker centered approach to the
study of dialogues. According to Sbis, the causal approach to speech acts means a
regrettable return of the subject, the triumph of the empirical speaker and the resuscitation of the use of introspection as a criterion of sense (1983: 100101). To
liberate pragmatics from its mentalist bias, she proposes a semiotic conception of
acts and actions. In accordance with the definition proposed by Greimas and
Courts (1979: 5), Sbis defines an act as a fair-tre, a make-be, i.e., the producing of a state (or of a change of state). She argues that the agent of an act thus
defined is not the cause of the speech act but simply its presupposition. Instead of
the study of speech acts, Sbis proposes the study of interactions between speakers
and hearers. The focus should be on the hearer as an active participant of the dialogue; it should be taken into consideration that the hearer has the choice (i) to select an acceptable interpretation of the speech act, and (ii) to either accept the
speech act, under such an interpretation, as a successful act, or to completely or
partly reject it as more or less inappropriate and unhappy (Sbis and Fabbri
1980: 305).
Instead of investigating the utterers agency in dialogic interaction, Sbis
(1983: 102, 1987: 259) deals with action as an ascribed concept: the interpreter
193
ascribes a certain change-in-the-world to the responsibility of the [utterer as the]

agent. In this perspective, agency shifts from the utterer to the interpreter, whose
decision it is to ascribe intentions and meanings to an utterer or not. With this shift
of perspective, the analysis can take into consideration ascriptions of agency to
quasi-utterers in Peirces sense (see 2.5). According to Sbis (1987: 259), the utterer of a sign, whose agency is now merely an agency ascribed by the hearer, can
thus be a human individual, an infra-individual instance (e.g., the Unconscious), a
super-individual construction (class, party, society, church, and the like), and even
a personified natural agent. Hence, in a semiotic speech act theory adopting these
premises, not only speakers are agents, but even beings other than humans can be
said to do things with words (Cooren 2006: 8). Two common examples of assertive and commissive speech acts in which agency is ascribed to nonhuman agents
are: This decision contests his result and This signature commits you to payment (Cooren 2006: 17).
Based on her assumption that acts result in transformations of states, Sbis
(1985: 529) arrives at a redefinition of basic concepts of speech act theory: (1) A
speech act is the production of an utterance insofar as it brings about a change for
which the utterer is taken to be responsible []. (2) An illocutionary act is the
production of a change of context at the conventional level of the addressers and
the addressees modal competences. (3) A perlocutionary act is the production of
a change of context at the material (psychological or praxeological) level of the addressees reactions. An example of her Greimasian approach to the illocutionary
force of speech acts is Sbiss (1985: 530) study of the discursive figure of manipulation, which she defines as the agency of a manipulating addresser who puts the
manipulated addressee into a position of not being able not to do. The application
of Greimass theory of modalities leads Sbis (1984: 100) to a typology of modal
changes of the speakers and hearers states in speech acts, e.g.: in exercitives (such
as granting or permitting), the speaker addresses the addressee in the deontic mood
of being allowed (can), and the effect in the addressee is one of the modality of
obligation (ought); in commissives, these two modal ascriptions are inverse.
Taking up Sbiss and Greimass elements of a semiotic pragmatics, Cooren
(2000: 81169) proposes a model based on the premise that speech acts involve a
narrative scenario typically involving three actants: the speaker as the agent, the
hearer as the recipient, and the utterance as the object of transfer (Cooren 2000:
86). Uttering is seen as an act of production in which the object produced by the
agent, i.e., the utterance, is irremediably severed or dissociated from its origin
so that the producer no longer has control over the object produced and acquires
an agency of its own (Cooren 2000: 8283). A speech act is always embedded in a
plurality of verbal and nonverbal texts; it does not only refer to, it also produces a
situation. Continuing to act in absence of its utterer, the speech act becomes the
product of multiple texts that speak on behalf of their producer, creating one or
several different situations (Cooren 2000: 87).
194
Winfried Nth
Following Greimass discourse grammar, which describes narrative roles by

means of the actantial model of verb valency of Tesnire (1959), Cooren develops
a typology of illocutionary acts based on verb valencies. According to this model,
assertives are bivalent and performatives are trivalent speech acts (2000: 89). The
latter involve an agent, an object, and a recipient. The tetravalent actantial model
of the factive verbs, which involve four actants, then allows opening up the traditional speaker/hearer schema by acknowledging action from a distance, which
could be called tele-action (Cooren 2006: 2). The utterance John asked Mary to
send his report to Bill is an example of a tetravalent teleactive speech act. To telecommunicate, thus means to delegate a speech act to another agent who communicates it on behalf of the delegating agent: one discursive agent mobilizes other
agents that serve as representative or delegates of the delegating agent (Cooren
2000: 6).
In view of the fundamental differences between Peirces pragmaticist semiotics
and Greimass structural discourse semiotics, it is remarkable that both schools of
semiotics, for very different reasons, agree in their criticism of the speaker-centered approaches to speech acts. It is equally remarkable that, quite independently
of each other, both schools of semiotics postulate models of semiotic agency in
verbal interaction radically distinct from those proposed by the classics of linguistic pragmatics. Nevertheless, the fundamental difference between the two semiotic approaches in this latter respect must not be ignored. Whereas Greimass
model attributes agency to actants in discourse, Peirce attributes agency and autonomy to the sign itself.
Notes
1. Cf. Posner (1991) and Nth (2000: 433438).
2. See Parret (1983), Posner (1997), Nadin (1987, 1993), Deledalle, ed. (1989), Driel, ed.
(1991), Hilpinen (1995, 1998), Pankow (1995), Pietarinen (2004, 2005, 2006), Rellstab
(2008), and Andersen (2009).
3. Cf. Liszka (1996: 315), Pietarinen (2004), and Houser (2009: 91).
4. CP refers to Peirce (193158).
5. Cf. Fisch (1986: 339), Liszka (1996: 109), and Bergman (2007).
6. EP 2 refers to Peirce (1998).
7. Oehler (1990, 1995), Colapietro (1995), Bergman (2000, 2004), Santaella and Nth
(2004), and Pietarinen (2006).
8. See also Prieto (1966, 1975; cf. Nth 2000: 228).
9. See 2.2 and Burkhardt (1990).
10. MS refers to Peirce (1979).
11. Cf. Allwood (1978: 4) and Levinson (1983: 103).
12. For Peirces distinction between presciss (from prescind) and precise see Spinks (1991:
1923) and CP 5.449, 1905. Roughly, prescission is giving attention to one element
and neglect of the other (CP 1.549, 1867).

13.
14.
15.
16.
17.
18.
19.
195
W 1 refers to Peirce (1981).

Cf. Oehler (1995: 268) and Johansen (1993).
Cf. Santaella (1988, 1990, 1994), Joswick (1996), and Pape (1996b).
Brock (1975, 1981a, b), Martens (1981), and Thibaud (1997).
Cf. Brock (1981b) and Hilpinen (1998).
For Greimas, see Nth (2000: 11219).
Sbis and Fabbri (1980), Sbis (1983, 1984, 1985, 1987, 1994, 2002), and Cooren
(2000, 2006).
References
Allwood, Jens
1978
On the Analysis of Communicative Action (Gothenburg Papers in Theoretical
Linguistics 38). Gothenburg: University, Department of Linguistics.
Andersen, Christiane
2009
Semiotics. In: Jef Verschueren and Jan-Ola stman (eds.), Key Notions for
Pragmatics, 217228. Amsterdam: Benjamins.
Austin, John
1962
How to Do Things with Words. London: Oxford University Press.
1954
Bergman, Mats
2000
Reflections on the role of the communicative sign in semeiotic. Transactions of
the Charles S. Peirce Society 36.2: 225254.
Bergman, Mats
2003
Peirces derivations of the interpretant. Semiotica 144: 117.
Bergman, Mats
2004
Fields of Signification. Explorations in Charles S. Peirces Theory of Signs.
(Philosophical Studies from the University of Helsinki 6), Helsinki, PhD dissertation.
Bergman, Mats
2007
The secret of rendering signs effective: the import of C. S. Peirces semiotic
rhetoric. The Public Journal of Semiotics 1.2: 211.
Brock, Jarrett
1975
Peirces conception of semiotics. Semiotica 14: 124141.
Brock, Jarrett
1979
Principal themes in Peirces logic of vagueness. Peirce Studies (Lubbock,
Texas) 1: 4150.
Brock, Jarrett
1981a
An introduction to Peirces theory of speech acts. Transactions of the Charles
S. Peirce Society 17: 319326.
Brock, Jarrett
1981b
Peirce and Searle on assertion. In: Kenneth Ketner, Joseph M. Ransdell, Carolyn Eisele, Max H. Fisch and Charles S. Hardwick (eds.), Proceedings of the
C.S. Peirce Bicentennial International Congress. 281287. Lubbock, TX:
Texas Tech Press (Graduate Studies Texas Tech University 23).
196
Winfried Nth
Bhler, Karl
1933
Die Axiomatik der Sprachwissenschaft. Frankfurt am Main: Klostermann.
Burkhardt, Armin, (ed.)
1990
Speech Acts, Meaning and Intentions. Berlin: Walter de Gruyter.
Buyssens, Eric
1943
La communication et larticulation linguistique. Bruxelles / Paris: Presses
Universitaires (revised edition 1964).
Carnap, Rudolf
1975
Introduction to Semantics and Formalization of Logic, 2 vols. in 1. Cambridge:
Harvard University Press [1942/43].
Colapietro, Vincent
1992
Purpose, power, and agency. The Monist 75: 423444.
Colapietro, Vincent
1995
Immediacy, opposition, and mediation: Peirce on irreducible aspects of
the communicative process. In: Leonore Langsdorf and Andrew R. Smith
(eds.), Recovering Pragmatisms Voice: The Classical Tradition, Rorty, and
the Philosophy of Communication, 2348. Albany, NY: State University
Press.
Cooren, Franois
2000
The Organizing Property of Communication. Amsterdam: Benjamins.
Cooren, Franois
2006
Between semiotics and pragmatics: Opening language studies to textual
agency. Paper presented at the annual meeting of the International Communication Association, Dresden. <http://www.allacademic.com/meta/p87081_
index.html, access 05-25-2009>.
Davidson, Donald
1980
Essays on Actions and Events. Oxford: Oxford University Press.
Deacon, Terrence
1997
The Symbolic Species. New York, NY: Norton.
Deledalle, Grard (ed.)
1989
Derrida, Jacques
1988
Limited Inc. Evanston, IL: Northwestern University Press.
Driel, Herbert van (ed.)
1991
Het semiotisch pragmatisme van Charles S. Peirce. Amsterdam: Benjamins.
Eisele, Carolyn
1987
Peirces pragmatics. In: Herbert Stachowiak (ed.), Der Aufstieg des pragmatischen Denkens im 19. und 20. Jahrhundert (Pragmatik: Handbuch des pragmatischen Denkens, vol. 2.), 8397 Hamburg: Meiner.
Fisch, Max H.
1986
Peirce, Semeiotic, and Pragmatism. Ed. Kenneth L. Ketner and Christian
J. W. Kloesel. Bloomington: Indiana University Press.
Gorl, Dinda
1993
Evolving through time: Peirces pragmatic maxims. Semiosis 71/72: 313.
Goudge, Thomas A.
1965
Peirces index. Transactions of the Charles S. Peirce Society 1: 5270.
197
Greimas, Algirdas J. and Joseph Courts

1979
Smiotique: Dictionnaire raisonn de la thorie du langage. Paris: Hachette.
Trad. 1982. Semiotics and Language: An Analytical Dictionary. Bloomington,
IN: Indiana University Press.
Grice, Paul
1957
Meaning. Philosophical Review 64: 377388.
Grice, Paul
1989
Habermas, Jrgen
1981
Theorie des kommunikativen Handelns, 2 vols. Frankfurt am Main: Suhrkamp.
Habermas, Jrgen
1995
Peirce and communication. In: Kenneth Laine Ketner (ed.), Peirce and Contemporary Thought. Philosophical Inquiries, 243266. New York: Fordham
University Press.
Hervey, Sndor G. J.
1982
Semiotic Perspectives. London: Allen and Unwin.
Hilpinen, Risto
1995
Peirce on language and reference. In: Kenneth Laine Ketner (ed.), Peirce and
Contemporary Thought. Philosophical Inquiries, 272303. New York: Fordham University Press.
Hilpinen, Risto
1998
Pragmatics and pragmatism: On C.S. Peirces pragmatic theory of meaning. In:
Paul Weingartner, Gerhard Schurz, and Georg Dorn (eds.), Die Rolle der Pragmatik in der Gegenwartsphilosophie The Role of Pragmatism in Philosophy,
151164 (=Akten des 20. Internationalen Wittgenstein-Symposiums. Schriftenreihe der Wittgenstein-Gesellschaft, vol. 26) Wien: Hlder-Pichler-Tempsky.
Houser, Nathan
1998
Introduction. In: Nathan Houser and Christian J. W. Kloesel (eds.), The Essential Peirce. Vol. 1, xvii-xxxviii. Bloomington: Indiana University Press.
Houser, Nathan
2009
Peirce, phenomenology and semiotics. In: Paul Cobley (ed.), The Routledge
Companion to Semiotics, 89100. London: Routledge.
Jakobson, Roman
1960
Linguistics and poetics. In: Thomas A. Sebeok (ed.), Style in Language,
1993
Dialogic Semiosis. Bloomington: Indiana University Press.
2002
Literary Discourse: A Semiotic-Pragmatic Approach to Literature. Toronto:
Joswick, Hugh
1996
The object of semeiotic. In: Vincent M. Colapietro and Thomas M. Olshewsky
(eds.), Peirces Doctrine of Signs, 93102. Berlin: de Gruyter.
Kappner, Stefan
2004
Intentionalitt aus semiotischer Sicht: Peirceanische Perspektiven. Berlin: de
Gruyter.
198
Winfried Nth

1980
Metaphors We Live By. Chicago: University Press.
Leech, Geoffrey N.
1983
1983
Liszka, J. Jakb
1996
A General Introduction to the Semeiotic of Charles S. Peirce. Bloomington: Indiana University Press.
Liszka, J. Jakb
2009
Rethinking the pragmatic theory of meaning. Cognitio (So Paulo) 10.1:
6180.
Luhmann, Niklas
1984
Soziale Systeme. Frankfurt am Main: Suhrkamp.
Martens, Ekkehard
1981
C. S. Peirce on speech acts. In: Kenneth Ketner, Joseph M. Ransdell, Carolyn
Eisele, Max H. Fisch and Charles S. Hardwick (eds.), Proceedings of the
C.S. Peirce Bicentennial International Congress, 289292. Lubbock, TX:
Texas Tech Press (=Graduate Studies Texas Tech University 23).
Montague, Richard
1970
Pragmatics and intensional logic. Synthese 22: 6894.
Morris, Charles W.
1938
Foundations of the Theory of Signs (Foundations of the Unity of Science:
Towards an International Encyclopedia of Unified Science, vol. 1.2.) Chicago,
IL: University Press [1970].
Morris, Charles W.
1946
Signs, Language, and Behavior. New York: Braziller. (Quoted from the reprint
in C. W. Morris. 1971. 73398.)
Morris, Charles W.
1964
Signification and Significance: A Study of the Relations of Signs and Values.
Morris, Charles W.
1971
Murphy, John P.
1990
Pragmatism from Peirce to Davidson. Boulder, CO: Westview Press.
Nadin, Mihai
1987
Pragmatics in the semiotic frame. In: Herbert Stachowiak (ed.), Der Aufstieg
des pragmatischen Denkens im 19. und 20. Jahrhundert, 148170. (Pragmatik: Handbuch des pragmatischen Denkens, vol. 2.) Hamburg: Meiner.
Nadin, Mihai
1993
Semiotics in action: The pragmatic level. In: Herbert Stachowiak (ed.), Spachphilosophie, Sprachpragmatik und formative Pragmatik, 219250. (Pragmatik: Handbuch des pragmatischen Denkens, vol. 4.) Hamburg: Meiner.
Nth, Winfried
2000
Handbuch der Semiotik, 2nd rev. ed. Stuttgart: Metzler.
Nth, Winfried
2002a
Charles Sanders Peirce, pathfinder in linguistics. Interdisciplinary Journal for
Germanic Linguistics and Semiotic Analysis 7.1: 114.
199
Nth, Winfried
2002b
Wrter als Zeichen: Einige semiotische Aspekte der Sprache. In: Jrgen
Dittmann and Claudia Schmidt (eds.), ber Wrter, 932. Freiburg: Rombach.
Nth, Winfried
2006
Representation and reference according to Peirce. Caderno. 9a Jornada do
Centro de Estudos Peirceanos / IIIrd Advanced Seminar on Peirces Philosophy
and Semiotics, 7183. (So Paulo, PUC-SP), So Paulo.
Nth, Winfried
2008
Natural signs and the natural ingredients of cultural signs. In: Martin Dring,
Hermine Penz and Wilhelm Trampe (eds.), Language, Signs and Nature: Ecolinguistic Dimensions of Environmental Discourse, 2137. Tbingen: Stauffenburg.
Nth, Winfried
2009
On the instrumentality and semiotic agency of signs, tools, and intelligent machines. Cybernetics and Human Knowing 16.12: 1136.
Nth, Winfried and Lucia Santaella
2009
Die Relevanz der Peirceschen Semiotik des Vagen fr die Theorie der Kommunikation. In: Siegfried J. Schmidt (ed.), Die (Un)mglichkeit der Kommunikation. Frankfurt am Main: Peter Lang (Special Issue SPIEL: Siegener Periodicum zur Internationalen Empirischen Literaturwissenschaft 26.1 [2007],
7396).
Oehler, Klaus
1990
Peirce oder Habermas? In: Elisabeth Walther and Udo Bayer (eds.), zeichen
von zeichen fr zeichen: festschrift fr max bense, 4247. Baden-Baden: Agis.
Oehler, Klaus
1995
A reponse to Habermas. In: Kenneth Laine Ketner (ed.), Peirce and Contemporary Thought, 267271. New York: Fordham University Press.
Pankow, Christiane
1995
Semiotics. In: Jef Verschueren, Jan-Ola stman and Jan Blommaert (eds.),
Handbook of Pragmatics Manual, 469476. Amsterdam: Benjamins.
Pape, Helmut
1996a
Charles Sanders Peirce. In: Tilman Borsche (ed.), Klassiker der Sprachphilosophie, 307324. Mnchen: Beck.
Pape, Helmut
1996b
Object and final cause in Peirces semeiotic. In: Vincent M. Colapietro and
Thomas M. Olshewsky (eds.), Peirces Doctrine of Signs, 103117. Berlin: de
Gruyter.
Parmentier, Richard J.
1985
Signs place in medias res: Peirces concept of semiotic mediation. In: Elizabeth Mertz and Richard J. Parmentier (eds.), Semiotic Mediation, 2348. Orlando, FL: Academic Press.
Parret, Herman
1983
Parret, Herman (ed.),
1994
Pretending to Communicate. Berlin: de Gruyter.
200
Winfried Nth

193158 Collected Papers, Vols. 16. ed. Charles Hartshorne and Paul Weiss, Vols. 78,
ed. Arthur W. Burks. Cambridge, MA: Harvard University Press (quoted as
CP).
1979
The Charles S. Peirce Papers, 30 reels, 3rd microfilm edition. Cambridge, MA:
The Houghton Library, Harvard University, Microreproduction Service
(quoted as MS).
1981
Writings of Charles S. Peirce. Vol. 1: 18571866. Bloomington, IN: Indiana
Univ. Press (quoted as W 1).
1998
The Essential Peirce. Selected Philosophical Writings. Vol. 2, 18931913).
Ed. by the Peirce Edition Project. Bloomington, IN: Indiana University Press
(quoted as EP 2).
2004
Grice in the wake of Peirce. Pragmatics and Cognition 12.2: 295315.
2005
Relevance theory through pragmatic theories of meaning. In: Bruno G. Bara,
Lawrence Barsalou and Monica Bucciarelli (eds.), Proceedings of the XXVII
Annual Conference of the Cognitive Science Society, 17671772. Mahwah, NJ:
Lawrence Erlbaum,
2006
Signs of Logic. Peircean Themes on the Philosophy of Language, Games, and
Communication. Dordrecht: Springer.
Plato
1953
The Dialogues of Plato, transl. Benjamin Jowett, 4th ed. (1st ed. 1871). Oxford:
Clarendon. Also: http://www.gutenberg.org/dirs/1/6/1/1616/1616.txt.
Posner, Roland
1985
La syntactique, sa relation la smantique et la pragmatique, la morphologie
et la syntaxe, et la syntagmatique et la paradigmatique. In: Herman Parret and
Hans-Georg Ruprecht (eds.), Exigences et perspectives de la smiotique. Recueil dhommages pour A.J. Greimas, vol. 1, 7396. Amsterdam: Benjamins.
Posner, Roland
1991
Research in pragmatics after Morris. Dedalus (Lisbon) 1: 115156.
Prieto, Luis J.
1966
Messages et signaux. Paris: Presses Universitaires.
Prieto, Luis J.
1975
Etudes de linguistique et de smiologie gnrales. Genve: Droz.
Ransdell, Joseph
1977
Some leading ideas of Peirces semiotic. Semiotica 19: 157178.
Rellstab, Daniel
2008
Peirce for linguistic pragmatics. Transactions of the Charles Sanders Peirce
Society 44.2: 312346.
Santaella, Lucia
1988
Charles Sanders Peirces object (of the sign). Versus 49: 5358.
201
Santaella, Lucia
1990
An introduction to Peirces concept of object. Cruzeiro semitico 13: 7582.
Santaella, Lucia
1994
Why things are objects of the sign. Cruzeiro semitico 18: 3746.
Santaella, Lucia
1999
Methodeutics, the liveliest branch of semiotics. Semiotica 124: 377395.
Santaella, Lucia and Winfried Nth
2004
Comunicao e semitica. So Paulo: Hacker.
Savan, David
1994
C. S. Peirce and American semiotics. In: Michael Shapiro and Michael
C. Haley (eds.), The Peirce Seminar Papers 2, 179208. Providence, RI: Berghahn.
Sbis, Marina
1983
Actes de langage et (acte d)nonciation. Langages 18.70: 99106.
Sbis, Marina
1984
On illocutionary types. Journal of Pragmatics 8: 93112.
Sbis, Marina
1985
Manipulation et sanction dans la dynamique des actes de langage. In: Herman
Parret and Hans-Georg Ruprecht (eds.), Exigences et perspectives de la smiotique. Recueil dhommages pour A.J. Greimas, Vol. 1, 529538. Amsterdam:
Benjamins.
Sbis, Marina
1987
Speech acts and context change. In: Thomas T. Ballmer and Wolfgang Wildgen
(eds.), Process Linguistics, 254280. Tbingen: Niemeyer.
Sbis, Marina
1994
Language and dialogue in the framework of the analytical philosophy of ordinary language. In: Edda Weigand (ed.), Concepts of Dialogue, 159170. Tbingen: Niemeyer.
Sbis, Marina
2002
Cognition and narrativity in speech act sequences. In: Anita Fetzer and Christiane Meierkord (eds.), Rethinking Sequentiality, 7197. Amsterdam: Benjamins.
Sbis, Marina and Paolo Fabbri
1980
Models ( ? ) for a pragmatic analysis. Journal of Pragmatics 4: 301319.
Searle, John R.
1969
Speech Acts: An Essay in the Philosophy of Language. Cambridge: University
Press.
Searle, John R.
1983
Intentionality. Cambridge: University Press.
1949
Press.
Short, Thomas L.
2007
Peirces Theory of Signs. Cambridge: University Press.
1986
202
Winfried Nth
Spinks, Cary W.
1991
Peirce and Triadomania. Berlin: de Gruyter.
Stetter, Christian
1989
Some remarks concerning the theory of speech acts and the idea of pragmatics. In: Gerard Deledalle (ed.), Semiotics and Pragmatics, 155165.
Strawson, Peter
1971
Intention and convention in speech acts. In: John R. Searle (ed.), The Philosophy of Language, 2338. Oxford: University Press.
Tesnire, Lucien
1959
Elments de syntaxe structurale. Paris: Klincksieck.
Thibaud, Pierre
1997
Between saying and doing: Peirces propositional space. Transactions of the
Charles Sanders Peirce Society 33: 270327.
Ventola, Eija
1987
The Structure of Social Interaction: A Systemic Approach to the Semiotics of
Service Encounters. London: Frances Pinter.
Watzlawick, Paul, Janet H. Beavin, and Don D. Jackson
1967
Pragmatics of Human Communication. New York: Norton.
1953
Philosophische Untersuchungen Philosophical Investigations. Oxford:
Blackwell.
7.

Introduction
According to what might be called the traditional paradigm in the philosophy of
language, linguistic interpretation is a linear and serial process. In particular, pragmatic interpretation (such as the working out of Gricean implicatures) is claimed
to commence only after the semantic interpretation of a given sentence has been
worked out. This traditional paradigm has come under attack from various directions. Some philosophers contend, for instance, that there is no such thing as a
purely semantic content or a semantically expressed proposition (or truth-condition) but that pragmatic inferences are necessary at every level of linguistic interpretation: this claim will be discussed in the first part (chapter 1) of this article. One
may even go one step further and maintain that semantic notions need to be spelled
out pragmatically in the first place, that linguistic interpretation requires comprehension of a particular type of action and that semantics is thus simply a part of
pragmatics: this will be explored in the second part (chapter 2) of our article.
1.
Pragmatic inferences and semantic rules
1.1.
Two approaches to language and meaning
In 1938, Charles Morris distinguished three different dimensions of semiosis in

his theory of signs:1
Syntax studies the relations of signs to other signs, semantics is concerned with the
relation between signs and objects, and pragmatics takes the relation between signs
and their interpreters as its subject matter. To partition the domain this way has
been common ever since. Morriss characterization, however, leaves the three
areas severely underdetermined. For example, pragmatics is commonly supposed
204
to cover such diverse phenomena as speech acts, deixis, presuppositions or implicatures, their unifying feature being only that they have something to do with the
relation between signs and interpreters (and that they cannot easily be accommodated within a semantic theory). Also, the characterization may be not as clear-cut
as Morris and others hoped it to be. In particular (and as a consequence), where
exactly to draw the line between semantics and pragmatics is a hotly contested
issue in current philosophical and linguistic debate.2 Moreover, the difficulties one
encounters when trying to sharpen the distinction between semantics and pragmatics point to a deeper problem about how to theoretically approach language in
general. It is common in the philosophy of language to contrast two different approaches to questions concerning language, meaning and interpretation. Robert
Brandom, for example, puts the difference thus: some philosophers view language
as a kind of practice or activity, others prefer a semantic Tarskian approach to language (cf. Brandom 2002: 4041). The former focus on how language is used; they
study language as a kind of practice, as something speakers do in order to achieve
certain aims. The latter, on the other hand, strive to develop a theory of content that
is closely tied to Alfred Tarskis truth-theoretical considerations. Underlying the
two different approaches to language are different views of what a language is,
which David Lewis aptly characterizes thus:
What is a language? Something which assigns meanings to certain strings of types of
sounds or of marks. It could therefore be a function, a set of ordered pairs of strings and
meanings. [] (Lewis 1975/1983: 163)
Those who favor this view of what a language is think of language as an abstract
system that systematically pairs certain strings of sounds or marks with certain
other entities, called meanings. The mapping can be done, it seems, without taking
language users into account. In the contrasting view, on the other hand, the language user takes center stage:
What is a language? A social phenomenon which is part of the natural history of human
beings; a sphere of human action, wherein people utter strings of vocal sounds, or inscribe strings of marks, and wherein people respond by thought or action to the sounds
or marks which they observe to have been so produced. (Lewis 1975/1983: 164)
Which of these two views of language will prove theoretically fruitful is the point
at issue between semantically and pragmatically oriented accounts of language and
linguistic interpretation. Very roughly, the pragmaticist will advocate a usagebased approach to language. S/he will, accordingly, think of language as a human
activity, a social phenomenon that helps us interact and communicate with our fellow human beings. His/her opponent, the semanticist, takes a language to be a
formal system that can be studied fairly independently of how human beings actually use language. His/her aim is to come up with a formal semantic theory that
allows us to assign every sentence of language L its meaning in L. More specifically, s/he will try to pair syntactically analyzed sentences (of a language) with
205
their truth-conditions, as s/he typically takes the meaning of a sentence to be its

truth-condition (cf., e.g., Portner and Partee 2002: 3).
Semantic interpretation was commonly taken to be independent of pragmatics.
Some even claimed that semantics comes first (cf. Manor 2001: 67f). Pragmatic interpretation (such as the derivation of Gricean implicatures) comes in only after the
semantic interpretation of the expressions in question has been worked out. But
new work on the context sensitivity of natural languages questions not only the semantics first-view but the idea of semantic independence as such. More specifically, in recent years the project of a formal semantic theory has come under attack
from what came to be called contextualist approaches to language and meaning.
Very roughly, the contextualist can be characterized as someone who takes context sensitivity to be a ubiquitous feature of natural languages, a feature much more
pervasive than deixis or anaphora and much less well-behaved than implicatures,
for example. S/he favors a pragmatically oriented, usage-based approach to meaning, arguing that linguistic (even truth-conditional) interpretation involves pragmatic inferences all the way down. Her/his opponent, on the other hand, advocates
a formal semantic approach to meaning according to which truth-conditional interpretation is a purely syntactic-semantic endeavor. S/he would, presumably, agree
with the following characterization:
[] in general, formal theories can be characterized as fundamentally syntax-driven
theories, which claim that it is possible to deliver an account of the propositional or
truth-conditional content of a sentence in natural language simply via formal operations
over the syntactic features of that sentence, that is, over the lexical items it contains and
their mode of composition. (Borg 2004b: 3)
The pragmatically oriented contextualist and his/her semantically oriented opponent negotiate what was traditionally taken to be key semantic notions: the notion
of what is said in the utterance of a sentence, the notion of the proposition expressed by the utterance, the notion of a sentences truth-conditions, and at the
most basic level the notion of meaning itself. Semanticists try to save one or another of these notions from pragmatic intrusion.
1.2.
Formal semantics
Let us first try to get a better grasp of what form a formal semantic theory might
take before we consider, in a second step, the objections the contextualist raises
against this kind of approach. Very roughly, formal semantic theories try to develop, for as large a fragment of natural language as possible, a formal representation that contains all syntactically and semantically relevant information. In a
first step, the logical form of each sentence in the relevant fragment of the language
in question has to be worked out. In a second step, the elements of logical form
have to be interpreted, i.e., each element has to be assigned a semantic value in ac-
206
cordance with the respective interpretation rules. More specifically, the idea is that
there are interpretation rules that assign semantic values to the basic expressions in
the language. Composition rules then tell us which value to assign to complex expressions, the semantic value of a complex expression being a function of the semantic values of its parts. In this manner we work our way up until, after finitely
many steps, we are able to assign semantic values (i.e., truth-conditions) to whole
sentences. Compositionality obviously plays an important role here, for if language were not compositional, no recursive procedure for assigning semantic
values to expressions of arbitrary complexity would be available. Yet such a recursive procedure is exactly what the semanticist is looking for: a procedure that
yields for every well-formed sentence (of finite length) its meaning, i.e., the proposition expressed or its truth-conditions, after finitely many steps. And to interpret
an expression is to follow such a recursive procedure. Here is a somewhat lengthy
quote from Jason Stanley, a (moderate) proponent of formal semantics, to illustrate
the point:
Successful interpretation involves assigning denotations to the constituents of the logical form, and combining them in accord with composition rules that do not vary with
extra-linguistic context. The denotations that successful interpreters will assign to constituents of a logical form will be constrained by the linguistic conventions governing
those elements. [] What results from a successful application of this first stage of
interpretation is a unique proposition, a fully truth-evaluable entity. (Stanley 2002:
149150)
One of the attractive features of a formal semantic approach is that it seems to provide a good explanation of the productivity of language. Obviously, competent
speakers are able to produce novel sentences and to understand sentences they
have not encountered before. This fact stands in need of explanation. The formal
semanticist seems able to provide an explanation by pointing out the recursive
character of linguistic interpretation.
Yet what semantic value do interpretation rules assign to the basic expressions
of a language? What is the semantic value, i.e. the meaning, of a linguistic expression? It is the answer to this question on which philosophical theories of meaning differ. According to the reference theory of meaning it is the expressions reference, often called its extension: what the expression is used to talk about or refer
to. Unfortunately, the reference theory of meaning faces a couple of serious and
well-known objections, some of which already attracted Gottlob Freges attention
and brought him to distinguish between an expressions reference (what he called
Bedeutung) and its sense (Sinn, cf. Frege 1891, 1892a and 1892b). Freges insights, in turn, loomed large in the development of so called intensional semantics.
Rudolf Carnap, who attended some of Freges lectures in Jena, strived to find a
suitable method for the semantic analysis of meaning (Carnap 21956: 2) and proposed the concept of extensions and intentions of an expression as a suitable tool.
This distinction between an expressions extension and intension soon caught on.
207
Very roughly, the idea is this: It may not be adequate to equate an expressions
meaning with its actual extension because the latter may depend on the way the
world is, which may be subject to change. Yet isnt the task of an expressions
meaning to pursue this kind of dependence? In other words, isnt it the case that
meaning determines the expressions extension in any possible world? Such functions from possible worlds to extensions are called intensions. The intension of an
expression is supposed to determine its extension (i.e. its truth-value, in the case of
whole sentences) in any possible world. In other words, intensions take possible
worlds as arguments and provide extensions as truth-values. Hence, intensions
seem to be better candidates than extensions for what we intuitively think of as an
expressions meaning.
One of the most influential later developments in intensional semantics is Montague Grammar. It was originally developed by Robert Montague who elaborated
on ideas introduced by Rudolf Carnap, Alfred Tarski and Alonzo Curch and subsequently modified and expanded by other logicians, philosophers and linguists
(cf. Montague 1970a, 1970b, 1973, and also Partee 1973, 1975; Porter and Partee
2002).
In the course of these considerations it became more and more obvious that the
extension of an expression not only depends on the way the world is but also, in
some cases at least, on quite specific features of the context in which the expression
is used. The most prominent examples are indexical expressions such as I,
here or now. What these expressions refer to depends on the context of utterance; similarly for demonstrative pronouns such as this or that. Insight into
this double-dependence of extension on context and world respectively has led to
the development of what came to be called two-dimensional semantics. According
to a renowned view held by David Kaplan (1989), for example, the overall import
of a linguistic expression breaks down into two components: its character and its
content. Both can be construed as functions, the character of an expression being a
function from the context of utterance into the expressions content, the content
being a function from possible circumstances of evaluation (very roughly, something like a possible world or an index, terminology differs here) into the extension
of the expression in the respective circumstances.3 This opens up a new way of accommodating context sensitive expressions within a formal semantic theory: just
associate with each context sensitive expression a function or rule (a character)
that determines, for any possible context of its use, the expressions content (its extension, in the case of indexical or demonstrative expressions) in that context.
Thought of in this way, linguistic (truth-conditional) interpretation is (context-sensitivity notwithstanding) still a purely syntactic-semantic enterprise. No pragmatic
processes seem to be required; a result that nicely fits in with the formal semantic
project as sketched above.
208
1.3.
Context-sensitivity and pragmatic inferences
Still, as mentioned above, the formal semantic project came under attack when it
became obvious that there is much more context-sensitivity in natural language
than the formal semanticist is willing to acknowledge. Moreover, it is far from
clear that she could accommodate all the context-sensitivity within her formal
semantic approach even if she wanted to. The main problem is that pragmatic inferences seem to be required after all in order to truth-conditionally interpret utterances containing context-sensitive expressions and to assign these expressions a
semantic value in context. Yet the nature of pragmatic inferences does not square
well with the allegedly deductive, algorithmic nature of semantic interpretation
(cf. Borg 2004a: 218219).
1.3.1.
The non-monotonic nature of pragmatic inferences
The need for pragmatic inferences in linguistic interpretation has long been acknowledged in philosophical and linguistic theorizing. Ever since Paul Grice developed his theory of conventional and conversational implicatures the need for
pragmatic inferences had become obvious. According to Grice (1989), what a
speaker means by an utterance divides up into what she says and what she implicates. What she says is (subject to certain qualifications) closely tied to the conventional meaning of the words used. The implicatures (at least the conversational ones), on the other hand, have to be pragmatically worked out. Usually, in
generating an implicature, a speaker exploits certain maxims her hearers expect
her to comply with, maxims to the effect that she should try to be relevant, succinct, truthful and informative. By violating or flouting one of these maxims, a
speaker can communicate something other or even more than what she explicitly
said.
To use one of Grices own examples, if you ask me whether our common friend
Smith has a girlfriend at the moment, and I reply by saying that he has been paying
a lot of visits to Zurich lately, you might reason as follows. What she said was not
directly relevant for my question. But then, I have no reasons to assume that she is
not cooperative. So she must have meant to communicate something other than
what she said. What could that be? Maybe she meant to suggest that Smith has a
girlfriend in Zurich. The conclusion is defeasible, though. It might need revision
in the light of new information. And this line of reasoning is an example of a pragmatic inference. Since Grice, it has been common to think of pragmatic inferences
as being restricted to cases such as these. More specifically, pragmatic inferences
are commonly taken to come into play in linguistic interpretation only after the
semantic, truth-conditional content of the sentences has been worked out. Yet that
is exactly what the contextualist denies. She claims that pragmatic inferences are
required all the way down. And the problem is that the defeasible nature of prag-
209
matic inferences clashes with the deductive nature of semantic interpretation. This
is problematic for various reasons:
Firstly, because in these cases it is not enough to just follow any allegedly semantic rule or algorithm, because there is no semantic rule or algorithm to follow.
(Though that is not to say that there might not be any syntactic rule, i.e. a rule of
grammar).
Secondly, a different conception of context is called for than the one invoked in
syntactic-semantic interpretation if our aim is to accommodate all different forms
of context-sensitivity. According to a very common conception of context (e.g.
Lewis 1980; Kaplan 1989), an utterance context is taken to comprise a neat list of
contextual features such as a world, a speaker, a time and a place of utterance, and
thus to determine at least the values of indexical expressions.4 Yet such a conception of context will not do for present purposes. For one thing, because the idea of
context determining anything stands in dire need of explanation. More importantly,
what is needed is a conception of context according to which a context is seen as
something that provides information: context has to provide all the information
that speaker and interpreters have to draw on in order for communication and interpretation to be successful. That seems to be the idea behind Kent Bachs notion
of wide context:
Wide context does not literally determine anything. It is the body of mutually evident information that the speaker exploits to make his communicative intention evident and
that his audience relies upon, taking him to intend them to do so, to identify that intention. (Bach and Bezuidenhout 2002: 285)
Bach emphasizes that wide context only enables the hearer to figure out the semantic value of a given expression but does not determine it. Content is ascertained
on the basis of information provided by wide context. This epistemic notion of
context as providing speaker and hearer with information about how to interpret a
given utterance seems to be better suited for modelling all different forms of context-sensitivity.5
Thirdly, and most importantly, pragmatic inferences often take the form of an
inference to the best explanation or rather to the best interpretation.6 As a result,
they are liable to be defeated by new information. And there is no limit to the
amount of information that may turn out to be relevant (cf. Borg 2007: 356 and
Borg 2004b: 79; Recanati 2004: 54). More specifically, pragmatic inferences are
defeasible, non-monotonic, in the following sense: It may be the case that although
conclusion C follows (not in the deductive sense, of course) from a set P of premises, it no longer follows from a set M, such that M P, i.e. such that P is a proper
subset of M. In other words, adding new premises might turn a good inference
into a bad one. (Defeasible reasoning is, of course, not confined to cases of linguistic interpretation. There is, by now, a huge literature on this mode of reasoning;
see for an overview Koons 2009). But if that is so, then there may not be an algo-
210
rithmic procedure for pragmatic interpretation as there is no guarantee that the task
can be completed after finitely many steps. Inferences concerning speaker intentions are examples of pragmatic inferences. Here is an example of Borgs to illustrate the defeasible nature of intentional interpretation:
So, imagine that you see Sally filling a glass of water from the tap. Then you might reason as follows: Sally is getting a glass of water form the tap. The best explanation for
this action is that Sally is thirsty and wants a drink; therefore Sally is thirsty and wants a
drink. Clearly this is a non-demonstrative piece of reasoning and it is susceptible to the
influence of an open-ended range of contextual factors. For instance, say that you know
that Sally has just come in with Sourav and that Sourav is wearing running gear and
looks out of breath, then the best explanation for Sallys action might be that Sourav is
thirsty and Sally is getting a drink for him. Or imagine that Sally has just glanced at her
potted plant, then the best explanation might be that she wants to water her plant. (Borg
2004b: 7879)
Interpreting a persons actions in the way described clearly requires pragmatic inferences. And the interpretation of linguistic utterances is similar in that respect to
the interpretation of actions. For many linguistic expressions are such that in order
to figure out what has been said, speaker intentions at least to the extent that
speakers manage to make them manifest need to be taken into account. Actually,
even indexical and demonstrative expressions seem to require pragmatic inferences for their interpretation; what place or time exactly a speaker refers to with an
indexical expression seems to depend, partially at least, on which region in space
or what stretch of time he intended to refer to. (But let us put these problems aside
for the time being.) Yet intentional interpretation is not the only case where pragmatic inferences come into play. In interpreting utterances, hearers not only draw
on contextually provided information about (manifest) speaker intentions; other
contextual information has to be taken into account, too.
1.3.2.
Context-sensitivity some examples
For the purpose of illustration, let us consider a few more examples of context-sensitivity in natural language. Suppose a speaker says
The suitcase is too heavy.
at one time while trying to put a suitcase into the trunk of a car, at another time
while packing his hand luggage for a flight to Paris (for a similar example, cf. Bach
1994: 268). Or suppose a speaker says:
Everyone came to the party. (Stanley 2005: 223224)
at one time while discussing whether all the important people in his department
came to the party, at another time when commenting on how crowded the party
was.
As opposed to sentences such as triangles have three sides or bachelors are
unmarried, our example sentences are context-sensitive in that they can be used to
211
say different things relative to different contexts of utterance; and their truth-conditions differ accordingly. Here is another example.
Jill cant continue. (Borg 2004b: 228)
This sentence can be used to say that Jill cannot continue school, or that she cannot
continue dance classes, or university education, and so on. What it is used to say on
a particular occasion of utterance usually depends on context and can, accordingly,
be figured out by taking context into account. Moreover, whether what is thereby
said is true or false will depend on context, too. Similarly if a speaker says
Tom is tall.
The speaker might have used the sentence to say that Tom is tall for a fifth grader or
that he is tall for an NBA player, etc. Moreover, his utterance may well be true in
the first case but false in the second.7
Accordingly, the contextualist will claim that in order to figure out what has
been said (not what has been implicated in the Gricean sense!) on a particular occasion of their use, most natural language sentences require pragmatic inferences
on the part of the interpreter. Moreover, the contextual clues the interpreter has to
draw on include not only information about the speakers intentions but also information about the purpose or point of the conversation, the participants shared
background assumptions and interests, etc.
Even more interesting from a contextualist point of view are still subtler forms
of context sensitivity. Many natural-language expressions can be used to depict
slightly different actions or entities in different contexts of their use (owing to the
fact that language is used for classification, and classification itself is context-sensitive). Accordingly, they can be interpreted differently, again depending on the
purpose or point of the conversation, the participants shared background assumptions, etc. That is just to say that there is ample polysemy in natural language,
where [p]olysemy is understood here in a broad sense as variation in the construal
of a word on different occasions of use (Croft and Cruse 2004: 109). Compare the
following utterances:
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
Jane opened the window

Bill opened his mouth.
Sally opened her book to page 56.
Mike opened his briefcase.
Pat opened the curtains.
The child opened the package.
The carpenter opened the wall.
The surgeon opened the wound. (Carston 2002: 325)
To open a wound is not the same thing as to open a book. And a wall is usually
opened not in the way one opens ones mouth. The word open can be used to depict a variety of different actions; and an interpreter has to contextually work out,
i.e., pragmatically infer the appropriate understanding. There may be default inter-
212
pretations, though, depending on the object of the verb. But these preferred interpretations are not inevitable. Also, one might think that in all these cases there is
something like a semantic core meaning. In the above example, manipulate such
that the inside/interior/thing underneath can be seen might be a promising candidate. But then the question arises of whether open is supposed to be synonymous
with manipulate such that the inside/interior/thing underneath can be seen. If it is
not supposed to be synonymous, the meaning of open differs from its core meaning, a puzzling result. Yet if it is supposed to be synonymous then one might
wonder what the point of exact synonyms could possibly be in natural language?
(And let us put aside the problem that giving a synonym is not the same as giving
the meaning of an expression, for one might know that a and b are synonymous, without knowing the meaning of either a or b.) Moreover, new work on
language acquisition seems to show that children, when learning a language, operate on the assumption that every difference in form (of linguistic expressions)
marks a difference in meaning (cf., e.g., Clark 1993: 64; Tomasello 2003: 7275ff;
Bloom 2000: 65). Otherwise language acquisition would be a mystery. So the idea
of exact synonyms and, consequently, of core meanings in the case of polysemy is
rather problematic.
Whether a sentence containing a polysemic expression is correctly applicable
to a particular situation depends on the participants interests and purposes, their
background assumptions and interests, etc. Given such and such purposes, interests, etc, it may be correctly applied, given different purposes, interests, etc., it may
be not correctly applied. Yet couldnt we specify further and try to make some of
the purposes, interests or common assumptions explicit? The problem is, as Anne
Bezuidenhout has pointed out, that
[t]here is no sentence that we can produce that can settle all questions about how some
original sentence is to be understood, since language doesnt function that way. It is not
self-interpreting. (Bezuidenhout 2002: 113)
In performing pragmatic inferences, one may have to draw on contextual information concerning the circumstances of utterance, speaker intentions (to the extent that they are manifest), the participants interests and background assumptions, the purpose or point of the conversation, and so on. More generally, we
interpret someones utterance against the background of a shared system of knowledge and a common view of how nature works and how our culture works
(Searle 1980: 226227).
This idea has been elaborated, among others, by Charles Travis (e.g., 1985,
1996, 1997), Julius Moravcsik (1998), and Anne Bezuidenhout (e.g, 2002), to
name just a few. Moravcsik, for example, takes polysemy to affect almost all natural language expressions (cf. Moravcsik 1998: 33).8 Here is an example of his to illustrate the point:
213
The verb walk admits of polysemy. On the one hand, there is a common meaning core,
locomotion with legs in appropriate position. But one has to add: appropriate distance,
covered in an appropriate time. For what counts as a walk for a toddlers first attempts
does not count as a walk for a normal, healthy adult, and the walk of a recovering patient
in a hospital is still a different matter. What counts as a walk depends on different senses
of the word. (Moravcsik 1998: 3536)
How the verb walk is to be understood, i.e., what counts as a walk, depends on
context, more specifically on the various ways we can and do interact with the
world, on our purposes, intentions, etc.: Our understanding of words is influenced
by the variety of ways in which we relate to the world. (Moravcsik 1998: 91) And
there is no limit to the different applications a given word may enjoy; there is no
way of telling in advance what the different senses of a word are that we need it to
have in order to be able to talk about everything we want to talk about (cf. Moravcsik 1998: 37). There are as many different uses to which we may put the expressions of our language as there are different purposes to which we may interact
with the world. The meanings of our words provide rough and ready guidelines for
their own future applications, in the light of new interests, purposes, assumptions
etc. Yet, as Moravcsik emphasizes, we cannot predict or exactly specify all past
and future human interactions with nature that have or will affect our uses of the
words of our language.
To echo John Austin, the question is: Why do we call different things by the
same name? (Austin 31979a: 69) A sensible answer seems to be: It allows us to use
a finite vocabulary to talk about or describe an (in principle) infinite array of situations. Jonathan Cohen puts the point thus:
Rather, it is an enormous convenience that the same word can often be uttered in one or
other of several different though related senses. Instead of having to learn a very much
larger number of words, each with fixed and context-independent meanings, we can
learn a relatively small number of words with variable meanings and then exploit their
verbal or situational contexts of utterance in order to disambiguate their actual occurrences. (Cohen 1985: 132)
Our expressions are flexible enough to fit all the various situations we encounter almost perfectly. Jon Barwise and John Perry speak of the efficiency of language
(Barwise and Perry 1983: 5; cf. also Goodman 21976: 80). The expressions of our
language answer to our human needs, they serve our practical purposes. They can be
contextually adjusted, if need be. And need there is as we never encounter exactly
the same situation twice over. And when we encounter a situation we have not encountered before, our expressions allow us to conceptually structure it by using familiar vocabulary. Sometimes, though, we have to unduly stretch the meaning of a
given expression so that it will become applicable to the case at hand. That yields
polysemy or metaphor, depending on the extent to which the new situation differs
from old ones. Yet metaphor, just as polysemy, seems to defy formal treatment. Both
phenomena seem to require pragmatic inferences for their interpretation.
214
Metaphor in particular is commonly used to conceptualize abstract and mental

phenomena, often with the help of vocabulary familiar from the realm of sense experience. Metaphor is, as Nelson Goodman nicely put it, a matter of teaching an
old word new tricks. (Goodman 21976: 69) That is why we are feeling blue, or
complain about someone being cold, and so on. We also use metaphors in the explanation of human behavior. We say that he broke down under pressure thereby
exploiting the metaphor the mind is a brittle object (cf. Lakoff and Johnson 1980:
28). We say that she is feeling up, that her spirits rose, or that she sank into a coma.
With the help of these orientational metaphors (as Lakoff and Johnson 1980: ch.
4 call them), we conceptualize mental phenomena by giving them a spatial orientation. But abstract phenomena too are commonly conceptualized with the help of
metaphorical expressions. We construe theories and arguments as buildings: The
theory needs more support, it lacks a foundation; the argument collapsed, it was not
solid enough (cf. Lakoff and Johnson 1980: ch. 10). Or we think of love as if it
were a journey: look how far we have come; yet now we are at a crossroad; from
here on we have to go separate ways as this isnt going anywhere; we are in a deadend street. We commonly use expressions from the concrete realm of sense experience to conceptualize abstract and mental phenomena (cf. also Keller and Kirschbaum 2003: 36 and 99). So if one were to say, as formal semanticists with their
fondness for a clear-cut distinction between literal and figurative meaning would
be inclined to do, that metaphorical utterances are always strictly speaking false,
that would render most of our talk about abstract or mental phenomena false. Polysemy and metaphor are central features of natural languages without which we
would not be able to structure the multifariousness of our experience and to conceptually copy with mental and abstract phenomena.
Some philosophers take these and similar considerations to show that word
meanings (or concepts) are not ready-made entities but have to be constructed in
context. According to Robyn Carston, for instance, often in linguistic interpretation an ad hoc concept is constructed and functions as a constituent of what is
explicitly communicated. (Carston 2002: 357) And, presumably, pragmatic inferences are required in the construction of these ad hoc concepts. But one might even
go one step further and try to do without something like linguistic meaning (as traditionally understood) altogether (cf., e.g., Recanati 2004: 146ff; Croft and Cruse
2004: 97ff). The idea would be that although words encode semantic information,
the information itself does not amount to something like a full-fledged meaning.
Rather, it provides the starting point for pragmatic processes, which result in a contextually specified interpretation of the expressions in question. Linguistic meaning would provide information that constrains interpretation, while pragmatic
processes are required to yield the contextually relevant contribution the expression makes on a particular occasion of its use. According to this conception of
linguistic interpretation, the notion of linguistic meaning itself needs to be pragmatically spelled out.
1.4.
215
Conclusion
The preceding considerations cast doubt on the idea of a functional, systematic pairing of sentences with meanings (truth-conditions) as propagated by formal semantic
approaches to language. This way of approaching language seems to ignore the
highly context-sensitive nature of linguistic interpretation. Many natural language
expressions require for their proper interpretation that context be taken into account.
This kind of context-sensitivity rests on a conception of context according to which
context provides information about the participants shared interests, intentions, purposes, background assumptions and so on. The information is extracted by pragmatically inferring the contextually appropriate interpretation of a given context-sensitive utterance. Pragmatic inferences are indispensable in guring out what a given
expression contributes to the truth-conditions of a sentence containing it (i.e., to the
proposition expressed). Accordingly, the value assigned to the expression hardly deserves to be called a (purely) semantic value. And the process of interpretation performed hardly deserves to be called (purely) semantic interpretation. Rather, pragmatic processes seem to intervene at (almost) all stages of linguistic interpretation.
And maybe even the notion of linguistic meaning itself is a pragmatic notion.
Pragmatic semantics
2.1.
The center of pragmatic semantics
Just cancel the second word (maybe) of the last sentence in the preceding paragraph, and you have already reached the very center of Pragmatic Semantics (PS).
Whereas traditional semantics (including formal truth-conditional semantics as its
most developed form) starts with linguistic expressions and then asks how to define the various kinds of their respective linguistic meanings, PS starts just the
other way round, namely with the understanding (i.e., the knowing of the meanings) of actions. From the point of view of PS, semantics is a special part of pragmatics (Action Theory), in the ideal case therefore a part of Action Logic.
Throughout, we will have to make use of the following two-fold distinction
Type
Token
Action
(3)
(4)
Action-Product
(1)
(2)
or of its already language-related specification, respectively:

Utterance
Type
Token
Utterance-Product
Expression (form)
(Concrete) Expression
216
Whereas traditional semantics starts from the right side, (1) above, PS starts from
and is primarily focussed on the left one.
Correspondingly, pragmatics is covering quite differing programs. Whereas,
as exemplified in chapter 1 above, according to traditional semantics it may mean
anything which goes beyond field (1), the program of Pragmatic Semantics as
sketched here is to define in the end even (1)-meanings themselves in terms of action-theoretical concepts.
In PS actions (and their meanings) come first. For this semantic paradigm,
there even need not be any expressions to start from. Actions, think of gestures, for
example, do not need to be product-producing actions (utterances of expressions)
in order to have meanings. Thus, in this approach expressions themselves may
be recognized or introduced as what they are: constructions on the basis of (some
continuity in) the meanings of the corresponding actions. Thus, PS is primarily a
Semantics of Actions.
2.2.
Rule-based versus intention-based pragmatic semantics
In Pragmatic Semantics we also find the division between nominalists and Platonists, depending on whether type-meanings (meanings of action-types, field (3) in
the schema above) can themselves be defined in terms of token-meanings ((4) in
our schema) or not. Since type-meanings (of the PS-relevant sort) are social facts
(i.e., state of affairs which, relative to population P, are facts iff it is common believe in P that they are facts), whereas the respective meanings of action-tokens are
just meanings for some particular subject, this (Platonist versus nominalist) division is a special case of the more general division within Social Philosophy between collectivist versus individualist approaches, according to whether or not
meanings as social facts are taken to be conceptually inexplicable with reference to
individualistic terms.
This distinction is very often identified with a distinction between the two possible readings of Wittgensteins basic equation of meaning = use, 9 which one may
refer to as the rule-based reading versus the intention-based reading. Whereas according to the rule-based reading we have, in order to understand (i.e. to know the
meaning of) a sign A, just to know the rules we have to follow in order to use A
correctly, the intention-based reading is referring to what in a particular situation A
is being used for. Which alternative will be the most fruitful one? This has been the
most debated question in the whole field of pragmatic semantics up to now (see
2.52.7 below).
2.3.
Intention-based semantics
This overview will start from Intention-Based Semantics, which is definitely following a strictly nominalistic / individualistic strategy by embedding PS into gen-
217
eral individualistic Action Theory including Rational Decision Theory as its most
important part (see Jeffrey 1983). The positions of the main collectivist alternatives can then easily be located by referring to their denials of and their alternatives to the respective aspects of Intention-Based Semantics.
For Intention-Based Semantics, as sketched by Herbert P. Grice (1957) and
further developed mainly by Stephen Schiffer (1972), Jonathan Bennett (1976),
David Lewis (1969), Franz von Kutschera (1975) and Georg Meggle (2010),
human languages are primarily used for interpersonal communication. Accordingly, linguistic meanings as social facts are primarily defined in terms of the communicative regularities or conventions holding in the respective social groups
(populations). But as the meanings of communicative action tokens may and often
do diverge from the meanings of the exemplified action type, communicative actions have first to be defined more generally, i.e. on the token level. Thus, the Intention-Based Semantics-program (even when restricted to the action side, i.e., not
yet including some expressions as their action-products) has to involve two steps:
(i) explication of general concepts of communicative acts (as tokens); and (ii) explication of (meanings of) communicative action types by means of (i). Whereas
step (ii) is at the very center of Intention-Based Semantics, (i) is its necessary preliminary stage defining communicative acts in terms of their respective communicative intentions.10
2.3.1.
General communication theory
Communicative acts are instrumental acts, with the most basic types of communicative acts being either directives or informatives, depending on whether Ss (the
speakers) primary communicative aim is to get H (the hearer) either to do or to believe something. But notice that this way of conveniently speaking of speaker
and hearer does not imply that the respective communicative acts are restricted
to acoustic utterances. The differentia specifica of communicative acts versus instrumental acts in general is that communicative acts are acts which, at least from
the point of view of S, will be successful if and only if they will be understood by
their addressee H. This entails the so called reflexivity-condition of communication: For an act to be a communicative one, its actor S has to intend it to be
understood by H (as being a communicative act). A General Communication Theory has to explicate these postulates in a non-circular way.
For this explication, one best starts from the more general communicative
acts: Instrumental acts are acts done for some end (or, to put it differently, done
with specific intentions). More precisely (but already strongly adapted to the
special cases of decisions under certainty, i.e., cases where the deciding person
expects the results of her actions with a maximal degree of probability), person X
is doing f with the intention of bringing it about that the state of affairs A will be
the case: (1) X does f, (2) X wants that A, and (3) X believes (expects) that A
218
will be brought about as and only as a consequence of her doing f. Instrumental

acts will thus be successful iff the belief (3), let us call it the expectation of success, turns out to be true, i.e., if A will in fact come about in the way expected
by X, i.e., as and only as a consequence of Xs doing f. And you understand this
act iff you know its ends, i.e., iff you know the intentions with which f is or was
being done by X.11
As there are no restrictions as to what ends one may try to achieve by means
of communication, we have to distinguish between the intentions with which S is
performing her communicative act (i.e. her communication related intentions) in
general versus her acts communicative intentions in particular, this latter term
referring only to those intentions that are necessary for an act to be a (respective)
communicative act.
Unlike instrumental acts, communicative acts do not only involve the expectation of success, but also the expectation of understanding (i.e., the expectation
that H will understand Ss communicative act), both expectations connected in the
differentia specifica way already mentioned above: that S expects that success will
be brought about as and only as a consequence of understanding. Thus, for the
communicative act to be successful this expectation must turn out to be true, i.e.
communicative success must in fact come about in the way expected by S, i.e., as
and only as a consequence of Hs understanding.
The most crucial and contested question to be settled is: What do we have to
know in order to understand Ss communicative act? To answer this question, one
has to notice that the Reflexivity-Condition (RC) above already implies what we
might call the absolute openness of communication. Whatever condition has to be
satisfied in order for an act to be a communicative one (lets say condition ), it follows from RC that this very condition itself has to be intended to be recognized
by H as being satisfied as well, which is itself a necessary condition for the act to be
a communicative one, etc. Logically, this openness implied by RC will have no
upper end. For such an unrestricted (and in this sense absolute) openness lets
write openness* for short. From RC we get straight to O*:
O*
Communicative intentions must be open* ones.
On the grounds of this condition, Grices first proposal that from Ss point of view
only Hs recognition of Ss primary communicative intention is necessary, can immediately be seen to be inadequate. Grices communications are not open enough.
But Grices Basic Model (BM) can easily be improved:12
BM
Ss doing is an attempt at communication, addressed to H, and of the content that H is to do r, iff (i) S intends, by doing the act, to make H do r; and
(ii) S believes (expects) that she will achieve this aim iff she is letting H
openly* know that S has this intention (i)
219
Communication is thus regarded as proposed by Grice plus the factor of openness*. This kind of openness*, however, relates only to communicative intentions,
i.e., to those intentions which are necessary conditions for an act to be a communicative act. Sincerity is not a necessary condition for an act to be a communicative
one; communicative openness (being a case of openness*) is therefore absolutely
compatible with the insincerity of the communicative act in question (including
downright lies).
This concept of communicative acts (as defined in BM) is a very broad one.
Though it involves the expectations of understanding and of success, it implies nothing about the reasons these expectations may or will have to be based on in order
to be (rationally to be expected by S to be) satisfied. This is fine as far as communicative acts (and their subjective meanings) in general are concerned. But things
change when we come to the special cases of communication by means of (types
of) actions with some intersubjective communicative meaning.
2.3.2.
Meanings of action types
Meaning and understanding (i.e. knowledge and recognition of meaning) are two
sides of one and the same coin. This holds both for subjective and intersubjective
meanings. (Inter-) subjective meanings of (the type f of) an action are relative to
the (class of) persons who understand the actions (of type f) as having these (inter-)
subjective meanings. In addition, meanings of both kinds are relative to (types of)
situations.
The fact, that some gesture (of type) f, lets say thumbs up, as used in the
population P in situations of type , means p (e.g.: you are fine) may be
equated roughly with the following regular meaning condition of f in P relative to
: Whenever (or when in sufficiently many cases) in a -situation a S (as a member
of P) addresses H (as another member of P), making a gesture (of type) f, it holds
that by doing f, S is trying to communicate to H that p (You are fine). Since
(type-) meanings (as social facts) have to be known to be (intersubjective) meanings, the above regular-meaning-definiens will have to be postulated as being common knowledge in P (regular meaning in a narrow sense of f in P relative to ).13
Regular meanings of f (in P for ) are, in -situations and for all members of P, the
best possible reasons for the respective Ss expectation of understanding.
Conventional meanings result from the likewise rationally founded expectation
of success: H will believe what S is trying to make her believe by this very action f
(namely that p) iff H will understand Ss communicative action f. Normally this
will hold only if (S expects that) H believes that, as far as p is concerned, neither S
will be deceptive to her (H) nor that S himself will be at fault.
And these expectations will be stable (within P and relative to -situations) in
the long run only if they will always (or in most or at least in sufficiently many
cases) turn out to be satisfied. And this again will be true only if both S and H will
220
be following a joint strategy, which in turn will be stable only if this strategy is
based on joint interests, as for example, that H is to believe p iff it is in fact the case
that p. Iff, in P and relative to -situations, the communicative expectations of success are based in this way, then the regular meanings of the respective communicative actions will also be conventional ones. This explication is roughly the same
as David Lewiss explication of signalling-conventions (in Lewis 1969).
Now, and this is the decisive step from actions to expressions as their products,
the meanings of (complete or unstructured) signs or expressions (you may think of
the gesture thumbs up again) can be equated with the conventional meanings of
the communicative acts of producing these expressions.
2.3.3.
Linguistic meaning
Unfortunately, when turning to languages and linguistic meanings of structured

signs, Lewis opted out of the whole Intention-Based Semantics-program by putting his proposals on the shoulders of traditional realistic semantics. How this predicament may be avoided in relation to a simple language (of first order predicate
logic) is shown in Kutschera (1975), who provides us with an illustration of the
schema according to which a grammar can be constructed. For a discussion of
such a pragmatic reconstruction see Bennett (1976) and Meggle (2010), the former
trying to couch it in more behavioristic terms, the latter in more formal terms. Following this approach, the meanings of so called realistic semantics (even in its
most developed form of intensional semantics) can be shown to be successfully explicable in terms of Intention-Based Semantics. But, as in many other places, explicability does not imply reduction.
2.4.
Classical speech act theory
Though rightly looked at as being a paradigm case of pragmatics, (classical)

speech act theory was definitely not (intended to be) a special case of PS. This was
clear from the start because John L. Austin (1962), in his basic distinction between
locutionary, illocutionary and perlocutionary (aspects of) acts, was referring to the
use of a sentence in its literal meaning in the sense of Frege. From Austins point
of view, speech act theory should only be interested in the illocutionary functions
of speech acts. This view is shared by his most influential interpreter, John R.
Searle. In this sense, Searles slogan that a theory of language is part of a theory of
action (Searle 1969: 17) can only be misleading.
As to the class of illocutionary acts, Austin did not give us an explicit definition, only an open list of examples. But there are two elements in his characterizations of illocutions which obviously contradict the explication of communicative
acts (in the wide sense of directives and informatives) as sketched in 2.3.1 above.
According to Austin, illocutionary acts are (i) essentially conventional acts; and, as
221
such, (ii) they have to exclude any perlocutionarity.14 Whereas (i) may be
weakened (cf. Bach and Harnish 1979) by distinguishing between essentially conventional illocutionary acts (including Austins favorites of marrying or baptizing
someone) on the one side, and mere communicative actions on the other side, the
non-perlocutionarity of illocutions remained the central dogma of classical
speech-act-theory (even with regard to communicative acts).
There is some truth in this dogma, but this truth can be easily reduced to the distinction between communication related intentions in general (i.e. perlocutionary
intentions in a broad sense) versus communicative intentions in particular (i.e. perlocutionary intentions in the narrow sense) (cf. 2.3.1 above). With regard to the
broad sense of perlocutionary, the non-perlocutionary dogma is (by definition)
true; but in this sense it is (as implied in 2.3.1) no objection to the presented definition of (the communicative variant of) illocution at all.
2.5.
Speech act based PS
It was William P. Alston (1964) who first tried to build a PS on speech act theory.
His basic idea was that two (unstructured) expressions and can be said to have
the same meaning iff they have the same illocutionary potential, i.e., could be used
to perform the same illocutionary acts, where the same illocutionary acts was
meant to refer not only to acts of the same illocutionary types (assertions, directives, questions etc.) but also to these acts as having also the same contents (assertions that p, directives to bring about q, questions as to whether it is the case etc.).
In other words: Illocutionary potential in Alstons sense is already a content-determined illocutionary act potential (CDIAP). And though his basic idea was presented by him only in the form of an identity criterion for the meanings of and , it
was clear that Alston wanted to identify and thus also define the meanings of expressions by means of their CDIAPs. That was the essential difference: Whereas in
Austins theory meanings of expressions are presupposed as being given, Alstons
idea was to define these expression meanings themselves by means of the meanings of illocutionary actions. If this is the line along which meaning should be
analyzed, then the concept of an illocutionary act is the most fundamental concept
in semantics and, hence in the philosophy of language (Alston 1964: 37 [emphasis
mine]).
If we substitute a communicative act for an illocutionary act, Alstons
speech act approach would seem to coincide with the Intention-Bases Semantics as
sketched in 2.3 above. But notice that even Alston in 1964 subscribed to the two
dogmas of classical speech act theory. However, nearly 40 years later, he has extended his basic idea of meanings as CDIAPs into a full blown PS of its own (Alston 2000). Beside Alston, the most detailed and best developed rule-oriented PS is
Eike von Savignys theory (1983). Whereas the intention-based PS as sketched
above starts with the communicative meaning of concrete utterances (action-
222
tokens, i.e. field (4) of the schema in 2.1 above), Savignys rule-based PS starts
with the conventional meaning of utterance types (action-types, i.e.field (3) in the
same schema). According to this approach, differences in meaning are correlated
with differences in the conventional commitments of the persons involved in the
respective utterance-situation. And these commitments in turn are defined by
means of rules as explicated by Hart (1961) with reference to regularities, sanctions and acceptance of these sanctions. Finally, sentence meanings are to be introduced as theoretical constructs in order to systematically explain the salient structures to be found in these regularities.
To systematically compare the IB- versus these and other Rule-Based Approaches to PS with regards to their various pros and cons is one of the biggest desiderata in the whole field of modern philosophy of language.
2.6.
Variety of language games
As far as Wittgensteins meaning = use equation is concerned (see 2.2 above),

the usual distinction between the rule-reading versus the intention-reading of it
should not be the last word. At least not in the sense of an either-or distinction because there is some chance that both readings need not be really exclusive. First, as
noticed in 2.6, the two readings may refer to different explication-levels, the intention-reading to the token-one, the rule-reading to the type-one. Next, as noticed in
2.4, only a special class of perlocutionary intentions is relevant for the definition of
communicative/illocutionary acts: communicative intentions (see 2.3.1). As these
intentions are necessary conditions for some action to be a communicative action
(a directive, for example), they may be looked at as being part of the internal relation, which, according to Wittgenstein ( 21), is constitutive of the kind of
meaning his various language games are meant to be exemplifications of. According to this reconstruction of Wittgensteins meaning as use theory,15 the relevant intention-reading is part of the rule-reading itself.
But there are many other language games (many other illocutions) whose internal relations do not include any reference to perlocutionary intentions. One
might therefore say that meanings (illocutions) do not per se entail perlocutions.
But from this it does not follow that perlocutions (as part of meanings) are per se
excluded. There is more than just one kind of meaning.
2.7.
Evaluations
This difference is of the utmost importance for any evaluation of the various PSapproaches.
What kinds of illocution are the most important ones? The IB-semanticists (as
sketched in 2.3 above) take it that interpersonal communication is at the very basis.
But they do not deny, of course, that languages may have many other functions as
223
well. Some neo-Griceans, interpreting Grices original starting-point of speakers

meaning as covering much more than communicative intentions, do accordingly
extend this basis and include, e.g., the intention to express ones thoughts or feelings (Davis 2003). Others, like Austin, think it best to start with fully institutionalized or even ritualized kinds of performances. Finally, there are scholars like Brandom (1994), who take the whole PS-project (of starting with primitive acts or
intentions and then constructing something like linguistic meanings out of them) to
be on the wrong track because they claim that language has to be presupposed before attributing even these most primitive elements.
Before we postulate the one PS which embraces all sorts of things language
may be used for, we should keep in mind that we just cannot come up with the complete theory of language and still hope for its consistency.
Notes
1. For an overview of the development of semiotics, see e.g., Scholz 2003, for semiotics
and pragmatics Nth, this volume, and for a general historiographic overview of pragmatics Koyama, this volume.
2. For an account of the relationship between semantics and pragmatics cf. Saeed, this volume.
3. Kaplan also speaks of the character of indexical expressions as of a rule that fully determine[s] the referent for each context. (1989: 491)
4. For this specific conception of context cf. Lewis A context is a location time, place,
and possible world where a sentence is said. (1980: 21), and Kaplan, who adds the
speaker or acting participant as a further constituent, argues that an index representing a
context is a 4-tuple <w, x, p, t>, consisting of a world (w), an actor (x), a place (p) and a
time (t) such that x is in w at t in p (Kaplan 1989: 509).
5. Stalnakers notion of context might be promising here: So I propose to identify a context (at a particular point in a discourse) with the body of information that is presumed,
at that point, to be common to the participants in the discourse. (1998: 98)
6. For more on inferences to the best explanation in general, see, e.g., Lipton (1991).
7. Some context-sensitive sentences may have preferred or default completions, though. In
the absence of any special stage setting, we will interpret the sentence: Sams daughter
got married and became pregnant, for example, as meaning that Sams daughter first
married and then became pregnant. For more on the notion of a preferred or default interpretation see Levinson (2000).
8. See also Searle (1978: 220221) and (1980: 221).
9. Which may be treated as something like the cornerstone of the whole of PS.
10. For formal explications of step (i) see Meggle (1997) and for step (ii) Meggle (2010).
11. Following Max Weber one could also say that Action Understanding in this sense is to
know the actions subjective meaning / subjektiver Sinn.
12. With (BM) being already a reconstructed version of Grices own version (1957): S
meant something by x is (roughly) equivalent to S intended the utterance of x to produce some effect in an audience by means of the recognition of this intention.
224
13. Where Common Knowledge in P that p may be easily defined this way: (i) All members
of P know that p; (ii) All members of P know that (i) etc., with this etc to be defined
in formal terms as usual by recursion.
14. See Austin (1962: ch. 9), where perlocutionary effects are by definition restricted to effects going beyond (conventional) illocutionary effects.
15. Very forcefully proposed for the first time by Feyerabend (1955) and heavily relying on
Wittgensteins famous language game, cf. 2 in his Philosophical Investigations.
References
Alston, William P.
1964
Philosophy of Language. Englewood Cliffs, NJ: Prentice-Hall.
Alston, William P.
2000
Illocutionary Acts and Sentence Meaning. Ithaca and London: Cornell University Press.
Austin, John L.
1962
How to do things with Words. Oxford: Oxford University Press.
Austin, John L.
31979a
The Meaning of a Word. In: John Austin, Philosophical Papers, 5575. Oxford: Oxford University Press.
1979
Linguistic Communication and Speech Acts. Cambridge, MA: Harvard University Press.
Bach, Kent
1994
Semantic slack. In: Savas L. Tsohatzidis (ed.), Foundations of speech act
theory: Philosophical and linguistic perspectives, 267291. London/New
York: Routledge.
Bach, Kent and Anne Bezuidenhout
2002
Distinguishing Semantics and Pragmatics. In: Joseph Keim Campbell, Michael
ORourke and David Shier (eds.), Meaning and Truth: Investigations in Philosophical Semantics, 284309. New York/London: Seven Bridges Press.
Barwise, Jon and John Perry
1983
Situations and Attitudes. Cambridge, MA: MIT Press.
Bennett, Jonathan
1976
Linguistic Behaviour. Cambridge: Cambridge University Press.
Bezuidenhout, Anne
2002
Truth-Conditional Pragmatics. Philosophical Perspectives 16: 105134.
Bloom, Paul
2000
How Children Learn the Meanings of Words. Cambridge, MA: MIT Press.
Borg, Emma
2004a
Formal Semantics and Intentional States. Analysis 64: 215223.
Borg, Emma
2004b
Borg, Emma
2007
Minimalism versus Contextualism in Semantics. In: Gerhard Preyer and Georg
Peter (eds.), Context-Sensitivity and Semantic Minimalism: New Essays on
Semantics and Pragmatics, 339359. Oxford: Oxford University Press.
225
Brandom, Robert B.
1994
Making it explicit. Reasoning, Representing and Discursive Commitment.
Brandom, Robert, B.
2002
Pragmatics and Pragmatisms. In: James Conant and Urszula M. Zeglen (eds.),
Hilary Putnam: Pragmatism and Realism, 4059. London: Routledge.
Clark, Eve V.
1993
The lexicon in acquisition. Cambridge Textbooks in Linguistics, Cambridge:
Carnap, Rudolf
1956
Meaning and Necessity: A Study in Semantics and Modal Logic. 2nd ed. Chicago: The University of Chicago Press.
Carston, Robyn
2002
Cohen, Jonathan L.
1985
A Problem about Ambiguity in Truth-Theoretical Semantics, Analysis,
Vol. 45/3: 129134.
Croft, William and Cruse, Alan D.
2004
Cognitive Linguistics. Cambridge Textbooks in Linguistics, Cambridge: Cambridge University Press.
Davis, Wayne A.
2003
Meaning, Expression, and Thought. Cambridge: Cambridge University Press.
Feyerabend, Paul
1955
Wittgensteins Philosophical Investigations. Philosophical Review: 449483.
Frege, Gottlob
1891
Funktion und Begriff. Lecture held 1891 at a meeting of the Jenaischen Gesellschaft fr Medizin und Naturwissenschaft. Reprinted in Gottlob Frege
1986, 6th ed., 1839.
Frege, Gottlob
1892a
ber Sinn und Bedeutung. Zeitschrift fr Philosophie und philosophische Kritik 100, 2550. Reprinted in Gottlob Frege 1986, 6th ed., 4065.
Frege, Gottlob
1892b
ber Begriff und Gegenstand. Vierteljahrsschrift fr wissenschaftliche Philosophie XVI, 192205. Reprinted in Gottlob Frege 1986,6th ed., 6680.
Frege Gottlob
1986
Funktion, Begriff, Bedeutung. Edited by Gnther Patzig, 6th ed. Vandenhoeck
und Ruprecht, Gttingen.
Goodman, Nelson
1976
Languages of Art. An Approach to a Theory of Symbols. 2nd ed. Indianapolis:
Hackett.
Grice, H. Paul
1957
Meaning. The Philosophical Review 66: 377388. Reprinted in Grice 1989.
Grice, H. Paul
1989
Hart, Herbert L.A.
1961
The Concept of Law. Oxford: Clarendon Press.
226
Jeffrey, Richard C.
1983
The Logic of Decision. 2nd ed. Chicago: University of Chicago Press.
Kaplan, David 1989
Demonstratives. An Essay on the Semantics, Logic, Metaphysics, and Epistemology of
Demonstratives and Other Indexicals. In: Joseph Almog, John Perry, and Howard Wettstein (eds.), Themes from Kaplan, 481563. Oxford: Oxford University Press.
Keller, Rudi and Ilja Kirschbaum
2003
Bedeutungswandel: Eine Einfhrung. Berlin: de Gruyter.
Koons, Robert
2009
Defeasible Reasoning. In: Edward N. Zalta (ed.), The Stanford Encyclopedia
of Philosophy (Winter 2009 Edition). <http://plato.stanford.edu/archives/
win2009/entries/reasoning-defeasible/>.
1975
Conventions of Language and Intensional Semantics. Theoretical Linguistics
Vol 2/3: 255283.
1976
Einfhrung in die intensionale Semantik. Berlin: de Gruyter.
1980
Metaphors We Live By. Chicago: University of Chicago Press.
2000
Lewis, David
1969
Convention. Cambridge, MA: Harvard University Press.
Lewis, David
1975
Languages and Language. In: Keith Gunderson (ed.), Minnesota Studies in the
Philosophy of Science, Vol. VII, 335. University of Minnesota Press. Reprinted in David Lewis 1983: Philosophical Papers. Volume 1, 163188. Oxford: Oxford University Press.
Lewis, David
1980
Index, Context, and Content, in Stig Kangar and Sven hman (eds.), Philosophy and Grammar, 79100. Dordrecht: Reidel.
Lipton, Peter
1991
Inference to the Best Explanation. London: Routledge.
Meggle, Georg
1997
Grundbegriffe der Kommunikation. Berlin: de Gruyter.
Meggle, Georg
2010
Handlungstheoretische Semantik. (Habilitation thesis 1984). Berlin: de
Gruyter.
Montague, Richard
1970
English as a Formal Language. In: Bruno Vicentini et al. (eds.), Linguaggi
nella societ e nella tecnica, 189224. Mailand: Edizioni di Comunit. Reprinted in Montague 1974, 188221.
Montague, Richard
1970b
Universal grammar. Theoria 36: 373398. Reprinted in Montague 1974,
222246.
227
Montague, Richard
1973
The proper treatment of quantification in ordinary English. In: Jaakko J. Hintikka, Julius M. Moravcsik and Patrick Suppes (eds.), Approaches to natural Language, 221242. Dordrecht: Reidel. Reprinted in Montague 1974,
1734.
Montague, Richard
1974
Formal Philosophy: Selected Papers of Richard Montague. Edited by Richmond H. Thomason. New Haven/London: Yale University Press.
Moravcsik, Julius M.
1998
Meaning, Creativity, and the Partial Inscrutabilty of the Human Mind. CSLI
Publications, Stanford.
Manor, Ruth
2001
On the Overlap of Pragmatics and Semantics. Synthese 128: 6373.
Morris, Charles
1938
Charles Morris (eds.), International Encyclopedia of Unified Science. Vol. 1,
Nr.2, 77138. Chicago: University of Chicago Press. Reprinted in Charles
Morris, 1971, Writings on the General Theory of Signs, 1771. Mouton: The
Hague/Paris, quotations are to Morris 1971.
Partee, Barbara
1973
Some transformational extensions of Montague grammar. Journal of Philosophical Logic 2: 509534.
Partee, Barbara
1975
Montague grammar and transformational grammar. Linguistic Inquiry 6/2:
203300.
Portner, Paul and Barbara H. Partee
2002
Introduction. In: Paul Portner and Barbara H. Partee (eds.), Formal Semantics:
The Essential Reading, 116. Oxford: Blackwell.
Recanati, Franois
2004
Literal Meaning. Cambridge; Cambridge University Press.
Savigny, Eike von
1983
Zum Begriff der Sprache. Konvention, Bedeutung, Zeichen. Stuttgart: Reclam.
Schiffer, Stephen
1972
Meaning. Oxford: Oxford University Press.
Scholz, Oliver R.
2003
Semiotik und Hermeneutik. In: Roland Posner, Klaus Robering and Thomas A.
Sebeok (eds.), Semiotik/Semiotics: Ein Handbuch zu den zeichentheoretischen
Grundlagen von Natur und Kultur / A Handbook on the Sign-Theoretic Foundations of Nature and Culture, Vol. 3, 25112561. Berlin: de Gruyter.
Searle, John R.
1969
Searle, John R.
1978
Literal meaning. Erkenntnis 13: 20724.
Searle, John R.
1980
Background of Meaning. In: John Searle, Ferenc Kiefer, Manfred Bierwisch
(eds.), Speech Act Theory and Pragmatics, 221232. Dordrecht: Reidel.
228
1998
On the Representation of Context. Journal of Logic, Language, and Information 7. Reprinted in Stalnaker 1999 Context and Content. 96113. Oxford: Oxford University Press.
Stanley, Jason
2002
Making it Articulated. Mind and Language 17: 149168.
Stanley, Jason
2005
Semantics in Context. In: Gerhard Preyer and Georg Peter (eds.), Contextualism in Philosophy: Knowledge, Meaning, and Truth, 221253. Oxford: Oxford
University Press,
Tomasello, Michael
2003
Constructing a Language: A Usage-Based Theory of Language Acquisition.
Travis, Charles
1985
On what is strictly speaking true. Canadian Journal of Philosophy 15:
187229.
Travis, Charles
1996
Meanings Role in Truth. Mind 105: 451466.
Travis, Charles
1997
Pragmatics. In: Bob Hale and Crispin Wright (eds.), A Companion to the Philosophy of Language, 87107. Oxford: Blackwell.
1953
Philosophische Untersuchungen. Ed. by G.E.M. Anscombe and Rush Rhees.
Oxford: Basil Blackwell.
8.
Foundations of pragmatics in functional

linguistics
1.
Introduction
The relatively young discipline of linguistic pragmatics has had forbears in several
older branches of linguistics. This chapter focuses on important prefigurations of
parts of linguistic pragmatics from the late nineteenth century to around 1970
which were informed by a functional view of language. We view functional linguistics as the study of language that starts its research with a focus on the functions of language in social life: the effects of language use, differentiated as to
types of communicators, types of contexts and types of language uses. This includes the analysis of the primarily intended goals of the speech participants in
their speaking and reacting as well as the study of the long-time effects of language
use in the life of the individual and in the history of the language.
One could name quite a few functionally inclined linguists of that period whose
theories are to some extent congenial to modern pragmatics, starting with the
theorists Dwight Whitney (18271894), Michel Bral (18321915), Georg von
der Gabelentz (18401893) and Henry Sweet (18451912) and ending with, e.g.,
Andr Martinet (19081999) and Michael Halliday (*1925). However, in this article we will confine ourselves to results of functional linguistics that have actually
been transmitted to linguistic pragmatics and may be counted among its foundations. Such results are found foremost in the works of four theorists: the German
psychologist Karl Bhler (18791963), the Czech scholar of English Vilm Mathesius (18821945), the Russian general linguist and scholar of Slavonic languages
Roman Jakobson (18961982) and the British general linguist Michael Halliday.
Of these, we will discuss the relevant theories of the first three. Bhler proposed an
influential model of three fundamental language functions, which Jakobson
came to extend to the model with six such functions that has had a strong impact
on pragmatics, the ethnography of communication and social semiotics.1 Furthermore, both Bhler and Jakobson analysed the functioning of the deictic elements of language, emphasising their role in the situated nature of language
use. Mathesius, who was more specifically concerned with the comparison of living languages as to their possibilities of expression, provided the foundations of
the functional analysis of utterances in terms of the notions theme and rheme
(topiccomment), which has become one of the stock items of pragmatic
analysis. For their ideas in this domain, both Bhler and Mathesius were indebted
to the work of the older and less well-known psychologically oriented German lin-
230
guist Philipp Wegener (18481916), who held views that also show a strong affinity to those of modern linguistic pragmatics. In our sections 25, we will in turn
discuss the contributions by Wegener, Bhler, Mathesius and Jakobson. In the concluding section 6, we consider the questions of whether and how their insights into
the functionality of language can still be of value for linguistic pragmatics. In our
view, these insights still provide valid reasons to integrate the description of elements of the situational context into linguistic analysis itself and thus to question
the wisdom of completely separating research into the system of language from
pragmatics, the study of its use.
2.
Philipp Wegener
Philipp Wegener (18481916), a classics teacher, grammar school headmaster and

psychological linguist from Northern Germany, developed in a relatively small
body of work a remarkably modern, communication-oriented analysis of language.2 It resulted from his opposition against a particular claim of contemporary
historical linguistics. The Neogrammarian theorists had acknowledged the relevance of psychological factors for processes such as the borrowing of linguistic
elements by one language or dialect from another, but not for so-called internal,
primarily phonetic language change, which was thought to be caused purely by
physiological mechanisms. Motivated by his study of developments in the local
dialects, Wegener wished to demonstrate the concrete psychological and communicative background also of internal changes, and this forms the first topic in
his 1885 publication Investigations into the Fundamental Questions on the Use of
Language (Untersuchungen ber die Grundfragen des Sprachlebens).3 Wegener
uses the key concept of automatisation: changed phonetic realisations may appear in hindsight not to fulfil any special goals, but they are in fact the results of
conscious choices by speakers, made at one point in time, of alternative possibilities with special communicative effects, which subsequently have become automatised through habit. A communicative basis can also be seen in the ontogeny of
language. In time, the child learns to use sound patterns with meanings of a general
nature, but this develops only after a stage in which a particular vocal expression is
a purposive action, essentially connected to the childs momentary needs and
meant to appeal to the carers for help (Wegener [1885] 1991: 419).
Wegener goes on to claim that the general purpose of speaking is indeed to influence the actions, volition or thoughts of another person in a way the speaker
thinks worthwhile (Wegener 1991: 6467, 180). He has no use for the well-known
subjectivist picture of the individual speaker who, in speaking, merely mirrors
his own thoughts; such a view would make the idea of communication as accommodation between different minds all but void (Wegener 1902: 402). Focusing
rather on the role of the hearer, Wegener emphasises that the hearer is the true locus
231
of meaning, because he constructs meaning first and foremost not from linguistic
cues but according to his knowledge of the world and of the purposes of social action, helped further by his inferential abilities. Within this dialogical theory of
meaning, the function of the words and grammatical elements the speaker presents
is to help organise the hearers expectations (Wegener 1991: 157158, 180182).
The importance of the asymmetrical interaction between speaker and hearer is
highlighted in the detailed analysis of many grammatical constructions that Wegener provides in his book. As the general functional structure of utterances, Wegener proposes a bipartition consisting of exposition (Exposition) and predicate (Prdicat).4 The predicate carries the core function of the utterance, it
indicates the new and important material for the situation of utterance, and whatever effect is achieved in the hearer is connected primarily to this part. The exposition is everything in the utterance that the speaker has deemed necessary for
the hearer to situate the information contained in the predicate in the way it is
meant to be situated (Wegener 1911; 1991: 1922).
Elements of the situation itself may provide enough information to let the
predicate do its work, as Wegener explains. Any aspects of the hearers momentary
attention, such as a context of actions that have just been carried out (or are in the
process of being carried out), a topic from the social or institutional situation at
hand or an element from the general cultural situation, can be sufficient for him to
understand an unaccompanied predicate Well done as uttered by the speaker.
Language thus need not always be congruous to what the mind understands (Wegener 1991: 2127).5 However, an exposition of linguistic material is often
added to help achieve clarity as to the situated fact in relation to which the predicate has to be understood. Note that according to Wegener, the exposition need not
always precede the predicate; there may be many reasons for a speaker to choose
the reverse ordering. Intonation, but also grammar, can fulfil the task of indicating
to the hearer just what part of the utterance has which role.
Wegener (1991: 4760) also relates important phenomena of word semantics to
his communicative bipartition of utterances.6 It is not the same for a word to appear
in the exposition of an utterance or to appear in the predicate. The words of the exposition are used by the hearer to determine the real-world referent, and their
meanings should therefore be relatively congruous to the relevant thought. By
contrast, words chosen for the predicate may well have a variable, only loosely-fitting relation to their real-world referent and often have a metaphorical character.
For his precise understanding, the hearer can rely on inferences that are drawn
from elements of the situation and from expositional parts of the utterance. When
such inferences become automatised through repeated use, the word meaning loses
its flexibility, increases in its congruousness to thought and thereby becomes appropriate to fulfil a role also in the expositional parts of utterances. To give a modern example: it is often said, figuratively, that the family is the cornerstone of society. The phrase has become so well-known that the original predicate part (the
232
cornerstone of society) can now also be used in a role of exposition and still be
understood to have the family as referent: the cornerstone of society is in trouble.
But this does not yet hold unrestrictedly for cornerstone on its own; of course,
this could well be a further development. The process as sketched amounts to semantic change and, more specifically, the well-attested, continuous fading (abblassen) of metaphors (Wegener 1991: 5157).
It is an interesting fact that in this picture, pragmatic analysis can explain linguistic phenomena that are usually considered to be semantic or grammatical in nature. Pragmatics could thus be conceived of as an overarching discipline encompassing linguistics the full potential of such a conception does not seem to be
exhausted by modern pragmaticists.7
Although Wegeners approach was appreciated and taken into account by such
functionalist theorists as the important Polish anthropologist Bronisaw Malinowski (18841942) and the British linguists Alan Gardiner (18791963) and Michael Halliday,8 it never gained real notoriety in general linguistics and was not
used by the founders of modern linguistic pragmatics. Nor did the revival of Wegeners pioneering ideas in works by the German communication theorists Gerold
Ungeheuer (19301982) and Johann Juchem (19392003)9 succeed in making
them known beyond a rather small circle of pragmaticists. However, some of Wegeners proposals did reach modern pragmatics through an intermediary. Karl
Bhler helped spreading Wegeners analysis of the interweaving, in each utterance, of the speakers expression of thoughts, the intended effect on the hearer and
the representation of a part of reality. This tripartite analysis had a major influence
on the development of Bhlers theory of the functioning of language, as can
be seen already in one of his earliest linguistic publications (Bhler 1909b:
119123).10 It eventually led to Bhlers model of three language functions, the
organon model, to be discussed in section 3.1, which in turn was further differentiated by Roman Jakobson and became part of linguistic pragmatics in that extended version (see section 5.1).
3.
Karl Bhler
3.1.
Bhlers axiomatic view of language functions
In 1934, Karl Bhler (18791963), Ordinarius professor of psychology at the

University of Vienna, published a monograph entitled Theory of Language. The
Representational Function of Language (Sprachtheorie. Die Darstellungsfunktion
der Sprache). Its introductory First Part deals with The Principles of Language
Studies, which include four basic axioms. The first of these axioms, the organon-model, is introduced as the maxim of the three semantic functions (Leitsatz
von den drei Sinnfunktionen) of all linguistic structures (Bhler 1990: 28; 1934:
233
22).11 Bhler cites Platos definition in Kratylos of language as an organum

(means/tool) for a speaker to inform the other of something about the things
(Bhler 1990: 30; 1934: 24) to motivate the term organon-model. It was this
1934 version of the model of language functions that inspired a number of later
scholars to develop the notion of Functional Linguistics with a strong emphasis
on pragmatic as well as sociolinguistic aspects, and it is quoted as such in textbooks to this day.12 Before we look at its systematic theoretical status and further
adaptation, however, we need to analyse its development in Bhlers own work.
1934 did not mark the starting point of Bhlers work on meaning functions
by any means; rather, it was its climax and almost also its ending due to the interruption of Bhlers work in 1938 when Nazi Germany annexed Austria.13 Building on Edmund Husserls (18591938), Anton Martys (18471914) and Wilhelm Wundts (18321920) theories of meaning,14 as well as on his own work on
thought psychology in the context of the Wrzburg school around Oswald Klpe
(18621915),15 Bhler had published an article on the Critical assessment of new
sentence theories (Kritische Musterung der neueren Theorien des Satzes) as early
as 1918, in which he proposed a triadic model of language functions (Leistungen). It is based on the three main foundations of any meaning-constitutive
interaction, i.e.,
1) the speaker whose feelings and attitudes are given expression (Kundgabe),
2) the hearer for whom it provides a stimulus that elicits reactions (Auslsung),
3) objects or states of affairs that form the referents for representation (Darstellung) (Bhler 1918: 16).
After moving to the Chair of Psychology at Vienna University in 1922, Bhler
went on to develop this triadic model further in two major psychological studies. In
The Crisis of Psychology (Die Krise der Psychologie) of 1927, the model serves as
the basis for a critical assessment of contemporary psychology. The two basic semantic functions of expression (Kundgabe, which, as Bhler emphasises, also includes its receptive counterpart, Kundnahme) and guidance (which also involves
mutuality) are present even in animal communication; the ability to produce symbolic representations (of objects/states of affairs), on the other hand, is characteristic of human communication and helps to adapt the first two functions to a higher
level, so that all three of them are only fully activated in human communication
(Bhler 1927: 3233, 37, 5051). The three functions model thus forms a framework for conceiving of psychology as the integrated study of subjective experience, social behaviour and structures of objective sense (Bhler 1927: 29,
5762). The enormous conceptual range of this function-concept is evident also
in the further monograph on Theory of expression (Ausdruckstheorie) of 1933,
where Bhler refers to it as the basis to analyse the relationship between miming,
gesture and language (Bhler 1933a: iiiiv).
234
With specific respect to linguistics, Bhler explicitly claimed axiomatic

status for the functions model in his two articles The whole of language theory, its
system and its parts (Das Ganze der Sprachtheorie, ihr Aufbau und ihre Teile) and
The axiomatics of the language sciences (Die Axiomatik der Sprachwissenschaften), published in 1932 and 1933, respectively. He now called the functions
expression (Ausdruck), appeal (Appell), and representation (Darstellung)
(Bhler 1932: 106; 1933a: 7490) and incorporated them into a set of four principles, which prefigure the axiomatic introductory part of Sprachtheorie (although they appear in different order).16 Compared with the earlier versions of the
three functions-model, both articles share with the Sprachtheorie the detailed discussion and acknowledgement of Ferdinand de Saussures (18571913) Cours de
linguistique gnrale (published posthumously in 1916; German translation published in 1931).
Saussures famous lectures had motivated Bhler to reconceptualise language
theory from a semiological (Saussure), or as Bhler calls it, sematological
viewpoint (1990: 33, 42; 1934: 27, 3435). However, Bhlers interpretation of the
Cours differed in significant ways from contemporary structuralist readings.17
Whilst acknowledging the epoch-making impact of the Cours (Bhler 1990: 78;
1934: 7), he criticises its presentation of the speech circuit as a relapse into 19th
century psycho-physics and he insists that linguistic communication comprises
different types of signs (1990: 3134, 4243; 1934: 2528, 3437). Most important
in our context is his discussion of Saussures distinction of a linguistique de la
langue from a linguistique de la parole. Bhler puts the emphasis not on the langueaspect, but instead praises Saussure for having shown what would have to be discovered in order to be really able to initiate a linguistique de la parole (1990: 8;
1934: 7). This foregrounding of the speech/action aspect of language was at odds
with the predominant reading of the Cours as an endorsement of a system-oriented
core linguistics. It thus provides a first indication of Bhlers interest in the pragmatic orientation as an essential methodological perspective for linguistic research.
Besides the pragmatically interpreted Cours, Bhler cites three further works
as coming closest to providing a functionalist grounding of linguistics as an urgently needed complement to the old grammar (1990: 27; 1934: 23):
a) Philipp Wegeners Investigations into the Fundamental Questions of the
Use of Language ([1885] 1991) (discussed in our section 2) had taught Bhler
about the listeners role in the social action dimension of speech;18
b) Karl Brugmanns (18491919) essay on Indo-European pronoun systems
(cf. 1904) had provided him with a theory of demonstrative deixis as a framework
to analyse the speech situation;
c) Alan Gardiners Theory of Speech and Language (1932), which owed its
publication partly to Bhlers encouragement.19 Bhler praised it as the most interesting attempt to carry out a project similar to his own but adds the reservation
that Gardiner strives exclusively for a situational theory of language, whereas
235
he, Bhler, wants to account also for types of language use that are removed from
the situation (Bhler 1990: 28; 1934: 3435).20
Supported by these three authorities (as well as by the parole-oriented reading
of Saussures Cours), the organon-model appears in the Sprachtheorie as the
first of four axioms; it postulates three semantic functions (semantische Funktionen) for every instance of a linguistic sign (Sprachzeichen):
Es ist Symbol kraft seiner Zuordnung zu Gegenstnden und Sachverhalten, Symptom
(Anzeichen, Indicium) kraft seiner Abhngigkeit vom Sender, dessen Innerlichkeit es
ausdrckt, und Signal kraft seines Appells an den Hrer, dessen ueres und inneres
Verhalten es steuert wie andere Verkehrszeichen. (Bhler 1934: 28).
It is a symbol by virtue of its co-ordination to objects and states of affairs, a symptom
(Anzeichen, indicium: index) by virtue of its dependence on the sender, whose inner
state it expresses, and a signal by virtue of its appeal to the hearer, whose inner and outer
behaviour it directs as do other communicative signs. (Bhler 1990: 35).
It is in this version that the organon-model has entered pragmatic-functionalist

textbooks as a well-known historical reference (see above). However, its presentation in the Sprachtheorie as the first axiom (possibly meant to underline its preeminence) is less convincing than its introduction in the Axiomatics essay from a
year earlier. In that essay it appears as the last axiom and forms the conclusion of
the previous argumentation.21 In the Sprachtheorie, the model is followed first by
illustrative comments on various styles and registers in which either of the three
functional dimensions appears dominant (e.g. poetry, rhetoric, scientific discourse)
and a hint that its main thesis will be verified when all three books [on language]
that the organon model requires have been written (Bhler 1990: 39; 1934: 33).
This hint underlines the point of the subtitle of Sprachtheorie, The Representational Function of Language (Die Darstellungsfunktion der Sprache): this work
does not contain the complete Theory of Language whose necessity is implied
by the organon model; it still focuses mainly on the one function that had been the
object of traditional language philosophy and semantics.
This discrepancy, although openly announced, is regrettable from a pragmatic
point of view, for historical reasons (as the three promised books were never completed: The Crisis of Psychology and Expression Theory contain valuable linguistic
insights but have a much wider, psychologically defined remit) as well as in theoretical respects. As a largely programmatic formulation, the organon model lacks an
explication by Bhler that would have clarified its methodological implications. In
the Sprachtheorie itself, the model is only intermittently and cursorily referred to
and does not provide a systematic structuring principle of the book. The following
remarks therefore highlight aspects of the Sprachtheorie that can be most closely
linked to the three functions model, without claiming to give a systematic overview of Bhlers opus magnum.
Within the Axiomatics-part of the Sprachtheorie, the second axiom is the signcharacter of language, which Bhler takes over not just from Saussure but from
236
the sign-theoretical tradition going back to the medieval scholastic debates on representation and which he relates in a bold move to the contemporary phonemephone distinction. Bhlers formula for the principle of the semantic character of
signs is abstractive relevance (Bhler 1990: 4056; 1934: 3348). Building on
the organon- and sign-axioms, he attempts in axiom III to achieve an integration of
functional and systemic language aspects by way of a cross-classification based on
the parole-langue and ergon-energeia distinctions by Saussure and Humboldt.
This yields four perspectives on language as an object of scientific investigation,
i.e. language as (1) speech action (Sprechhandlung), (2) language work
(Sprachwerk), (3) speech act (Sprechakt) and language structure (Sprachgebilde) (Bhler 1990: 5780; 1934: 4869).
This schema has been criticised for being terminologically and methodologically inconsistent.22 Its English translation in the 1990 edition adds further problems as it suggests false friends such as Sprechakt and speech act. Bhlers
understanding of this aspect is far removed from Austinian, Searlean or Gricean
concepts; rather, it rests on the philosophical notion of sense-conferring acts
(sinnverleihende Akte), which Bhler had adopted from Husserl. Bhler praises
Husserls Logische Untersuchungen for highlighting sense-conferring acts as a
necessary object of language theory but he criticises the implication of an exclusively subject-centred perspective on meaning constitution, against which he insists
on the inter-subjective side of all use of signs (1990: 7374; 1934: 6365; see also
Nerlich and Clarke 1996: 194198, 211215).
The most interesting aspect of Bhlers schema from a pragmatic perspective is
the category of the speech action, which encapsulates the insight that all concrete speech is in vital union with the rest of a persons meaningful behaviour; it is
among actions and is itself an action (Bhler 1990: 61; 1934: 52). In the discussion in the Sprachtheorie Bhler gives only a relatively brief illustration by way
of famous quotations that embody a whole situation (e.g. Caesars alea jacta est)
and refers to developmental psychology; he also highlights the field-character of
every speech action and the importance of the act(ion) history, which includes the
language acquisition process (1990: 6566; 1934: 56). In the Axiomatics essay,
however, Bhler gives a more detailed explanation, drawing strongly on the action
theory developed by the fellow Wrzburg School psychologist Abraham Anton
Grnbaum (18851932), who had given a paper on Language as action (Sprache
als Handlung) at a workshop organised by Bhler at a psychology congress in
Hamburg in 1931 (Bhler 1933a: 5051). For Grnbaum, every action is a dynamic system of organism and environment (System: Organismus Umwelt); its
constitutive aspects are the agentive organisms need (Bedrfnis) and intention (Intention), the active completion (Erledigung) of the action and the agents reaction
to the completion (Erledigungsreaktion) (Grnbaum 1932: 165173). For the application to the speech action, Bhler re-labels the reaction to the completion as
fulfilment (Erfllung) and describes it further as consisting of the speakers con-
237
scious realisation that the process of meaning constitution has been concluded
(Bhler 1933a: 5051).23
In the Sprachtheorie, Grnbaums theory is not explicitly mentioned but
Bhler uses Grnbaums terminology in some parts and refers for further reading
to his own Axiomatics essay as well as to disputes among the schools of Gestalt
theory, where he takes a polemical stance against the supposed monism of the
Berlin School of Gestalt psychology around Max Wertheimer, Wolfgang Khler
and Kurt Koffka (Bhler 1990: 66; 1934: 56). In this allusion-laden discussion the
contours of Bhlers concept of speech as action are sometimes barely discernible.
There is just one hint that the topic is taken up again in chapter 10, which deals
with the integration of speaking into other meaningful behaviour, and more specifically into different types of surrounding fields (Umfelder) (Bhler 1990: 61;
1934: 52). This distance or detour of six chapters is motivated by Bhlers more
immediate focus on the concept of language fields, which overlaps with the notion of surrounding fields, thus leading to a potentially confusing double-terminology.
Within the Axiomatics-part of Sprachtheorie, the field concept comes into
focus in the fourth axiom, which deals with the syntax-semantics interface under
the heading Word and Sentence: the S[ymbol]-F[ield]-system of the type language (Bhler 1990: 8190; 1934: 6978), but the concept also dominates the remainder of the monograph. It serves Bhler to gain an overview of language phenomena that ranges from phonology and morphology via syntax and case theory to
text theory and stylistics. The principle unifying this theoretical and methodological tour de force is the distinction between, on the one hand, those aspects of linguistic signs that receive their precise meaning in the deictic field (Zeigfeld), i.e.
the spatio-temporal surroundings, the relationship of speaker and listener, their
shared knowledge and action goals, and on the other hand, those aspects of linguistic signs that are fully integrated in syntagms and are detachable from the immediate situation, thus forming a symbol field (Symbolfeld).24 As a programmatic outline, this two-field theory provides, as Garvin (1994) has pointed out,
an alternative to any rigid syntax-semantics-pragmatics distinction and has
proved to be a constant source of fresh insights. In terms of a theoretical focus,
however, Bhlers wide-ranging discussion covers too many data and theoretical
aspects and the important insights from the organon model and the speech action concept are taken up only intermittently in the field-theoretical chapters. In
the following section, we highlight these pragmatically relevant theory aspects
without claiming to provide a comprehensive reconstruction of Bhlers general
field-theoretical framework.
238
3.2.
Linguistic deixis and action
The two main concepts in the Sprachtheorie that are of relevance for linguistic
pragmatics are those of the deictic field and the surrounding fields, i.e. the context in which speech is meaningful as social action. Terminologically, both categories are evidently strongly integrated in Bhlers overarching field-theory,
but conceptually they also depend on the three-functions model. In both cases, the
multi-functionality of language and the fundamental importance of the speakerhearer relationship, which derive from the organon model, are constitutive of
Bhlers argumentation.
The origin of linguistic indexicality/deixis lies in the I-now-here-position of
the speaker, which serves as the starting point for spatial, temporal and social
orientation and co-ordination between the communication partners (Bhler 1990:
117136; 1934: 102120). The interlocutors shared horizon can then be extended
through the quasi-figurative use of the co-ordinates of the demonstratio ad oculos
(visual [or other sensory] demonstration) for the purpose of imaginationoriented deixis. When referring to objects that are not immediately given in the
discourse situation, speaker and hearer must either project them linguistically into
that situational context (e.g. by putting a problem before themselves) or, vice
versa, project their own situation co-ordinates onto an imagined context, or use a
mixture of both. Such imagination-oriented deixis is present already in everyday
language use and marks the first step in the gradual emancipation of language from
the immediate situational context (Bhler 1990: 158166; 1934: 140148). The
highest developmental stage of deictic language use is reached when indexical
language signs are used anaphorically and cataphorically as joints of speech, e.g.
as relative pronouns, text-deictic prepositions, prepositional adverbs or conjunctions (Bhler 1990: 438452; 1934: 385397). Here, the symbol-field is reconstructed along deictic co-ordinates; at the same time, deictic signs are used to
demonstrate relationships within a representational context.25
The second main pragmatically interesting discussion in the Sprachtheorie,
which Bhler had already flagged up in the context of the Axiomatics section when
dealing with the speech action, is chapter 10. Its title provides an additional typology of surrounding fields: The sympractical, the symphysical and the synsemantic field of Language Signs (1990: 175; 1934: 154). Again, as in the case of
the deixis-symbolising continuum, Bhler conceptualises the field-series as a cline
of abstraction. Linguistic signs can be tied closely to a specific situation or specific
objects (= the sympractical and symphysical fields) and so remain situation-bound.
The more strongly they are integrated into the linguistic con-text (= the synsemantic field) however, the more liberated they are from the specific circumstances
of the speech situation. In some genres, e.g. algebraic formula systems, the liberation from concrete, situation-related content is almost taken to the maximum (cf.
chapters 11, 25 of Sprachtheorie). The overlap between the synsemantic field
239
and the symbol field (Bhler 1990: 179189; 1934: 159168) blurs the boundaries
between the two field-theoretical concepts to some extent,26 which may be indicative of Bhlers difficulty to combine his focus on functional aspects of language with the perceived necessity of integrating these insights into the predominant contemporary tendencies of explaining language mainly as a system of
representation.
In the context of the surrounding fields, Bhler comes closest to a pragmatic
interpretation of language data in the passages dealing with the sympractical
field, which presuppose his three-functions axiom and Grnbaums action field
theory (see above). The sympractical field includes what would be called in modern terminology the situational context (as opposed to co-text), which communication partners need to know (and share) in order to be able to disambiguate
situation-bound utterances such as short commands, requests, greetings etc. Such
utterances had been viewed traditionally as deficient, on account of not being
syntactically well-formed sentences (Bhler 1990: 176179, 187189; 1934:
156159, 166167).27 Through examples such as that of a patron in a coffee house
asking for a particular type of coffee by way of uttering something like an espresso, please28 and of a passenger in a tram car choosing between types of
tickets, Bhler explains that an ellipsis interpretation is mistaken. In such cases,
the speaker does not have to produce semantically explicit or grammatically wellformed utterances, because the interlocutor, on account of shared knowledge of the
situation (the Umfeld), understands motives and plans for the action (in modern
terminology: the action schema or script) and uses the language sign only as a
diacritic. The situational context provides a sufficient fulfilment of all meaning
functions in the action field: the utterance is not elliptical but rather, in Bhlers
terminology, empractical (1990: 177179; 1934: 155158).
Generalizing from Bhlers hints, one can argue that language use is empractical in principle, insofar as it is always aimed at resolving ambiguities of intersubjective co-operation and based on some degree of shared background knowledge
(or at least, background assumptions). Whilst Bhler never formulated any communicative co-operation or relevance principles in the Sprachtheorie, his notion of the sympractical field as the basic action-structure in which language functions operate can be viewed at least as compatible with more recent pragmatic and
relevance-theoretical approaches.29
4.
Vilm Mathesius and the Prague Linguistic Circle
4.1.
Introduction
From 1930 onward, Bhler had scattered connections with members of the Prague
Linguistic Circle, the avant-garde group of linguists that propagated a view of language as a functional system with structural laws (Prague Linguistic Circle
240
[1929] 1983: 7779). Founded by the Czech linguist and scholar of English Vilm
Mathesius (18821945) in 1926, the Circle united under Mathesiuss presidency
some of his Czechoslovak colleagues and linguists of other nationalities who lived
in Prague or nearby, notably a number of Russian and Ukrainian migrs. Whereas
the functional point of view was the original contribution of Mathesius himself,
Roman Jakobson, the Russian vice-president of the Circle, added a structuralist,
much more system-oriented approach. Not surprisingly, Mathesiuss and Jakobsons legacies to pragmatics vary. The following section will detail Mathesiuss
functionalist findings that are relevant for pragmatics; in section 5, we will turn to
foundations of pragmatics deriving from Jakobsons work.
4.2.
Mathesiuss functional view of language
Since his earliest linguistic writings in the first decade of the century, Mathesius
had argued for a view of language as a means to an end.30 In the years of his university education in Prague, his general methodological outlook was formed by the
call for concreteness from the Czech philosopher Toms Masaryk (18501937),
who coupled this appeal with the desirability of democratising scientific and philosophical education (Masaryk 1887). Science had to give priority to the study of
close and accessible phenomena, including their comparative analysis; this
would prepare for historical research at a later stage.31 Accordingly, Mathesius
sought his linguistic inspirations in the works of opponents of Neogrammarian
dogmatism, such as Sweet, Otto Jespersen (18601943), von der Gabelentz and
Wegener (Mathesius [1936] 1966: 137), and he became an advocate of synchronic
language study.
The ideal of concreteness took shape in the emphasis which Mathesius laid
on the inherent synchronic variability of linguistic phenomena among individuals
of a language community as well as in the speech of a single individual; he spoke
of the static oscillation in speech and called this aspect of language its potentiality (Mathesius [1911] 1983a: 3). He stressed that the seeming homogeneity of
language is a product of the method of analysis and that without the phonetic phenomenon of synchronic oscillation, languages could not change (Mathesius
[1911] 1983a: 4, 32). These claims sound very modern and they were indeed approvingly cited in the influential paper on variability and language change by
Weinreich, Labov and Herzog (1968: 167169), although these authors also criticised Mathesius for not having pointed to the sociolinguistic regularities that are
inherent in variation.
In his ideas on synchronic oscillation in the domain of semantics, Mathesius
was at first strongly influenced by the stylistics of the Geneva linguist Charles
Bally (18651947) as well as by the idealistic approach of the German scholar of
Romance languages Karl Vossler (18721949). This stance implied an emphatic
interest in the expression of the speaking individual rather than in the social char-
241
acter of language, which he saw as having been overstated by nineteenth-century

linguistics (Mathesius [1911] 1983a: 2931, 3536; [1926] 1983b: 5253). In his
later work, Mathesius redressed the balance to some extent by way of an acknowledgement of the dominance, at least for language change, of the communicative
function of language vis--vis the original expressive function (Mathesius
[1926] 1983b: 5455). However, on account of his continued interest in stylistics
he kept a predilection for the expressive function and defined his functional
point of view as one that takes the meaning or function as its starting point and
tries to find out by which means it is expressed (Mathesius [1926] 1983b: 5658;
also [1929] 1983c: 122125). Mathesius thus came to neglect some of the necessary focus on the hearer as the key player in the constitution of meaning, something
which Wegener and Bhler had already reached in their work. This has to be
viewed as a general restriction of Praguian functional linguistics in its foundational
role for pragmatics.
Mathesiuss emphasis on the expressive function is understandable in the light
of his long-standing interest in the comparative study of word order patterns,
which was first kindled by the work of the French scholar of Ancient Greek Henri
Weil (18181909) (Weil [1844] 1887). Confronted with differences in word order
between Czech and English, Mathesius saw languages of different character.
Where a speaker of Czech usually filled the slot of the grammatical subject with the
agent of the action expressed by the verb, an English speaker appeared to prefer the
theme (Thema) of the utterance for this role. The Czech speaker therefore was
relatively free to order his sentence parts according to their contextual status of
theme or nucleus (rheme). For the English speaker to have a similar possibility readily at hand, the language had developed grammatical means for alternative choices of the grammatical subject, i.e. several types of passive constructions, which could be observed to make out a considerable proportion of English
verbal constructions (Mathesius 1929: 202203; [1929] 1983c: 126128, 133
134; 1930). The character of a language was thought to explain the links between
such descriptive facts of different types.
Mathesiuss characterology (or: language typology) thus relied on what he
called the actual articulation32 of utterances, which was to be distinguished from
its logico-grammatical articulation. The notion of actual articulation pointed to
the inherent situatedness of the structure in a specified context, and to its concrete quality; in short: it was a pragmatic structuring. Specifically, it consisted of
a basis or theme on the one hand and a nucleus, enunciation, or rheme
on the other.33 In a formulation from Mathesius (1975a: 81): The element about
which something is stated may be said to be the basis of the utterance or the theme,
and what is stated about the basis is the nucleus of the utterance or the rheme.
To an extent, Mathesiuss proposal of a functional sentence perspective resembles Wegeners ideas on the bipartition of utterances in exposition and
predicate that we discussed in section 2. Still, we note the less pronounced com-
242
municative emphasis with Mathesius: the hearer is absent in his formulation, and
even the speaker is no more than implied. Furthermore, Mathesiuss insight in the
role of the situation is much less developed. In any case, there appear to be no references by Mathesius to Wegeners theories on this subject, at least not from the
period in which Mathesius developed his notion of functional sentence structure
(the late 1920s). We have to decide that the relation is probably not a direct one.
It is certain however that Mathesiuss notions found favour in many quarters of
(pragmatic) linguistics and have become, under some of their original names or as
topic and comment, classics of pragmatic analysis (see, e.g., Lyons 1977:
501512; Levinson 1983: 8889).34
5.
Roman Jakobson
5.1.
Jakobsons model of linguistic functions
The Russian general linguist and scholar of Slavonic languages Roman Jakobson
(18961982) was intrigued by the seemingly autonomous functionality (teleology, self-movement) of language change all his life (Liberman 1987). For a
long time, Jakobson hoped to be able to account for it with the help of a Hegelian
logic of system change (Toman 1995: 171176). But in the milieu of the Prague
Linguistic Circle, he gradually came to recognise that the stages of a change in
progress had in fact to be conceived as different simultaneous facts, or styles,
within the language system. Especially after Jakobsons emigration to the United
States in 1941, his teleological point of view shifted to a focus on the less mysterious functions in a means-ends model of a synchronic system of systems (Jakobson 1963: 107).
In his new homeland, Jakobson found the study of language to have been, as he
remarked at a conference of anthropologists and linguists in 1952, strongly bulwarked by the impressive achievement of two conjoined disciplines the mathematical theory of communication and information theory (Jakobson 1953: 12).
This led him to discuss the factors of the communication process that were treated,
e.g., in the well-known publication by Shannon and Weaver (1949). In his paper,
Jakobson enumerates the four factors connected to the message in the information
theorists terminology (sender, receiver, topic of the message, code) and points to
the variability of their relative importance for a message. Apart from being associated with the factor of topic (information), a message may also be oriented toward the sender or the receiver, leading to a function of the message different
from a purely informative one, or, as the case may be, a hierarchical bundle of several functions (Jakobson 1953: 13). These three basic function varieties can be
considered equivalent to Bhlers (1990, 1934) functions of representation, expression and appeal, in the three-functions model that Jakobson subsequently
243
refers to as the traditional model of language (Jakobson 1960: 355; [1976] 1985:
115).35 It should be kept in mind however that Jakobson views the language functions as systemic facts throughout (facts of a system of systems). This distinguishes his theory from Bhlers approach (recall Bhlers championing of parole)
and even more from Wegeners; indeed, we are not aware of any references to Wegener in Jakobsons work.
Jakobson had come to linguistics via the study of Russian verse and of poetics
in general (cf. Rudy 1978; Toman 1995; Waugh and Rudy 1998: 22622265). Together with his compatriot, the literary scholar Jurij Tynjanov (18941943), he had
made a case for a parallel development of language, literature and other cultural
systems by structural necessity; there was to be no separation of the interconnected
codes of language and literature (Tynjanov and Jakobson ([1928] 1972). In a section of the 1929 Theses of the Prague Linguistic Circle, Jakobson and the Czech
literary scholar Jan Mukarovsky (18911975)36 had sought to contrast poetic language with communicative language. They had stated that poetic expression is
directed towards the way of expression itself and that the linguistic means of different levels, which in communicative speech tend to become automatised, in
poetic utterances tend to become foregrounded themselves (Prague Linguistic
Circle [1929] 1983: 94; cf. also 89). It is this idea that Jakobson takes up when he,
in his paper from 1953, adds a fourth, poetic function of language to Bhlers set
of three. For a message that is partly loosened from its referential meaning by
being oriented towards its own self, towards the palpability of its signs, the
poetic function of language is at work, be it as the predominant function of a literary fragment or as a subordinate, additional function of an otherwise prosaic
message (Jakobson 1953: 14).
Jakobson mentions also a fifth possibility: an orientation of the message toward
the factor of code, interesting because here the code (i.e., the langue) is not just an
abstraction but interacts with language use. The orientational set (explained as
Einstellung) for the code is witnessed in non-specialist discourse about such linguistic subjects as clarification of code elements, code switching, bilingualism,
areal diffusion, language variation and language change (Jakobson 1953: 1418).
This function is described with more clarity and termed the metalingual function
in a publication based on Jakobsons presidential address at a meeting of the Linguistic Society of America in 1956 (cf. Jakobson 1985: 116121). A sixth function
is mentioned in that address as well: Jakobson analyses anthropologist Malinowskis phatic communion function37 as being attached to messages with a set for the
factor of contact, or channel, as when interlocutors, without much information
transfer, check or revive the physical communication channel, or keep it going
(Jakobson 1985: 115).38
Jakobson (1960: 353357) sums up the resulting model of six language functions (with some terminological changes): the referential (denotative, cognitive)
function, the emotive (expressive) function, the conative39 function, the poetic
244
function, the metalingual (or glossing) function and the phatic function. They
are explained by the orientation of the message toward one of the six constitutive
factors of speech events, to wit the context referred to (the referent), the addresser,
the addressee, the message, the code, and the contact channel. It is in this version
that Jakobsons model of language functions has been widely cited by linguists of
several persuasions, including theoreticians of linguistic pragmatics (see, e.g.,
Hymes [1962] 1968: 99, 110124; Hymes 1983: 334; Lyons 1977: 5256; Levinson 1983: 41; Schiffrin 1994: 3233; Duranti 1997: 284287). It is clear, however,
that the model is focused on language as an undivided phenomenon, in which, just
as in Bhlers approach, the many facets of a pragmatic nature remain inclusive,
and therefore partly implicit. In order to gain an unimpeded view of pragmatic
facts, a systematic linguistic pragmatics had still to be drawn out.
5.2.
Jakobsons treatment of deixis
Among the further aspects of Jakobsons functionalism that are relevant for the development of linguistic pragmatics, his view of deictic phenomena is the most important. In a publication of 1957, which goes back to a paper delivered in Geneva
in 1950 (cf. Jakobson 1950), he treats the differentia specifica of indexical items
such as the personal pronouns I and you (Jakobson [1957] 1971). Like Bhler, Jakobson points to the fact that their precise meaning in context needs to take into account aspects of the situation, always unique, in which they happen to be used;
adopting Otto Jespersens (18601943) (1922: 123) term, he calls them shifters
(Jakobson 1971: 131). Jakobsons main interest however is focused on the question
of the status of these shifters as elements of the linguistic code, and he insists that
their symbolic meaning (for I, the addresser of the pronoun instance, and for you,
the addressee) should not be neglected either. Using Charles Sanders Peirces
(18391914) classification of sign modes and specifically following its interpretation by the American mathematician and philosopher Arthur Burks (19152008),
Jakobson argues in favour of according them the status of indexical symbols.40
First and second person pronouns in use are associated with the object they represent by an existential relation (i.e., pronoun token and object exist in the same
situation), therefore they indeed function as indices, just like acts of pointing
with the index finger would. However, on account of their conventional relation
with the role of speaker or hearer, which in different language codes holds for different forms (Latin ego, German ich, Russian ja), they partake just as well of the
symbolic mode of signs (Jakobson 1971: 132).41
Whereas the sober-minded Jespersen (1922: 123), in reaction to a remark by
the German idealist philosopher of subjectivity Johann Gottlieb Fichte
(17621814), had belittled attempts to derive deep philosophical conclusions from
the little linguistic trick involved in using the pronoun I correctly, Jakobson goes
on to develop its theoretical significance further. Considering the constitutive fac-
245
tors of speech situations, he notices for the shifters an overlapping of the factors
Code and Message: their general meaning, which is part of the Code, contains a
reference to the very Message of which the shifter in an instance of its use happens
to form a part (Jakobson 1971: 132). Jakobson explains this possibility by asserting
that both Code and Message can in principle refer as well as be an object of reference; in the specific case of the shifters, the Code has the function of referring,
while the Message is being referred to.42
This explanation takes up a similar remark by the French linguist mile Benveniste (19021976) in an article about pronouns that was published a year earlier
in the Festschrift For Roman Jakobson (Benveniste [1956] 1966a: 252). Benveniste, likewise anxious to find an aspect of the Saussurean langue (the Code) within
the typical discourse (Message) reality of first and second person pronouns, goes
further than Jakobson in postulating that it is precisely the use of elements symbolizing the roles of utterer-conceptualiser (nonciateur) and allocutee of a discourse event which leads to language users consciousness of their role as subjects. Those symbolic features of the language system then get parcelled out into
psychological reality. Benveniste also mentions that the process is strengthened by
various verbal forms that align with the perspective of the utterer of the discourse
(Benveniste 1966a: 254255; also [1958] 1966b: 258263). In his 1957 article,
Jakobson indeed uses the theory of the shifters to set up a new classification of verbal categories, in particular those of Russian. Like first and second person pronouns, verbal shifter categories (several forms in the paradigms of person, mood,
tense and evidentiality) have an encoded meaning in which the speech event (the
message) and/or one of its participants play a decisive role (Jakobson 1971:
133147).43
The relevance of the idea of the interweaving of Code (langue) and Message
(instance of discourse) to linguistic pragmatics becomes evident when the phenomenon of the shifters is described as a fact about speech events rather than the
linguistic code. It can be said then that by means of various indexical elements
of the language, speech events often encode, or index, important components of
the communitys social life. It was the American anthropologist Dell Hymes
(19272009) who realised that on this basis, Jakobsons speech functions could be
extended and seen to be differently organised for different communities, opening
up an additional realm of structure (Hymes 1983: 335). This was to result in the
ethnography of speaking and more broadly, the ethnography of communication
(Hymes [1962] 1968: 110129; 1964; 1972).
6.
Conclusion
Philipp Wegener, Karl Bhler, Vilm Mathesius and Roman Jakobson all of them
characteristically modernist theorists were fully aware of the pervasive relevance of language in countless domains of human life and culture. They projected
246
the pragmatic aspects which they had identified, such as the multifunctionality of
language, the functional structure of the sentence and the nature of deixis, onto a
rich picture of language that did not separate semantic from pragmatic phenomena
and often contained no fixed answer to the question as to what is language itself
and what is its use. Seen from a present-day perspective, their works may appear to
be lacking in systematic treatment of pragmatic facts. By the same token, however,
they have had and continue to have potential not for just one but many branches
of linguistic pragmatics. Wegeners ideas for the linguistic analysis of utterances
together with their embedding situations may still be used to further loosen the grip
of the received wisdom about the structure of linguistic utterances as phenomenal
entities per se. Similarly, Bhlers and Jakobsons theories of the functions of language, the associated models of the communication situation and the deictic signs/
shifters that derive their specific meaning from it offer the chance to review the
immunity of the theory of the language system to results of the analysis of language use.
Notes
1. Outside the discipline of linguistics, Bhlers model has been taken up by Jrgen Habermas, who in his theory of communicative action (cf. 1981) uses the three functional dimensions as reference points for a model of social communication. According to Habermas, every speaker, when making a meaningful utterance, intends it to be understood by
the hearer as legitimate (within the social context), truthful (i.e. as an expression of the
speakers own convictions) and true (in relation to objective states of affairs). Even
though these assumptions are often disappointed in actual communication, they act as
regulating norms for meaningful social exchange (Habermas 1981, vol. 1: 372377,
412). See further the article by Cooke in this volume.
2. For information on Wegener and his work, see especially Knobloch (1991); also Juchem
(1984, 1986), Knobloch (1988: 292297), Nerlich (1990: 153192; 1992: 8187), Graffi
(1991: 8789), Elffers (1993), Nerlich and Clarke (1996: 177183), Morpurgo Davies
(1998: 318320), Grimm-Vogel (1998), Tenchini (2008).
3. An English translation of Wegeners 1885 book appeared as part of Abse (1971); note
that Knobloch (1991: xliv) expresses a caveat about details of the translation.
4. In his publication from 1885, Wegener uses the term logical predicate; in later works he
switches to psychological predicate and statement (Mitteilung).
5. Knobloch (1991: xvi, xxxii) mentions Lazarus ([18561857] 1884) as a possible source
of Wegeners ideas in this area. About Moritz Lazarus (18241903), see Knobloch (1988:
268273, 411422) and Nerlich and Clarke (1996: 165168).
6. For details, see Hlzer (1987), Knobloch (1988: 296297; 1991: xviii-xix), Nerlich
(1990: 181185; 1992: 8687).
7. This is the opinion also of Knobloch (1988: 297) and Nerlich (1992: 87).
8. See the references to Wegeners work in Malinowski ([1923] 1946: 297), Gardiner (1932:
passim) and Halliday (1978: 109), particularly in relation to the notion of situation.
247
9. See Ungeheuer (1987, 1990) and Juchem (1984, 1986).

10. For details see Knobloch (1991: xxxix-xli).
11. Quotations from Bhlers works are referenced, where possible, to both the original
German publications and their English translation, i.e., in the case of the Sprachtheorie,
1934 for the German volume, 1990 for its translation.
12. See e.g. Schlieben-Lange (1979: 13); Levinson (1983: 41, 61); Brown and Yule (1983:
1): Coupland (2007: 12).
13. After the Sprachtheorie (1934), Bhler published several other linguistic articles up
until 1938 (Bhler 1935, 1936a, 1936b, 1938). Immediately after the occupation of
Vienna by German troops, Bhler was imprisoned by the Gestapo. His wife and academic collaborator, Charlotte Bhler, managed to effect his release and to organise their
emigration to the USA, but the continuation of their careers proved difficult: they only
gained posts as visiting or assistant professors and did not return to Europe after the war
(see Garvin 1966, Ch. Bhler 1984, Musolff 1997). The 1960 article Von den Sinnfunktionen der Sprachgebilde (Of the Semantic functions of linguistic structures) is an
excerpt from Sprachtheorie and of dubious editorial status (e.g. it cites the Organonmodell as Organmodell).
14 Husserl (19001901); Wundt (1900); Marty (1908). Bhler reviewed Martys work with
specific emphasis on the functional typology, see Bhler (1909a: 964946).
15. For Bhlers experiments on sentence comprehension in the context of Thought Psychology, see Bhler (1908), Wundt (1908), Wettersten (1988).
16. For the changes in the functions-terminology and the status and ordering of the
axioms, see Wunderlich (1969); Kamp (1977: passim); Beck (1980: 169192); Busse
(1975: 213215, 222229); Camhy (1980a: 9294); Swiggers (1981: 5455); Innis
(1982: 310); Eschbach (1984b: 9193); Henzler (1987: 354356); Knobloch (1988:
424430); Graumann (1988); Musolff (1990: 2631); Vonk (1992: passim); Elffers
(2005; 2008: 2124).
17. For the contemporary context of the Saussure-reception in relation to Bhler see
Scheerer (1980: 3640), Koerner (1984), Musolff (1990: 4043) and Ehlich (1996).
18. Details of what Bhler had learnt from Wegener can be found in Bhler (1909b:
119123).
19. See Gardiner (1932: viii).
20. For detailed discussion of these and other pragmatic-functional influences cf. Graumann (1984), Herrmann (1984), Knobloch and Schallenberger (1993), Musolff (1993)
and Nerlich and Clarke (1996: 182183, 226236, 341343).
21. See Bhler (1933b: 90); in Strkers edition (1976: 116117).
22. See Laziczius (1939); Lohmann (19421943); Camhy (1980b); Ortner (1986).
23. For detailed discussion of Grnbaums influence on Bhler see Knobloch and Schallenberger (1993: 8590).
24. For Bhlers field-theoretical approach (as opposed to the contemporary notion of
semantic fields) cf. Wolf (1932); Heger (1984); Musolff (1990: 61120); Garvin
(1994).
25. For Bhlers influence on modern pragmatically oriented studies of deixis see Hrmann
(1978: 394424); Weissenborn (1988); Conte et al. (1989); Innis (1992: 556557);
Ehlich (1996: 959).
26. Garvin (1994: 6064) interprets the relationship as a superordination of the more general surrounding field over the two subcategories of symbolic and deictic field; he
248
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
bases this reading on an excellent comparison with the contemporary gestalt-psychological theories, which Bhler himself highlighted (1990: 175176, 1934: 154155) as
well as with early semiotics (Charles Morris). However, in the Sprachtheorie this classification is not explicated nor consistently followed it constitutes a further theoretical
(and highly plausible) development, the implications of which are still to be worked out
in detail.
For critical assessments see Klein (1984), Ortner (1983).
Living in Vienna at the time, Bhler of course chose an appropriate local variety (einen
schwarzen a black one).
But compare Nerlich and Clarkes (1996: 236237) critical comment that Bhler, despite being one of the most important figures among language researchers engaged in
pragmatics avant la lettre, never crossed the last hurdle on his way to a fully developed theory of linguistic pragmatics, especially on account of his [clinging] too much
to his three functions of language, overlooking to some extent the polyfunctionality of
speech acts, rediscovered by Wittgenstein.
Toman (1995: 7779). About Mathesiuss work, see further Trnka ([1946] 1966), Vachek (1982), Danes (1987, 2003), Fronek (1988), Graffi (1991: 179183, 249250),
Toman (1995: 7186, 97101 and passim), Leska (1995), Nekula (1999).
See Toman (1995: 8486, 97101), also Fronek (1988).
This notion features in the original title of the Czech article Mathesius ([1939] 1975b);
see the translators remark in Mathesius (1975b: 479). It came to be translated as information-bearing structure of the sentence, shortened to information structure and
alternating with functional sentence perspective (FSP).
The terminology differs; see Mathesius ([1929] 1983c: 126128; 1929: 202; 1930: 58).
The term rheme came into use only later, through the writings of the Brno scholar of
English Jan Firbas (19212000). See the explanation by Vachek in Mathesius (1975a:
185 n71), further Vachek (1982: 123124) and Danes (1987: 2326; 2003: 4041).
The Prague version of the information structure of the utterance has remained a distinctive one. See a short description in Hajicov (1994).
Jakobson referred to Bhlers three functions already in 1932 (Jakobson [1932] 1966:
2728). In its Theses of again some years earlier, the Prague Linguistic Circle had distinguished, under the heading On the functions of langugage, intellectual speech
from two types of emotional speech and had mentioned the different hierarchy of
functions for each utterance (Prague Linguistic Circle [1929] 1983: 8889). Jakobson
probably wrote (part of) this section (nr. 3a) of the Theses (cf. Toman 1995: 289 n13).
For the attribution to Jakobson and Mukarovsky, see again Toman (1995: 289 n13).
See Malinowski ([1923] 1946: 315316). Cf. Nekula and Ehlers (1996: 190, 192) on the
at first marginal place of the phatic function of language in Jakobsons model.
On Jakobsons model of language functions, see Holenstein (1976: 153164); Waugh
(1976: 2526); Waugh and Monville-Burston (1990: 1516); Nekula and Ehlers (1996);
Nerlich and Clarke (1996: 284286); Waugh and Rudy (1998: 22602261).
Conative: with the notion of striving to effect some act by the addressee; cf. Jakobson ([1932] 1966: 22, 27).
See Burks (1949). Among the signs having the property of indexicality, Burks views the
indexical symbol as the most fundamental type; a pure index cannot in effect point
to a well-defined object and is therefore a less fundamental type of sign (Burks 1949:
678688).
249
41. Jakobsons (1971: 132) statement to the effect that Bhler acknowledged the indexical
side of personal pronouns only seems not fully justified. It is when Bhler introduces the
indexical field (Zeigfeld) that he contrasts indexical and appellative words as to the
field in which, specifically, their meaning fulfilment (Bedeutungsprzision, Bedeutungserfllung, see above, 2.2) is located: in the indexical field and the symbol-field of
an utterance, respectively (Bhler 1990: 9194; 1934: 7881). This does not imply that
personal pronouns would not have a symbolic meaning aspect to begin with, such as
the sender of the word in question (Sender des aktuellen Wortes) (Bhler 1990: 91;
1934: 78); indeed, the notion of meaning fulfilment presupposes the opposite idea.
42. Considering the three other possibilities of combining the choices of Code/Message and
referring/referred-to item, Jakobson theorises further in the following way. When the
functions are turned round and it is rather the Message that refers and the Code that is referred to, we have to do with the elucidation of words and sentences (what would be
named the metalingual function of language a few years later); when a Message refers
to a Message, the result is the well-known phenomenon of reported speech; and lastly,
with Code referring to Code we could describe the general nature of the meaning of
proper names, Jakobson holds (Jakobson 1971: 130131).
43. On Jakobsons notion of shifters, see Holenstein (1976: 158, 162163); Waugh (1976:
2425, 32, 48, 52); Waugh and Monville-Burston (1990: 1719); Nerlich and Clarke
(1996: 282284).
References
Abse, D. Wilfred
1971
Speech and Reason: Language Disorder in Mental Disease. With a Translation
of The Life of Speech by Philipp Wegener. Charlottesville, VA: University
of Virginia Press.
Beck, Gtz
1980
Sprechakte und Sprachfunktionen: Untersuchungen zur Handlungsstruktur
der Sprache und ihren Grenzen. Tbingen: Niemeyer.
Benveniste, mile
1966a
La nature des pronoms. In: mile Benveniste, Problmes de linguistique gnrale, tome 1, 251257. Paris: Gallimard. First published The Hague: Mouton
[1956].
Benveniste, mile
1966b
De la subjectivit dans le langage. In: mile Benveniste, Problmes de linguistique gnrale, tome 1, 258266. Paris: Gallimard. First published Paris: PUF
[1958].
1983
Brugmann, Karl
1904
Die Demonstrativpronomina der indogermanischen Sprachen: Eine bedeutungsgeschichtliche Untersuchung. In: Berichte ber die Verhandlungen der
Kniglichen schsischen Gesellschaft der Wissenschaften zu Leipzig, Philologisch-Historische Classe, No. 6: 3151.
250
Bhler, Charlotte
1984
Karl Bhler: Eine biographische Skizze. Transl. by Achim Eschbach. In:
Achim Eschbach (ed.), Bhler-Studien, Vol. 1, 2530. Frankfurt am Main:
Suhrkamp.
Bhler, Karl
1908
Tatsachen und Probleme zu einer Psychologie der Denkvorgnge: II) ber
Gedankenzusammenhnge. Archiv fr die gesamte Psychologie 12: 123; III)
ber Gedankenerinnerungen. Archiv fr die gesamte Psychologie 12: 2492.
Bhler, Karl
1909a
Review of Anton Marty, Untersuchungen zur Grundlegung der allgemeinen
Grammatik und Sprachphilosophie (1908). Gttingische gelehrte Anzeigen
171: 947979.
Bhler, Karl
1909b
ber das Sprachverstndnis vom Standpunkt der Normalpsychologie aus. In:
Bericht ber den 3. Kongre fr experimentelle Psychologie (Frankfurt 1908),
94130. Leipzig: Wilhelm Engelmann.
Bhler, Karl
1918
Kritische Musterung der neuern Theorien des Satzes. Indogermanisches Jahrbuch 6: 120.
Bhler, Karl
1927
Die Krise der Psychologie. Jena: J.A. Barth.
Bhler, Karl
1931
Phonetik und Phonologie. Travaux du Cercle Linguistique de Prague 4: 2253.
Bhler, Karl
1932
Das Ganze der Sprachtheorie, ihr Aufbau und ihre Teile. In: Gustav Kafka
(ed.), Bericht ber den 12. Kongre der Deutschen Gesellschaft fr Psychologie (Hamburg 1931), 95122. Jena: G. Fischer.
Bhler, Karl
1933a
Ausdruckstheorie: Das System an der Geschichte aufgezeigt. Jena: G. Fischer.
Bhler, Karl
1933b
Die Axiomatik der Sprachwissenschaften. Kant-Studien 38: 1990. (Reprinted
1969 and 1976 under the same title in a truncated version without the introductory chapter, edited by Elisabeth Strker, 25117). Frankfurt am Main: Klostermann.
Bhler, Karl
1934
Sprachtheorie: Die Darstellungsfunktion der Sprache. Jena: G. Fischer.
Bhler, Karl
1935
Forschungen zur Sprachtheorie: Einleitung. Archiv fr die gesamte Psychologie 94: 401412.
Bhler, Karl
1936a
Psychologie der Phoneme. In: Daniel Jones and Dennis B. Fry (eds.), Proceedings of the Second International Congress of Phonetic Sciences, 162169.
Bhler, Karl
1936b
Das Strukturmodell der Sprache. Travaux du Cercle Linguistique de Prague 6:
312.
251
Bhler, Karl
1938
Der dritte Hauptsatz der Sprachtheorie: Anschauung und Begriff im Sprechverkehr. In: Henri Piron and Emile Meyerson (eds.), 11e Congrs International de Psychologie (Paris 1937). Paris: Rapports et Comptes Rendus,
196203.
Bhler, Karl
1960
Von den Sinnfunktionen der Sprachgebilde. In: Richard Wisser (ed.), Sinn und
Sein: Ein philosophisches Symposion; 95112. Tbingen: Niemeyer:
Bhler, Karl
1990
Theory of Language: The Representational Function of Language. Translation
by Donald Fraser Goodwin. Amsterdam: Benjamins.
Burks, Arthur W.
1949
Icon, index, and symbol. Philosophy and Phenomenological Research 9:
673689.
Busse, Winfried
1975
Funktionen und Funktion der Sprache. In: Brigitte Schlieben-Lange (ed.),
Sprachtheorie, 207240. Hamburg: Hoffmann and Campe:
Camhy, Daniela G.
1980a
Karl Bhlers Sprachtheorie. PhD. Diss. Graz: University of Graz.
Camhy, Daniela G.
1980b
Sprechereignis als Handlung: Bhlers Beitrag zu handlungstheoretischen Anstzen in der Sprachphilosophie. In: Rudolf Haller and Wolfgang Grassl (eds.),
Sprache, Logik und Philosophie, 401405. Wien: Hlder-Pichler-Temsky.
Conte, Maria Elisabeth, Jnos Snder Petfi and Emel Szer (eds.)
1989
Text and Discourse Connectedness. Amsterdam: Benjamins.
Coupland, Nikolas
2007
Style: Language, Variation and Identity. Cambridge: Cambridge University
Press.
Danes, Frantisek
1987
On Prague School functionalism in linguistics. In: Ren Dirven and Vilm
Fried (eds.), Functionalism in Linguistics, 338. (Linguistic and Literary
Studies in Eastern Europe 20.). Amsterdam: Benjamins.
Danes, Frantisek
2003
Vilm Mathesius Konzeption der funktionalen Linguistik. In: Marek Nekula
(ed.), Prager Strukturalismus: Methodologische Grundlagen, 3147. (Slavica:
Monographien, Hand-, Lehr- und Wrterbcher 3.). Heidelberg: Universittsverlag Winter.
Duranti, Alessandro
1997
Linguistic Anthropology. (Cambridge Textbooks in Linguistics.) Cambridge:
Ehlich, Konrad
1996
Sprache als System versus Sprache als Handlung. In: Marcelo Dascal, Dietfried Gerhardus, Kuno Lorenz, Georg Meggle (eds.), Sprachphilosophie: Ein
internationales Handbuch zeitgenssischer Forschung, vol. 1, 952963.
(Handbcher zur Sprach- und Kommunikationswissenschaft 7.1). Berlin/New
York: de Gruyter.
252
Elffers, Els
1993
Philipp Wegener as proto-speech act theorist. Linguistics in the Netherlands
1993: 4959. Amsterdam: Benjamins.
Elffers, Els
2005
Reichling and de Groot: Two Dutch reactions to Bhlers organon-model. Historiographia Linguistica 32: 87116.
Elffers, Els
2008
Interjections and the language functions debate. The Henry Sweet Society
Bulletin 50: 1729.
Eschbach, Achim (ed.)
1984a
Bhler-Studien. 2 vols. Frankfurt am Main: Suhrkamp.
Eschbach, Achim
1984b
Karl Bhlers Axiomatik und das Axiomensystem der Zeichentheorie. In: Carl
F. Graumann and Theo Herrmann (eds.), Karl Bhlers Axiomatik: Fnfzig
Jahre Axiomatik der Sprachwissenschaften, 53104. Frankfurt am Main:
Klostermann.
Fronek, Josef
1988 T.G. Masaryk and Prague School linguistic structuralism. In: Josef Novk (ed.), On
Masaryk: Texts in English and German, 8196. (Studien zur sterreichischen
Philosophie, 13). Amsterdam: Rodopi.
Gabelentz, Georg von der
1875
Weiteres zur vergleichenden Syntax: Wort- und Satzstellung. Zeitschrift fr
Vlkerpsychologie und Sprachwissenschaft 8: 129165, 300338.
Gardiner, Sir Alan Henderson
1932
The Theory of Speech and Language. Oxford: Clarendon Press. (2nd ed. with
additions 1951.)
Garvin, Paul L.
1966
Karl Bhlers contributions to the theory of linguistics. The Journal of General
Psychology 75: 212215.
Garvin, Paul L.
1994
Karl Bhlers Field Theory in the Light of the Current Interest in Pragmatics.
In: Svetla Cmejrkov and Frantisek Stcha (eds.), The Syntax of Sentence and
Text: A Festschrift for Frantisek Danes, 5966. Amsterdam: Benjamins.
Graffi, Giorgio
1991
200 Years of Syntax: A Critical Survey. (Studies in the History of the Language
Sciences 98.) Amsterdam: Benjamins.
Graumann, Carl F.
1984
Wundt Mead Bhler: Zur Sozialitt und Sprachlichkeit menschlichen Handelns. In: Carl F. Graumann and Theo Herrmann (eds.), Karl Bhlers Axiomatik: Fnfzig Jahre Axiomatik der Sprachwissenschaften, 217247. Frankfurt
am Main: Klostermann.
Graumann, Carl F.
1988
Aspektmodell und Organonmodell: Die Problematik des Verhltnisses zwischen
Sprachwissenschaft und Psychologie bei Karl Bhler. In: Achim Eschbach
(ed.), Karl Bhlers Theory of Language, 107124. Amsterdam: Benjamins.
Grimm-Vogel, Irmingard Hildburg
1998
Philipp Wegener 18481916: Wesen Wirken Wege. Dissertation Rheinischen Friedrich-Wilhelms-Universitt zu Bonn.
253
Grnbaum, Abraham Anton

1932
Sprache als Handlung. In: Gustav Kafka (ed.), Bericht ber den 12. Kongre
der Deutschen Gesellschaft fr Psychologie (Hamburg 1931), 164176. Jena:
G. Fischer.
Habermas, Jrgen
1981
Theorie des kommunikativen Handelns. 2 vols. Frankfurt am Main: Suhrkamp.
Hajicov, Eva
1994
Topic/Focus and related research. In: Philip A. Luelsdorff (ed.), The Prague
School of Structural and Functional Linguistics: A Short Introduction,
245275. (Linguistic and Literary Studies in Eastern Europe 41.). Amsterdam:
Benjamins.
1978
Language as Social Semiotic: The Social Interpretation of Language and
Meaning. London: Edward Arnold.
Heger, Klaus
1984
Zeigfeld und Symbolfeld. In: Carl F. Graumann and Theo Herrmann (eds.),
Karl Bhlers Axiomatik: Fnfzig Jahre Axiomatik der Sprachwissenschaften, 105116. Frankfurt am Main: Klostermann.
Henzler, Stefan
1987
Die Funktionen der Sprache im Gebrauch: Zu Konsistenz und Funktionalismus von Karl Bhlers Sprachtheorie. In: Werner Abraham and Ritva rhammar (eds.), Linguistik in Deutschland, 353365. Tbingen: Niemeyer.
Herrmann, Theo
1984
Sprachstrukturen und ihre aktionale Realisation: Psychologische Erwgungen
zu Karl Bhlers Axiomatik der Sprachwissenschaften. In: Carl F. Graumann
and Theo Herrmann (eds.), Karl Bhlers Axiomatik: Fnfzig Jahre Axiomatik
der Sprachwissenschaften, 143172. Frankfurt am Main: Klostermann.
Holenstein, Elmar
1976
Roman Jakobsons Approach to Language: Phenomenological Structuralism.
Transl. by Catherine Schelbert and Tarcisius Schelbert. Bloomington: Indiana
University Press.
Husserl, Edmund
19001901 Logische Untersuchungen. Vol. I: Prologomena zur reinen Logik. Vol. II/1:
Untersuchungen zur Phnomenologie und Theorie der Erkenntnis. Vol. II/2:
Elemente einer phnomenologischen Aufklrung der Erkenntnis. Halle: Niemeyer.
Hrmann, Hans
1978
Meinen und Verstehen: Grundzge einer psychologischen Semantik. Frankfurt
am Main: Suhrkamp.
Hlzer, Heike
1987
Die Metapher: Kommunikationssemantische berlegungen zu einer rhetorischen Kategorie. Mnster: Nodus.
Hymes, Dell
1964
Introduction: Toward ethnographies of communication. In: John J. Gumperz
and Dell Hymes (eds.), The Ethnography of Communication, 134. (Special
issue of American Anthropologist 66, 6, part 2.)
254
Hymes, Dell H.
1968
The ethnography of speaking. In: Joshua A. Fishman (ed.), Readings in
the Sociology of Language, 99138. The Hague: Mouton. First published
Washington, D.C. [1962].
Hymes, Dell
1972
Models of the interaction of language and social life. In: John J. Gumperz
and Dell H. Hymes (eds.), Directions in Sociolinguistics: The Ethnography of
Communication, 3571. Oxford: Blackwell.
Hymes, Dell
1983
The pre-war Prague school and post-war American anthropological linguistics.
In: Dell Hymes, Essays in the history of linguistic anthropology, 331345.
Innis, Robert E.
1982
Karl Bhler: Semiotic Foundations of Language Theory. New York/London:
Plenum Press.
Innis, Robert E.
1992
Karl Bhler (18791963). In: Marcelo Dascal, Dietfried Gerhardus, Kuno Lorenz, Georg Meggle (eds.), Sprachphilosophie: Ein internationales Handbuch
zeitgenssischer Forschung. Vol. 1, 550562. (Handbcher zur Sprach- und
Kommunikationswissenschaft 7.1.) Berlin/New York: de Gruyter.
Jakobson, Roman
1950
Les catgories verbales. [Summary]. Cahiers Ferdinand de Saussure 9: 6.
Jakobson, Roman
1953
[From the point of view of linguistics.] In: Results of the Conference of Anthropologists and Linguists, 1121. (Indiana University Publications in Anthropology and Linguistics; International Journal of American Linguistics, Memoir 8.)
Baltimore: Waverly Press.
Jakobson, Roman
1960
Jakobson, Roman
1963
Efforts toward a means-ends model of language in inter-war continental linguistics. In: Christine Mohrmann, F. Norman and Alf Sommerfelt (eds), Trends
in modern linguistics, 104108. Utrecht/Antwerp: Spectrum.
Jakobson, Roman
1966
Zur Struktur des russischen Verbums. In: Eric P. Hamp, Fred W. Householder
and Robert Austerlitz (eds.), Readings in Linguistics II, 15. Chicago/London:
The University of Chicago Press. First published Prague: Prague Linguistic
Circle [1932].
Jakobson, Roman
1971
Shifters, verbal categories, and the Russian verb. In: Roman Jakobson, Selected Writings, vol. 2: Word and Language, 130147. The Hague/Paris: Mouton. First published Cambridge, MA: Harvard University [1957].
Jakobson, Roman
1985
Metalanguage as a linguistic problem. In: Roman Jakobson, Selected Writings,
vol. 7: Contributions to Comparative Mythology: Studies in Linguistics and
Philology, 19721982, ed. by Stephen Rudy, 113121. Berlin: Mouton de
Gruyter. First published [1976].
255
Jespersen, Otto
1922
Language: Its Nature, Development and Origin. London: George Allen and
Unwin; New York: Holt.
Juchem, Johann G.
1984
Die Konstruktion des Sprechens: Kommunikationssemantische Betrachtungen
zu Philipp Wegener. Zeitschrift fr Sprachwissenschaft 3: 318.
Juchem, Johann G.
1986
Wegener und Wundt. Kodikas/Code 9: 155166.
Juchem, Johann G.
1987
Einleitung. In: Gerold Ungeheuer, Kommunikationstheoretische Schriften,
vol. 1: Sprechen, mitteilen, verstehen. Ed. by Johannes G. Juchem, 117. (Aachener Studien zur Semiotik und Kommunikationsforschung 14.). Aachen:
Rader Verlag.
Kamp, Rudolf
1977
Axiomatische Sprachtheorie: Wissenschaftstheoretische Untersuchungen zum
Konstitutionsproblem der Einzelwissenschaften am Beispiel der Sprachwissenschaftstheorie Karl Bhlers. Berlin: Duncker and Humblot.
Klein, Wolfgang
1984
Bhler Ellipse. In: Carl F. Graumann and Theo Herrmann (eds.), Karl Bhlers
Axiomatik: Fnfzig Jahre Axiomatik der Sprachwissenschaften, 117141.
Frankfurt am Main: Klostermann.
Knobloch, Clemens
1988
Geschichte der psychologischen Sprachauffassung in Deutschland von 1850
bis 1920. (Reihe Germanistische Linguistik 86.) Tbingen: Niemeyer.
Knobloch, Clemens
1991
Introduction. In: Philipp Wegener, Untersuchungen ber die Grundfragen des
Sprachlebens. New edition by Konrad Koerner, with an introduction by Clemens Knobloch, xili. (Classics in Psycholinguistics 5.) Amsterdam: Benjamins.
Knobloch, Clemens and Stefan Schallenberger
1993
Sprechhandlung und Sprachbedeutung in der Sprachpsychologie um 1930.
Histoire pistmologie Langage 15/1: 81109.
Koerner, E. F. Konrad
1984
Karl Bhlers theory of language and Ferdinand de Saussures Cours. Lingua 62: 324.
Lazarus, Moritz
1884
Geist und Sprache: Eine psychologische Monographie. 3rd ed. Berlin:
Dmmler. First published Berlin [18561857].
Laziczius, Julius von
1939
Das sog. dritte Axiom der Sprachwissenschaft. Acta Linguistica 1: 162167.
Leska, Oldrich
1995
Vilm Mathesius (18821945). [In Czech.] Slovo a Slovesnost 56: 8197.
1983
Liberman, Anatoly
1987
Roman Jakobson and his contemporaries on change in language and literature
(the teleological criterion). In: Krystyna Pomorska, Elzbieta Chodakowska,
256
Hugh McLean and Brent Vine (eds.), Language, Poetry and Poetics: The
Generation of the 1890s: Jakobson, Trubetzkoy, Majakovskij, 143155. Berlin:
Mouton de Gruyter.
Lohmann, Johannes
19421943 Karl Bhlers Drittes Axiom. Acta Linguistica 3: 516.
Lyons, John
1977
Semantics. 2 vols. Cambridge: Cambridge University Press.
Malinowski, Bronislaw
1946
The problem of meaning in primitive languages. In: Charles K. Ogden and Ivor
A. Richards, The Meaning of Meaning: A Study of the Influence of Language
upon Thought and of the Science of Symbolism. 8th ed, 296336. London:
Kegan Paul. First published London: Kegan Paul [1923].
Marty, Anton
1908
Untersuchungen zur Grundlegung der allgemeinen Grammatik und Sprachphilosophie. Vol. 1. Halle: Niemeyer.
Masaryk, Toms Garrigue
1887
Versuch einer concreten Logik: Classification und Organisation der Wissenschaften. Wien: C. Konegen.
Mathesius, Vilm
1929
Zur Satzperspektive im modernen Englisch. Archiv fr das Studium der
neueren Sprachen und Literaturen 155: 202210.
Mathesius, Vilm
1930
On linguistic characterology with illustrations from Modern English. In: Actes
du premier Congrs International de Linguistes la Haye, du 1015 avril
1928, 5663. Leiden: Sijthoff.
Mathesius, Vilm
1966
Ten years of the Prague Linguistic Circle. Transl. by Josef Vachek. In: Josef
Vachek (ed.), The Linguistic School of Prague: An Introduction to its Theory
and Practice, 137151. (Indiana University Studies in the History and Theory
of Linguistics.) Bloomington/London: Indiana University Press. First published in Czech, Prague [1936].
Mathesius, Vilm
1975a
A Functional Analysis of Present Day English on a General Linguistic Basis.
Ed. by Josef Vachek, transl. by L. Duskov. (Janua Linguarum, Series Practica
208). The Hague/Paris: Mouton; Prague: Academia. First published in Czech,
Prague [1961].
Mathesius, Vilm
1975b
On information-bearing structure of the sentence. Transl. by Tsuneko Olga Yokoyama. Harvard Studies in Syntax and Semantics 1: 467480. First published
in Czech, Prague [1939].
Mathesius, Vilm
1983a
On the potentiality of the phenomena of language. Transl. by Josef Vachek. In:
Josef Vachek and Libuse Duskov (eds.), Praguiana: Some basic and less
known aspects of the Prague Linguistic School, 343. (Linguistic and Literary
Studies in Eastern Europe 12.). Amsterdam: Benjamins. First published in
Czech, Prague [1911].
257
Mathesius, Vilm
1983b
New currents and tendencies in linguistic research. In: Josef Vachek and
Prague Linguistic School, 4563. (Linguistic and Literary Studies in Eastern
Europe 12.). Amsterdam: Benjamins. First published Prague [1926].
Mathesius, Vilm
1983c
Functional linguistics. Transl. by Libuse Duskov. In: Josef Vachek and
Prague Linguistic School, 121142. (Linguistic and Literary Studies in Eastern Europe 12.). Amsterdam: Benjamins. First published in Czech, Prague
[1929].
Morpurgo Davies, Anna
1998
History of Linguistics. Vol. 4: Nineteenth-Century Linguistics. Ed. by Giulio
Lepschy. (Longman Linguistics Library.) London/New York: Longman. First
published in Italian, Bologna [1992].
Musolff, Andreas
1990
Kommunikative Kreativitt: Karl Bhlers Zweifelderlehre als Ansatz zu einer
Theorie innovativen Sprachgebrauchs. Aachen: Alano.
Musolff, Andreas
1993
Karl Bhlers and Alan Gardiners concepts of metaphor in the context of their
theories of speech and language. Beitrge zur Geschichte der Sprachwissenschaft 3: 255272.
Musolff, Andreas
1997
Karl Bhler. In: Jef Verschueren, Jan-Ola stman, Jan Blommaert, Chris Bulcaen (eds.), Handbook of Pragmatics 1997, 115. Amsterdam: Benjamins.
Nekula, Marek
1999
Vilm Mathesius. In: Jef Verschueren, Jan-Ola stman, Jan Blommaert, Chris
Bulcaen (eds.), Handbook of Pragmatics 1999, 114. Amsterdam: Benjamins.
Nekula, Marek, and Klaas-Hinrich Ehlers
1996
Jakobsons Bilanz des Strukturalismus: Ein Vortrag, der jahrzehntelang nicht
gedruckt werden konnte. Beitrge zur Geschichte der Sprachwissenschaft 6:
187199.
Nerlich, Brigitte
1990
Change in Language: Whitney, Bral, and Wegener. (Routledge History of
Linguistic Thought Series.) London/New York: Routledge.
Nerlich, Brigitte
1992
Semantic Theories in Europe 18301930: From Etymology to Contextuality.
(Studies in the History of the Language Sciences 59.) Amsterdam: Benjamins.
1996
Ortner, Hanspeter
1983
Karl Bhlers Stellung in der Ellipsenforschung. Zeitschrift fr Germanistische
Linguistik 11: 149165.
Ortner, Hanspeter
1986
Bhlers Vierfelderschema (Das dritte Axiom der Sprachtheorie) Grundgedanken und Rezeption. Kodikas/Code, Ars Semeiotica 9: 211226.
258
Prague Linguistic Circle

1983
Theses presented to the first Congress of Slavists held in Prague in 1929.
Transl. from Czech by Josef Vachek. In: Josef Vachek and Libuse Duskov
(eds.), Praguiana: Some basic and less known Aspects of the Prague Linguistic
School, 77120. (Linguistic and Literary Studies in Eastern Europe 12.). Amsterdam: Benjamins. First distributed as loose sheets in Czech, Prague [1929].
Rudy, Stephen
1978
Jakobsonian Poetics of the Moscow and Prague Periods. Doctoral dissertation
Yale University.
Saussure, Ferdinand de
1931
Grundfragen der allgemeinen Sprachwissenschaft. Berlin: W. de Gruyter.
Transl. by Hermann Lommel of: Cours de linguistique gnrale, ed. by
Charles Bally and Albert Sechehaye, 1916.
Scheerer, Thomas M.
1980
Ferdinand de Saussure: Rezeption und Kritik. (Ertrge der Forschung 133.)
Darmstadt: Wissenschaftliche Buchgesellschaft.
Schiffrin, Deborah
1994
Approaches to Discourse. Oxford: Blackwell.
Schlieben-Lange, Brigitte
1979
Linguistische Pragmatik. Stuttgart: W. Kohlhammer.
1949
Press.
Swiggers, Pierre
1981
Two dimensions in Karl Bhlers sign theory. Kodikas/Code, Ars Semeiotica 4:
5356.
Tenchini, Maria Paola
2008
Aspetti funzionali e pragmatici nel pensiero di Philipp Wegener. Con la traduzione antologica di Philipp Wegener, Untersuchungen ber die Grundfragen
des Sprachlebens. (Quaderni del Centro di Linguistica dellUniversit Cattolica 11.) Brescia: Editrice La Scuola.
Toman, Jindrich
1995
The Magic of a Common Language: Jakobson, Mathesius, Trubetzkoy, and the
Prague Linguistic Circle. Cambridge, MA: MIT Press.
Trnka, Bohumil
1966
Vilm Mathesius. Transl. by Vladimir Honsa. In: Thomas A. Sebeok (ed.),
Portraits of Linguists: A Biographical Source Book for the History of Western
Linguistics, 17461963, vol. 2, 474489. Bloomington: Indiana University
Press. First published in Czech, Prague [1946].
Tynjanov, Jurij, and Roman Jakobson
1972
Probleme der Literatur- und Sprachforschung. Translated by Rolf Fieguth und
Inge Paulmann. In: Texte der russischen Formalisten, vol. 2: 386391. First
published [1928].
Ungeheuer, Gerold
1987
Kommunikationstheoretische Schriften, vol. 1: Sprechen, mitteilen, verstehen.
Ed. by Johannes G. Juchem. Aachen: Rader Verlag.
259
Ungeheuer, Gerold
1990
Kommunikationstheoretische Schriften, vol. 2: Symbolische Erkenntnis und
Kommunikation. Ed. by H. Walter Schmitz. Aachen: Alano Verlag.
Vachek, Josef
1982
Vilm Mathesius living heritage to world linguistics. Philologica Pragensia
25:121127.
Vonk, Frank
1992
Gestaltprinzip und abstraktive Relevanz: Eine wissenschaftshistorische Untersuchung zur Sprachaxiomatik Karl Bhlers. Mnster: Nodus.
Waugh, Linda R.
1976
Roman Jakobsons Science of Language. Lisse: Peter de Ridder Press.
Waugh, Linda R., and Monique Monville-Burston
1990
Introduction: the life, work, and influence of Roman Jakobson. In: Roman
Jakobson, On Language. Ed. by Linda R. Waugh and Monique MonvilleBurston, 145. Cambridge, MA: Harvard University Press.
Waugh, Linda R., and Stephen Rudy
1998
Jakobson and structuralism. In: Roland Posner, Klaus Robering and Thomas
A. Sebeok (eds.), Semiotik: Ein Handbuch zu den zeichentheoretischen Grundlagen von Natur und Kultur, vol. 2, 22562271. (Handbcher zur Sprach- und
Kommunikationswissenschaft 13.2.). Berlin/New York: de Gruyter.
Wegener, Philipp
1902
Review of Berthold Delbrck, Grundfragen der Sprachforschung (Strassburg:
Trbner, 1901). Literarisches Centralblatt 12 (22. Mrz 1902): 401410.
Wegener, Philipp
1911
Exposition und Mitteilung: Ein Beitrag zu den Grundfragen des Sprachlebens.
In: Festschrift zur Feier des 350 jhrigen Bestehens des Gymnasiums zu
Greifswald, Bd. 2: Wissenschaftliche Aufstze, 121. Greifswald: Kunike.
Wegener, Philipp
1991
Untersuchungen ber die Grundfragen des Sprachlebens. New edition by
Konrad Koerner, with an introduction by Clemens Knobloch. (Classics in Psycholinguistics 5.) Amsterdam: Benjamins. First published Halle: Niemeyer,
[1885].
Weil, Henri
1887
The Order of Words in the Ancient Languages compared with that of the
Modern Languages. Transl. by Charles W. Super. Boston: Ginn. First published in French, Paris: Joubert [1844].
Weinreich, Uriel, William Labov and Marvin I. Herzog
1968
Empirical foundations for a theory of language change. In: Winfred P. Lehmann and Yakov Malkiel (eds.), Directions for Historical Linguistics: A Symposium, 95195. Austin/London: University of Texas Press.
Weissenborn, Jrgen
1988
Von der demonstratio ad oculos zur Deixis am Phantasma: Die Entwicklung
der lokalen Referenz bei Kindern. In: Achim Eschbach (ed.), Karl Bhlers
Theory of Language, 257276. Amsterdam: Benjamins.
Wettersten, John
1988
Klpe, Bhler, Popper. In: Achim Eschbach (ed.), Karl Bhlers Theory of
Language, 327347. Amsterdam: Benjamins.
260
Wolf, Kthe
1932
Darstellungsfelder in der Sprache. In: Gustav Kafka (ed.), Bericht ber den
12. Kongre der Deutschen Gesellschaft fr Psychologie (Hamburg 1931),
449453. Jena: G. Fischer.
Wunderlich, Dieter
1969
Karl Bhlers Grundprinzipien der Sprachtheorie. Muttersprache 79: 5262.
Wundt, Wilhem
1900
Die Sprache. 2 Parts (= Vol. 1 of Vlkerpsychologie). Leipzig: Engelmann.
Wundt, Wilhem
1908
Kritische Nachlese zur Ausfragemethode. Archiv fr die gesamte Psychologie
11: 445459.
9.
Foundations: ethnomethodology and

Erving Goffman
1.
Genesis and foundations of ethnomethodology
1.1.
Harold Garfinkels perspective on peoples everyday life activities
The most crucial assumption of ethnomethodology is that social reality exists only
through the everyday activities of a given societys members. Ethnomethodologists thus focus on the question of how social order arises in everyday life and is recognized as such. Answering this question entails identifying the actual practices
used by members of a society to create order. The goal of ethnomethodological
studies is to identify and reconstruct the methods used by individuals in everyday
life to produce meaningful and differentiable activities and to indicate how they interpret the activities of others.
Ethnomethodology was founded in the 1960s by Harold Garfinkel, who received his doctorate in 1952 for a dissertation entitled The perception of the other:
A study in social order under the direction of the sociologist Talcott Parsons (cf.
Heritage [1984] 1992; Schegloff 1992). While Garfinkel viewed Parsons studies
(alongside those of Alfred Schtz and Edmund Husserl) as crucial in the development of his own perspective on the world of everyday activities (Garfinkel
[1967] 1996: ix), he also insisted on distancing himself from his former mentors
theoretical assumptions about the crucial question of how social order arises. In
Parsons structural functionalism, social order is made possible by the orientation
of the individual to intersubjectively valid norms. According to Parsons, these cultural values and rules serve as a framework for the actions of the individual, who
perceives and interprets social reality on the basis of these shared norms. Garfinkel, however, rejected the notion that the individual is geared to given norms and
asked instead how he goes about interpreting situations. Garfinkel thus shifted the
focal point away from collectively held norms as constitutive of social order to
identify the continuous interpretation and attribution of meaning as central to the
genesis of social order.
Hence in his own analysis of social order, Garfinkel concentrated on the assessments made by ordinary people and the activities they produce which can be perceived by others (Heritage 1992: 736). Garfinkel understood ethnomethodology
in contrast to the then prevalent assumption of an objective reality. In the preface
to his book Studies in Ethnomethodology (1996), he remarks that the objective
reality of social facts should not be seen as given, but rather as the result of an ongoing accomplishment of the members of a society. Social order thus arises out of
262
the interactive activities of the participants. These activities also serve as accounts to explain how the activities of the participants should be understood.
On the basis on these reflections, everyday life became a legitimate subject of
sociological analyses. Moreover, the new focus directed attention at examining
how the order of everyday life, including the world of work, is organized. For as
Garfinkel (1996: 3575) pointed out in the late 1960s, although the familiar common sense world of everyday life occupied a central place as a topic of sociological inquiry, most scholarly literature included little data and few methods which
could be used to investigate the features of familiar scenes. His main concern
was thus to affirm and to (re)discover common sense activities in a societys everyday life as a relevant object of study.
1.2.
Early ethnomethodological investigations
One of the first ethnomethodological studies (see e.g. Turner 1974b; Sudnow
1972) examines the activities which jurors as non-professional adjudicators engage in so that others will recognize them as jurors and take their actions seriously
(Garfinkel 1974: 16). Using tape recordings and interviews with jurors as a basis,
this study focuses on becoming a juror (Garfinkel 1996: 110). This process includes, for example, using rules of the official line instead of the rules of daily
life (Garfinkel 1996: 110113). When describing their actions, the jurors present
their decisions in correspondence with the official rules as right, as based on law
and evidence, not personal preferences or interests. Moreover, they justify their
decisions by stressing the integrative features of the deliberations, thereby
neglecting the anomic or random ones (Garfinkel 1996: 113). These examples
demonstrate how a continuous perspective, that of the jurors, is successively generated.
Another early study about suicidal deaths examines the question of how the
involved parties, in this case coroners, arrive at the diagnosis suicide in cases of
sudden, unnatural death (Garfinkel 1996: 11). This work analyzes the process of
inquiry through which investigators establish what really happened as they provide an account of how a particular person died (Garfinkel 1996: 15). Ethnomethodological studies are thus concerned with how activities of everyday life are organized and how this organization is accomplished through peoples everyday life
activities.
The approach as well as the name ethnomethodology can be traced back to the
study on jurors (see above) which Garfinkel carried out together with Saul Mendlovitz in Chicago during the 1950s before going on to the University of California
at Los Angeles (UCLA) in 1954. When he wrote up the study about the tapes and
interviews, he was looking for a suitable name for the doing methodology as observed in the jurors (Garfinkel 1974: 16). The prefix ethno- is intended to refer to
a societys body of general knowledge that serves the individual as a point of orien-
263
tation. Hence the term ethnomethodology designates the procedures and practices,
that is, ethno-methods, which are valid for the members within a particular society
and which produce structure. While the approach of ethnomethodology is analogous to, for example, ethnobotany, it is not intended to stand for a specific knowledge domain such as, for example, ethno-physics (Garfinkel 1974: 1617). The
genesis of the term highlights one of the main thoughts behind this sociological research approach as laid out in Garfinkels 1967 volume Studies in Ethnomethodology: the search for the everyday methods through which a societys members
structure their activities and make them comprehensible and visible as such should
be oriented towards situations and conditions of everyday life (1996: vii). This direct orientation to the accomplishment of everyday activities was achieved with
the help of tape recordings and ethnographical observations from the field. Basing
their work on common sense activities and daily situations, Garfinkel and other
ethnomethodologists of his day vehemently denounced all forms of standardization of ethnomethodological research and stressed what they were not intending
to do: neither theories, nor advice, nor recipes, but rather the practices and features
of everyday life would be the focus of their investigations (Garfinkel 1996: viii; see
also Turner 1974a).
This avoidance of solely theoretical assumptions to focus on everyday activities represents one of Garfinkels central ideas, which can be linked to classical
American pragmatism: The relevance of real lived experiences, habitual practices,
and language with its indexically and socially constitutive characteristics had already been established in the work of Peirce, Mead, Dewey, and others. Seen from
this perspective, the questions arises as to whether certain elements of Garfinkel
and ethnomethodology can be seen as the fulfillment of the unfinished business
of pragmatism (Emirbayer and Maynard 2006).
In Studies in Ethnomethodology (1996) Garfinkel presents various studies
from the preceding twelve years, thereby demonstrating the possibilities for his
form of analysis of everyday activities. In addition to introductory articles and the
study of the everyday practices of jurors mentioned above, the volume also includes two of Garfinkels best-known areas of investigation: one about a psychiatric clinic and the methods used there to identify patients for treatment, and another known as Agnes. Agnes is a pseudonym for a young woman who was
born and raised as a male person (Garfinkel 1996: 120). At the age of 17, Agnes assumed a female identity by altering her appearance and dressing as a woman; three
years later she underwent a sex-change operation. Garfinkels study uses Agnes
story to illustrate that, contrary to commonly held notions about the fundamental
difference between the sexes, the perception of normal sexuality is actually the
result of a constant situational production which, in Agnes case, was constituted
by the daily management of herself as natural female (1996: 184).
264
1.3.
Basic assumptions of ethnomethodology: making activities accountable
While the establishment of a full-fledged research program was not a concern of

the first ethnomethodologists (Turner 1974a), several of their assumptions can be
seen as central.
As mentioned above, ethnomethodology assumes that the members of a society
provide accounts of their own actions. The main focus of ethnomethodological
studies is to show how people make their activities observable, reportable and visible so that social settings become accountable for their participants. Another basic
thought behind ethnomethology is that the actions themselves indicate how they
should be understood as they are being carried out. Because they constitute both a
part of the situation in progress and a means of interpreting it, these human actions
can be characterized by a basic reflexivity. Hence the context in which an activity
should be understood is locally produced in situ by the activities themselves. This
reflective character of the accounts makes it possible to assign sense and meaning
in spite of the indexicality of the expressions and the vagueness of language. Another central concept in the work of Garfinkel known as indexicality emphasizes
the contextuality of language. Because indexical expressions are so common and
in spite of the efforts of many scholars who are ironically mentioned by Garfinkel
irreplaceable, their properties provide important clues for understanding everyday
actions. Concerned only with objective expressions or context-free language, the
exact sciences describe indexical, that is, contextual expressions as simply
awkward (Garfinkel 1996: 5) in spite of their frequency of occurrence and their
utility; from this perspective, indexical expressions appear to be irremediable nuisances (Garfinkel 1996: 7). Garfinkel, however, is most interested in exploring the
rational properties (1996: 911) of indexical expressions and thus the following
central question: How is it possible to gain understanding in spite of the recognizable and irremediable contexuality of language and human action?
In order to illustrate this contextuality and reflexivity of language and human
activity, Garfinkel had his students conduct experiments in familiar scenes of
everyday affairs known as breaching experiments. He wanted his students to
examine ordinary conversations and episodes to test the degree to which everyday
events are characterized by seen but unnoticed features (Garfinkel 1996: 3649).
Garfinkel borrowed the term seen but unnoticed from the phenomenologist
Alfred Schtz, whose work about the world of everyday life influenced Garfinkel
profoundly (Heritage 1992: 3774). Schtz referred to the background expectancies which the members of a society use to understand and interpret human actions as the attitude of daily life. Garfinkels aim was thus to find out how commonplace scenes become familiar for the members of a society. The experiments
were designed to discover the mutual expectations and assumptions which people
share as a kind of basis. Garfinkels method was to examine what happens when
people depart from the usual everyday order by breaking social rules or norms, or
265
in his words: to start with familiar scenes and ask what can be done to make
trouble (Garfinkel 1996: 37). Garfinkels students were asked to engage someone
in an ordinary conversation and to then respond to indexical phrases which are
normally not further scrutinized by asking the person to clarify their remarks. For
example when asked How are you?, they would respond: How am I in regard to
what? My health, my finances, my school work []? (Garfinkel 1996: 44). Unaware of the experiment, most people reacted to this unexpected change in the normal course of conversation by showing confusion or annoyance or by changing the
subject. It also became clear that one explanation tended to result in another, indicating that indexicality cannot be suspended or healed. Reflexivity, which is, as
mentioned above, constitutive of all actions, normally prevents these kinds of deviations in everyday life and assures that the participants in a conversation alternately receive hints to help them deal with vagueness as a matter of course.
2.
Conversation Analysis
2.1.
Harvey Sacks Lectures on Conversation
These basic assumptions about the production of order through everyday activities
have also been fundamental for conversation analysis (CA), which grew out of ethnomethodology from the mid to late 1960s to became an independent area of research (see e.g. Sudnow 1972; Turner 1974b). Harvey Sacks, who, like Garfinkel,
was at UCLA from 1963, is known as the founder of CA. Through his work, he
paved the way for a growing interest in everyday conversations as an object of
study. Garfinkel and Sacks were engaged in close scientific exchange since they
first met in a seminar by Talcott Parsons at Harvard in 1954 (Schegloff 1992: xiii;
see also Bergmann 2000a). Known as the cornerstone of CA, Lectures on Conversation is composed of lectures which Sacks held between 19641972. Following his early death in 1975, Sacks former student Gail Jefferson edited the lectures
(Sacks 1989, 1992 Vol. I & II). Sacks used recorded conversations (including telephone recordings from an emergency psychiatric hospital) as a basis for intensely
analysing the structural characteristics of verbal interactions. In the course of his
work, he introduced and advanced the position that with respect to both verbal and
nonverbal activities, the generation of social order in everyday life is a legitimate
subject of sociology (Sacks 1972). Topics covered in his extensive analyses include lies to be expected in conversations (e.g. as an answer to the question How
are you?, Sacks 1992, Part V: 8, 9), the techniques used to choose a conversations
next speaker, and the procedures used by members to categorize someone or an
event. These techniques are used automatically by the participants in a conversation and assure, as a rule, two things: that conversations proceed not chaotically,
but according to a certain order, and that each party is able to assign meaning to the
utterances of the other. These order providing mechanisms are the focus of CA.
266
Sacks dealt extensively with questions about the everyday use of categories of
membership (Sacks 1992, inter alii Part I: 6, Part V: 11, 12). In his view, the procedures and devices which members use to categorize others constitute a central
machinery of social organization. Sacks developed an apparatus called MIR device (M for membership, I for inference-rich, R for representative) (1992:
41) which includes three categories of questions: questions about the chosen category sets (e.g. race, sex, age, occupation); questions about the conclusions which
are made possible by the categories used, e.g. baby and mommy in the wellknown example The baby cried. The mommy picked it up (see Sacks 1992,
Part III: 12; Sacks [1972] 1974); and questions about the activities ascribed to
particular members as category bound, e.g. crying to babies and picking up to
mothers.
2.2.
The sequential organization of verbal interaction
Sacks thus took what Garfinkel (and others) discussed about the relevance of
everyday activities for the production of social order and applied it to verbal interactions. CA empirically examines how order is generated in conversations. Proceeding from the basic principle of ongoing accomplishment (see above), the
question of the successive generation of various ordering structures comes into
focus. The processual creation of ordering structures during interactions is conceptualized as doing (see Sacks 1984 on doing being ordinary, Garfinkels analysis
of Agnes from 1996 on doing gender, Aya 2004 on doing interviewing).
A concentration on sequential organization constitutes a central and pivotal
point of analyses in the work of Sacks and within CA. The ever-present question
about the localisation of individual elements of conversation puts the focus on the
direct context of single utterances. By analysing the sequence of utterances generated locally within the situation, it is possible to determine what an utterance refers
to, what problem a particular turn solves, and which turn can be expected to follow.
Sequential organization is the core of the methodical approach of CA, which tries
to identify the processes which give order to conversation. In conjunction with
Emanuel Schegloff, a former fellow student who later became a colleague, and
Gail Jefferson, a former student, Sacks developed various research methods for
studying sequential organization. These methods constitute the centrepiece of CA
and prominently demonstrate how CA examines interactions within the context of
their own order, which is produced turn by turn. These assumptions about the selfreferential and reflexive processes involved in the production of a conversation
make it possible to describe the autonomous nature of interactions and to designate
them as systems (see below).
The first basic structure describes how people take turns in ordinary discourse
according to a turn-taking-machinery which functions automatically during conversation. Analysing this machinery indicates how the simplest systematic for
267
the organization of turn-taking in conversation functions and how the latent problem of choosing the next speaker at a transition relevance place can be solved,
e.g. through self-selection or other selection (Sacks, Schegloff, and Jefferson
[1974] 1978). The local, sequential organization of a conversation underscores
Garfinkels assumptions about the reflexivity of all actions in everyday life: understanding in verbal interaction is possible because the utterance of one speaker reflects and makes accountable his understanding of what was said before. One of the
central methods used by speakers to organize their conversation in the course of
conversation are so-called utterance pairs or adjacency pairs (Schegloff and
Sacks 1973; Schegloff 1972). Because some utterances provoke certain responses,
they are said to operate according to a property called conditional relevance
(inter alii Schegloff 1968), which compels the next speaker to respond in an expectable way in terms of both content and form (as a rule, questions are followed
by answers and greetings are followed by return greetings). Understanding is made
possible not only in the sequence of utterances and the structural order (question/
answer, accusation/defence), but also by virtue of the fact that utterances are retrospectively repaired. The sequential repair of utterances constitutes a special feature
of conversation: an utterance can be readily revised or repaired with respect to
formulations or details in a second turn. As shown in various studies (Schegloff,
Jefferson, and Sacks 1977; Schegloff 1979; Jefferson 1974), repairs can be differentiated according to who initiates them and who actually carries them out. The
analyses indicated that participants prefer self-initiated self-repair (as opposed to
other-repair), which means that the speaker of the utterance requiring repair makes
the correction without prompting from another. This idea has also been developed
in further studies of conversation under the name of a general preference organization and can be described with reference to common adjacency pairs such as
agreement/disagreement (inter alii Pomerantz 1984). In addition to these three central areas (turn-taking, adjacency pairs, repairs), a number of structural characteristics and specific tasks of verbal interaction (such as recipient design, laughter,
opening and ending of conversation, stories) as well as conversation types (everyday conversations, consultations, trouble talks, e.g. Jefferson 1984) have been analyzed since the beginnings of CA (see Schenkein 1978; Psathas 1979; Atkinson
and Heritage 1984; Hutchby and Wooffitt 1998). The initial work of Sacks, Sudnow, Schegloff, and Jefferson in Berkeley and at UCLA placed a sustained focus
on the regularities of everyday speech (also in working life) as well as on the
(ethno)-methods by means of which this verbal order is generated.
2.3.
The analysis of empirical data
Because talk-in-interaction (Schegloff 1987) had not been the subject of such
minute analysis up until this point, another achievement of the first CAs was to develop and promote appropriate methods for archiving and preserving conver-
268
sations (Bergmann 1981, 1985). Using recordings of natural data (initially telephone conversations) as a basis, transcripts were made which attempted to record
entire exchanges word for word. Hence word interruptions, reformulations and
ums were not edited out, but rather became a part of the transcript. Both the recordings and the transcripts provide the data used in the analysis of authentic conversations (Hutchby and Woofitt 1998; ten Have 2007). Both then and now, these
natural data stand in strong contrast to data which have been contrived or generated
in experimental settings (inter alii Heritage 2001). With the help of sound recordings and, more recently, video recordings (on this topic see Heath and Luff 2000;
Goodwin 2000), CA preserves interactions by recording them for future use. These
data transcripts then form the basis for exploring the central issue of ordering structures. Seen from this perspective, reconstructive methods (e.g. narrative interviews, memorized observations) commonly used prior to CA (and even today) do
not constitute data which have been recorded verbatim from an actual situation.
Already tainted by the perspective of the one reporting it, data collected in this way
cannot provide a basis for analyzing the genuine course of interactions (Bergmann
1985).
The strictly empirical approach of CA starts from the data themselves
(Schegloff and Sacks 1973: 291) in order to reconstruct the mechanisms und
methods through which the participants of verbal interaction generate turn-by-turn
order. This focus on the course of the interactions themselves disaffirms the practice of arbitrarily applying a methodological or normative apparatus which is always supposed to be universally valid. Just as ethnomethodologists reject theoretical claims and programmatic assertions (see Turner 1974b; Garfinkel 1996), there
is a consensus within CA about the inappropriateness of these kinds of standardizations. In the view of CA, verbal interactions should be studied using methods
that arise out of the analysis of the data: to use hypotheses or theoretical links as a
starting point would be to narrow ones perspective on the order of a given set of
interactions. As it seeks to identify the specific features of a particular interaction
using turn-by-turn analysis of the data, this position can also be understood as a
kind of analytic mentality (Schenkein 1978: 6).
This absolute concentration on the order of interaction as recorded by the transcript (which now includes nonverbal elements as well) emphasizes the special
empiric understanding of CA and its perspective on interactions. The fact that the
methods of the participants are concretely reconstructed from the data themselves not only makes it possible to study authentic processes by reconstructing
the creation of order and the sequential structure of talk turn by turn. This method
of analysis with its step by step procedure also highlights the autonomy of interaction. In other words, the particular empiric understanding points to certain assumptions about the object, the interaction itself. With each successive turn, the order of
interaction shows which options emerge for the participants, what the interaction
requires, and what interactively appears as relevant. From this perspective, self-
269
referential and reflexive processes of interaction move to the centre of CA. This
focus on the autonomy of interaction makes it possible to connect CA with more
recent system theory (Luhmann 1984, 1997), which presupposes the self-referential and autopoietic organization of a system (see Hausendorf 1992; Schneider
1997; Domke 2006). These interconnections correspond to CAs fundamental interest in interaction and open up empirical links to system theorys strictly theoretically based thoughts on the formation of systems.
One of the leading achievements and characteristics of CA is the way in which
it unlike mechanistic communication models points out the internal dynamics
of interactions. However, this focus on interaction as an autonomous system also
comprises a great challenge: the context of the interaction and the participants
must be taken into account appropriately and methodically examined. It is in these
points that CA differs from Erving Goffmans perspective on interactions.
3.
Erving Goffmans Interaction Order
Alongside Garfinkels ethnomethodology, Erving Goffmans interaction analysis

paved the way for numerous studies on the subject of everyday interactions and
analyses of how social order is generated by these interactions. Goffman taught at
the University of California at Berkeley from 19571968 before taking a position at
the University of Pennsylvania. His first book, The Presentation of Self in Everyday
Life was published in 1959 and is considered to be one of his best-known works.
Goffman had a great influence on Harvey Sacks, who studied with Goffman in Berkeley before going on to UCLA. In his introductions to Sacks lectures, Schegloff
emphasises that both sociologists influenced each other, an aspect which needs to
be explored more closely (Schegloff 1992: xxii). The special nature of their relationship is underlined by Goffmans reported answer to the question of whether
Sacks studied under him: What do you mean; I was his student! (Schegloff 1992:
xxiii). In spite of numerous ties and points of common ground (Drew and Wootton
1988; Widmer 1991) Goffmans work can neither be subsumed under the name of
ethnomethodology nor that of CA, but rather occupies an independent area of empirical social research with its own form of analysing everyday activities. However,
this does not mean that a Goffman School or methodology actually developed
out of his studies. It was primarily by pointing out the crucial role and function of
the interaction order as an institution in its own right (Goffman 1983b) that Goffman came to have such an important influence on conversation analysts like Sacks
and Schegloff (Heritage 2001; see also Bergmann 1991, 2000b).
According to Goffman, the features of traffic rules for pedestrians can be
studied in a crowded kitchen as well as anywhere else, and distinguishing between public and private settings is not crucial for understanding social interactions (Goffman 1983b). In his view, solely an understanding of face-to-face inter-
270
action as an independent domain with independent structures is necessary for the

analysis of everyday situations in the presence of others. However, this does not
mean in reverse that it can be effective to analyse the structure and rules of conversations out of context. On the contrary: in his studies on interaction order (initially
Interaction Ritual, 1967), Goffman focuses on the rituals and processes (such as
deference and demeanor, see also 1967) which participants adhere to in their
everyday presentations of self in order to create face and engage in moral enterprises (Drew and Wootton 1988: 67). From this perspective, the importance of
interactions for the formation of socially relevant structures (such as face, role) becomes clear. In order to differentiate social situations, Goffman uses, among
others, the term encounter to describe participants engaged in copresence-based,
mutually ratifying interaction, and the term gathering for unfocused situations
(such as waiting for a train) (Goffman 1964, 1983b).
Goffmans analyses work under the assumption that the individual self is created and shaped in every everyday face-to-face interactions. In his early essays
On Cooling the Mark Out (Goffman 1952) and On Face-work (Goffman
1955), Goffman introduces the concept of face-work as an essential process of
presenting the self and interacting with others. He developed the concept further in
later studies. For example in Stigma (1963b), he explores how persons who deviate
from normality (such as the handicapped or homosexuals) try to come to terms
with their threatened identity in everyday life. The exclusion of the individual in
public places, e.g. via the claiming of territories (such as seats or phone boxes) is
the focus of Behaviour in Public Places (Goffman 1963a). Goffman uses central
concepts from the theater to describe the production of self (Goffman 1959). Like
actors in a play, people constantly engage in impression management in order to
embody a certain role for their audience. This everyday dramaturgy includes
various methods and attributes such as the front stage, where the performers play
their roles in full view of the audience, and the backstage, a less accessible area
which is not available for public viewing.
The key points outlined above indicate a fundamental difference between Goffmans ideas and the core of CA. In CA, the focus is on the sequential structure of
interaction and the empirical analysis of interaction structures. This approach
brings the autonomy of interaction to the fore, which allows for an understanding
of interactions as systems as described in 2.3. Goffman, by contrast, is more interested in the individual and the effects of situations on his activities. The pertinent
concepts of face and ritual put the focus on the individual in relation to his surroundings. For Goffman, interaction within the structures of these surroundings is
less relevant than the individual participant in his emerging social identity. Hence,
by looking at the pertinence of interaction for both CA and Goffman, significant
differences can be identified. In light of the mutual influence between CA and
Goffman, these differences have prompted numerous attempts to explain and distinguish the two approaches (see Bergmann 1991).
271
Goffman addresses linguistic systems and dialogic structures more concretely

(see Bergmann 1991) in his later works. The essays collected in Forms of Talk
(Goffman 1981) identify the various possibilities for differentiating social situations and encounters and discuss the role of the participants. Of particular relevance are his thoughts on participation status and participation framework (Goffman 1981: 137), concepts which stand in contrast to the rigid ascriptions of a
speaker/hearer model (see Auer 1999: 155163; Bergmann 1991: 314317). In one
of the essays entitled Footing, which first appeared in 1979 in the magazine Semiotica, Goffman describes a range of different types of speakers and hearers
(Goffman 1981: 124157). The participant who is engaged in talking is at that moment merely the sounding box in use (Goffman 1981: 144); he or she functions
as an animator. The animator and the author of what is said are not necessarily
identical and must be differentiated. A third function is that of the principal, the
one who is responsible for what is said. Taken together, these ascriptions reveal
what Goffman refers to as the production format of an utterance (Goffman 1981:
145). Hearers can also be differentiated into different modes (Goffman 1981:
131137). There are two kinds of ratified hearers: those who are directly addressed
by the speaker and those who are unaddressed. Non-ratified participants may be
listening intentionally, which Goffman calls eavesdropping, or unintentionally,
which he refers to as overhearing. All of these elements constitute the footing of an
interaction. Displayed interactively, footing is characterized by constant change.
The footing of an interaction also acts as a kind of orientation aid similar to the
more general frame of events, a concept which Goffman developed in Frame
Analysis (Goffman 1974) with reference to the earlier work of Bateson. For Goffman, frames involves definitions of situations and cognitive structures which
guide participants and indicate what might be expected of them. Frame analysis
(e.g. Goffman 1974) includes Goffmans analyses of response cries, a term which
was also the title of an article (Goffman 1978) later published in Forms of Talk.
Speakers produce exclamatory utterances like Oops! or Brr! in everyday life
as self talk without directly addressing these interjections to a particular person.
Using his distinction between social situation and encounter (see above), Goffman
describes these response cries as creatures of social situation, not states of talk
(Goffman 1981: 121) and thus identifies them as conventionalised expressions and
not as primitive, unsocialised (Goffman 1981: 120. The open state of talk is yet
another of Goffmans important designations which is useful for describing certain
communicative situations. An open state designates a conversation in which the
participants have the right but not the obligation to talk, as is the case when individuals are involved in a joint physical task or a common meal (Goffman 1981:
134135; also 1983a).
272
4.
Recent developments
Taking the fundamental studies described here as a starting point, the analysis of
verbal interaction has developed further in many respects. In the following, a few of
the most prominent out of a large number of works will be mentioned, starting with
the studies of work which grew out of ethnomethodology. In the introduction to a
volume of studies (1986) which he edited, Garfinkel emphasises that the analysis of
work processes (e.g. within a jazz ensemble studied by Harvey Sacks) makes it
clear that there exists a locally produced order of works things (Garfinkel 1986:
vii) which has largely been ignored. The specific features of individual occupational activities have become the focus of studies on everyday practices in the working world including both verbal and nonverbal activities (see Bergmann 2000b).
So-called workplace studies, which concentrate primarily on the relationship
between society and technology, have also been influenced by studies of work
(Knoblauch and Heath 1999). This area of research focuses on computer supported
work environments and the consequences of the mechanisation of various processes (inter alii Heath and Luff 2000). The next step will be to extend the methods
used in these studies to carry out large-scale document and video analyses; this endeavour will provide adequate material for future works of conversation analysis.
Other important studies which were motivated by conversation analysis analyse the characteristics of work realms or specific forms of interactions by focusing
on the verbal component (see Drew and Heritage 1992 on Talk at work; Clayman
and Heritage 2002 on News Interview). In more recent studies, this concentration on speech has given way to a more holistic approach (Dausendschn-Gay
and Krafft 2002), which tries to incorporate facial expressions, gestures and the environment of interactions into each analysis. Conversation analytical work with
video recordings has led to a discussion of classical CA focus areas (such as turntaking, conversational openings) and elaborated the interplay and necessary coordination between speech, gesture and objects for various areas including medicine,
everyday conversations and the media (inter alii Goodwin 2000; Aya 2004; Kissmann 2009; Schmitt 2007; Streeck 2009).
5.
Influence of conversation analysis on linguistic pragmatics
Conversation analysis is one of the two cornerstones of pragmalinguistics. In retrospect, it appears to be the less prevalent of the two, but in the long run it may be the
more sustainable. In simplistic terms, one could say that linguistic pragmatics has
essentially followed two major lines in its reorientation. On the one hand, it has
taken its impulse from the philosophy of language, incorporating and integrating
not only ordinary language philosophy along with the speech act theory of Austin
and Searle, but also the concept of meaning and understanding as explicated by
273
their attributed intellectual father Ludwig Wittgenstein and further developed by

Grice. On the other hand, linguistic pragmatics has been influenced by sociological
impulses which were mediated chiefly through conversation analysis and its historical backdrop, including names ranging from Harold Garfinkel and Alfred
Schtz to Max Weber. In addition, the microsociological studies of Erving Goffman with their direct relevance to every kind of interaction have impacted the field
to a lesser degree.
No linguist working in the field of pragmatics has been able to avoid coming to
grips with speech act theory and its string of tradition because speech act is the
central concept that has defined not only the whole linguistic subdiscipline, but
also the refounding of the entire field of linguistics. The study of conversation
analysis (and especially Goffman), however, has been less obligatory for two reasons. First, because it applies only to verbal interaction, CA was thought to represent a limited subject area. Secondly, CA was in competition with other approaches which also attempted to analyze spoken interaction or which did not
make a rigorous distinction between (spoken) interactions and (written) texts. Another possible explanation for the resistance to CA may be the challenges it posed
to adherents of speech act theory from the philosophy of language who were accustomed to working with constructed and isolated example sentences. The consistent empirical approach to collecting data and diligently transcribing them
required a fundamentally different kind of methodology than was required by
grammar-oriented research. Finally, many were put off by what they saw as a kind
of unorthodox leftist movement, a reputation which CA earned with its wild
1960s California image, its antitheoretical mistrust of academic omniscience, and
its uniquely different perspective: the adherents of CA were searching for the
methods of the members of society themselves. Others, however, were attracted
by this unconventionality. For example, Bergmann (1981: 38) admires das Unsystematische und Unabgeschlossene, das Ungeschtzte und Riskant-Artistische, das
Fragmentarische, Labyrinthhafte und Antiakademische [the unsystematic and inconclusive, the unprotected and risky-artistic, the fragmentary, labyrinthian and
anti-academic]. In any case, it cannot be denied that the actual subject matter of CA
was paradoxically just what the more theoretical and less disputed branch of pragmatics actually had in mind without analyzing it empirically: ordinary language.
While CA researchers repeatedly denied having a primary or special interest in language and conversation (Schegloff and Sacks 1973, cited in Coulthard 1977: 52;
see also Sacks comment as quoted in Auer 1999: 137), they nonetheless were
compelled to and wanted to attract interest by designating themselves as conversation analysts, especially in light of the new pragmatic orientation of the field
of linguistics and the relevance and central importance of their subject matter, as
highlighted here by Levinson: It is not hard to see why one should look to conversation for insight into pragmatic phenomena, for conversation is clearly the
prototypical kind of language usage [] (1983: 284).
274
The gradual triumph of CA, particularly in the field of linguistics, was thus inevitable to a certain extent. The appearance of important studies by sociological researchers in linguistic journals (Sacks, Schegloff, and Jefferson 1978; Schegloff,
Jefferson, and Sacks 1977 in Language) or in a socioliguistic context (e.g. Sacks
1974, first in Gumperz and Hymes 1972) was no accident. The proximity and
kinship to newly developing sociolinguistic approaches (e.g. the ethnography of
speaking) created a situation in which the various communication-based approaches benefited each other in an area of overlap between sociology, anthropology, and linguistics; this interconnectedness also strengthened interest in works on
conversation analysis. Recognition of CA expanded beyond a small circle of interested scholars to the entire (Anglo-Saxon) field of linguistics after the publication
of the two chapters The ethnography of speaking and Conversation analysis
(following the chapter entitled Speech acts) in An Introduction to Discourse
Analysis by Malcolm Coulthard (1977). Coulthard founded the Birmingham
School together with John Sinclair on the basis of their study of classroom interaction (Sinclair and Coulthard 1975).
Within the field of German linguistics, two research structures in Bielefeld and
Freiburg were primarily responsible for the spread of studies in conversation
analysis. As early as 1973, researchers in Bielefeld published two volumes of texts
by American sociologists, including studies by Garfinkel, Sacks und Psathas on
ethnomethodology and by Hymes on the ethnography of speaking (Arbeitsgruppe
Bielefelder Soziologen 1973). The work of the group in Bielefeld also led to the
publication of a highly regarded article by the linguist Werner Kallmeyer and the
sociologist Fritz Schtze on Conversation Analysis (Kallmeyer and Schtze
1976). The Freiburg project Dialogstrukturen [dialogue structures] from 1974
under the direction of Hugo Steger investigated spoken language, thereby incorporating CA research in its investigation of spoken language (Berens et al. 1976; for
details see Schwitalla 2001a). Thanks to this early engagement in the field, the
1980 Annual Conference of the Institute for German Language on the topic of dialogue research included an overview article on ethnomethodological conversation
analysis (Bergmann 1981). This topic was also covered in later publications, including German guides to dialogue analysis and conversation linguistics (Bergmann 1994, 2001) as well as an introduction to linguistic interaction by Auer
(1999) on the basis of several classic authors in the field.
The research paradigm conversation analysis was effectively linguistically
canonized in the 1980s with the publication of the chapter on conversational structure in Pragmatics by Stephen Levinson (1983), the internationally best-known
book on the topic. Levinson deals extensively with the methods and results of CA
in this chapter, and similar introductions published later follow his example (e.g.
Green 1989; Mey 1993; Grundy 1995; Meibauer 1999). Levinsons work also
served as a model for other writers because he discusses CA in contrast to other approaches, which are combined under the label discourse analysis (DA).
6.
275
Conversation analysis vs. other approaches in the linguistics

of spoken interaction
While Levinson advocates CA, others have taken a position between reservation
and rejection. As a whole, the field of discourse analysis is quite diverse, especially
since it sometimes includes analyses of both written texts (not considered here) and
larger units such as discourse (in the sense of Foucault). This discussion, however,
will continue to focus on spoken language. Alongside the previously mentioned
approaches (Dell Hymes ethnography of speaking, Sinclair and Coulthards Birmingham School), John Gumperz developed his interactional sociolinguistics;
Levinson also cites Van Dijk (1972) and Labov and Fanshel (1977) as typical
representatives of discourse analysis; Bublitz (1991) and Ventola (2001) provide a
general overview of conversation and discourse analysis, and Hausendorf discusses the situation in German-speaking regions (2001). Some approaches are
based more strongly on speech acts: in Germany, for example dialogue analysis
and dialogue grammar (e.g. Hundsnurscher 1995, 2001; Hindelang 1994). The
volume Handbuch der Dialoganalyse [Handbook of Dialogue Analysis] (Fritz and
Hundsnurscher 1994) includes a discussion of the Geneva model of discourse
analysis (Moeschler 1994, 2001) and an action theoretically based concept of
linguistic communication analysis (Gloning 1994), which is associated with
Heringers practical semantics. Konrad Ehlich and Jochen Rehbein take their
own separate direction with their functional pragmatics in Diskursanalyse
(Rehbein 2001). Yet another example of DA is the critical discourse analysis
(Kritische Diskursanalyse) of the Wiener Schule under Ruth Wodak (e.g.
Wodak et al. 1990), an approach which is based on Faircloughs Critical Discourse Analysis (CDA) (e.g. Fairclough 1995).
Writing from the perspective of CA, Levinson (1983: 286294) targets text
grammatical or speech act theoretical approaches of DA with his criticisms of
methods and theoretical tools (Levinson 1983: 294). Claiming that DA sometimes falls back on constructed examples, he criticizes a lack of empirical focus as
well as a fundamentally intuitive approach which ultimately cannot tolerate the
possibility of falsifiability. He also takes issue with DAs acceptance of the familiar
theoretical principles of well-formedness (see also Kohrt 1986: 7475), a viewpoint which sees discourses as nothing more than strings of interlinked sentences.
In the eyes of DAs critics, this oversimplifying grasp (Kohrt 1986: 80) can
easily lead to the construction of rigid or even normative structures (Bublitz 1991:
271). According to Levinson, the practice of assigning individual utterances to one
particular speech act and then constructing structurally definable sequence mechanisms (as in sentences) is bound to fail because it cannot do justice to the complex
reality of the unpredictable dynamic interactions which are themselves embedded
in complex contexts (Levinson 1983: 288294). Proponents of CA defend their
method by pointing out that it is strictly empirical and inductive, with as little ap-
276
peal as possible to intuitive judgements (Levinson 1983: 287), and that it does not
immediately categorize individual cases, but rather attempts to compile functionally similar cases from the data in order to prove the validity of an analysis (Bergmann 2001: 923). The most effective argument for CA is definitely its indisputable
success in the recognition of the structures which organize speech.
However, CA has also had its critics. For example, Levinson (1983: 287)
makes reference to Labov and Fanshel (1977: 25) and Coulthard and Brazil (1979)
as he points out: DA theorists can accuse CA practitioners of being inexplicit, or
worse, plain muddled, about the theories and conceptual categories they are actually employing in analysis. It is indeed questionable whether all existing order
can be deduced solely from the object of investigation. In other words, it is doubtful that utterances are each reflexive and indexical enough to be understood without being categorized by the interpreter from the outside. In fact, it is commonly
recognized that by drawing on his or her analytical mentality, the ethnomethodologist is forced to draw repeatedly on the intuitive understanding which he or she
possesses as a competent member of a linguistic community (Bergmann 1981: 23).
At the same time, this demonstrative refusal to draw on additional outside knowledge in an attempt to immunize CA against possible allegations of interpretive arbitrariness has not only led several critics to label the field as anti-psychological
and individual-psychological (Redder 1990: 7), but it has also caused some to
cross the line to (unjustifiably) accuse CA of secret positivism (Flader and Trotha 1988). A further criticism cites an alleged failure to take social background into
account (e.g. Fairclough 1989: 12; see also Bublitz 1991: 271), but this assessment
is not particularly convincing in view of CAs sociological and sociocritical origins. Moreover, more recently, a growing number of studies in CA have focused on
institutional and work contexts as well as political communication.
On the whole, the field of linguistic pragmatics has not really seen an intense
dispute over theoretical and methodological issues surrounding CA; the predominant pattern has been rather a more or less reflected and more or less selective appropriation of methods and results. Evidence of these integrative tendencies can be
found in many textbooks and introductions, but they arise above all out of the experiences and needs of concrete research. Hence at the end of the 1990s, Hausendorf (2001: 977) came to the following conclusion with respect to the situation in
the German-speaking realm: To the extent to which the methodological praxis of
conversation analysis consolidated itself and the findings and conclusions of empirically sophisticated i.e. complex and long-term projects were presented in
the course of the 1990s, the methodological struggle between the various approaches has faded away.
7.
277
Influence of Erving Goffman on linguistic pragmatics
While CA quickly attained prominence within linguistic pragmatics on account of

its direct relevance for the study of spoken language, the linguistic importance of
Goffmans research initially appeared to be limited to marginal questions. However, his work has been widely recognized and discussed within many other disciplines because of its broad spectrum of topics (self-portrayal, asylums, interaction
rituals, role distance, gender advertising, frame analysis), so that gradually, his
studies have also entered the field of linguistics indirectly via other routes. At least
four strands of Goffmans work have been adopted by pragmatics in one form or
another over the course of time.
First, Goffmans concept of face was further developed by Brown and Levinson ([1978] 1987) into a politeness model with universal applicability. This model
was adopted in numerous studies on the topic of language and politeness, but it
was also vehemently criticized for its weak empirical basis of single utterances as
well as for the questionable way in which the model was universalized (for summary see Fraser 2001; Bublitz 2009: 270). However, the complexity of Goffmans
concept (see Tracy 1990) comprises at its core something that goes beyond mere
aspects of politeness to encompass the ritual negotiation of identities and relationships between interaction partners. This second Goffmanian strand was received
by conversation analysis quite early in a few instances (e.g. Holly 1979, 2001),
but it has taken more time to develop the concept fully for linguistics with respect
to routines (Coulmas 1981), rituals (e.g. Rauch 1992; Werlen 2001), conflict
settlement (e.g. Schank and Schwitalla 1987; Schwitalla 2001b) and further issues. The third important strand constitutes Goffmans frame analysis, which has
been applied more or less prominently in many pragmalinguistic studies (e.g.
Tannen 1984). Finally, the fourth strand involves Goffmans explicit investigations of conversations (Forms of Talk), in particular the deconstruction of
participant roles. This deconstructive aspect of Goffmans work was adopted by
Levinson (1988) and formed the basis of a new conceptualization of the idealized
and simplistic speaker-hearer-constellation (see Schwitalla 2001c), e.g. to analyse multiple addressing (Khn 1995) or testimonials in a television commercial
(Holly 2007). Moreover Goffmans analysis of different intensities of focus on
talk generated the useful concept of open state of talk, which is connected to
Bhlers observations on action-accompanying talk and is also essential for the
analysis of certain uses of speech (see e.g. on the reception of television texts
Holly, Pschel, and Bergmann 2001). The fact that four articles in the Handbook
of Linguistics of Text and Conversation focus primarily on Goffmanian concepts
(Werlen 2001; Schwitalla 2001b, 2001c; Holly 2001) is an index of how widely
Goffmans research has been received in the field of pragmalinguistics. Nonetheless, his importance for the analysis of spoken language continues to be underestimated in pragmalinguistics, probably because of his independence as a scholar,
278
which fostered a mentality against the formation of schools, and his conceptual
flexibility.
8.
The convergence of various approaches
Although ethnomethodological conversation analysis constitutes a markedly independent and distinctive school of thought in the study of spoken language, there is
a gradual tendency toward increasing convergence with other related approaches
which are anchored more firmly in the fields of anthropology, sociology or linguistics; Heritage (2001) combines the three under the label ethno-sciences.
These ethno-sciences share in common not only work with empirical data, but also
a series of core ideas which he characterizes in the following:
This research embodies the core notions that (i) communicative meaning is inherently
contextual in character, (ii) social context is unavoidably dynamic and is managed
through the participants actions, (iii) the specific contextual significance of actions is
structurally achieved by means of rules and practices of conduct which are systematically related and organized as systems, (iv) the contextual significance of action also involves a web of inferences which are inescapable, very often involve personal, moral
and social accountability, and thus connect interaction with culture, social structure and
personality, and (v) all this is managed through the integrated significance of talk, paralanguage and body movement. (Heritage 2001: 916)
With these kinds of common convictions as a basis, infighting, which also has seen
adherents of CA criticize a less than rigorous handling of methodological questions (Streeck 1983), could relax. At the same time, this convergence is being promoted in the field of linguistics by an increasing return to the original competencies, resulting in a common focus on linkages between grammatical and interactive
structures (Uhmann 1997; Ochs, Schegloff, and Thompson 1997), in particular
with reference to prosodic-intonational structures as foreseen in the concept of
interactional linguistics by Elisabeth Couper-Kuhlen und Margret Selting. This
approach sees itself as an interface between linguistics und interaction theory
(above all functional linguistics, conversation analysis, contextualization theory,
anthropological linguistics) (Selting and Couper-Kuhlen 2000: 76). Other efforts
concentrate more closely on the connection between grammar and interaction (e.g.
Deppermann, Fiehler, and Spranz-Fogasy 2006), so that the discipline of conversation linguistics reencounters many older orientations in the investigation of
spoken language (see Schwitalla 2001a), be it on the basis of Ehlich und Rehbeins
Functional Pragmatics which was originally intended to focus on grammar as
well, or on the basis of construction grammar (Gnthner and Imo 2006; Deppermann 2007).
279
References
Arbeitsgruppe Bielefelder Soziologen (ed.)
1973
Alltagswissen, Interaktion und gesellschaftliche Wirklichkeit. 2 vols. Reinbek:
Rowohlt.
Atkinson, John Maxwell and John Heritage (eds.)
1984
Structures of Social Action: Studies in Conversation Analysis. Cambridge:
Auer, Peter
1999
Sprachliche Interaktion. Eine Einfhrung anhand von 22 Klassikern. Tbingen: Niemeyer.
Aya, Ruth
2004
Konversationsanalytische Medienforschung. Medien und Kommunikationswissenschaft, 52 (1): 529.
Berens, Franz-Josef, Karl-Heinz Jger, Gerd Schank and Johannes Schwitalla
1976
Projekt Dialogstrukturen. Ein Arbeitsbericht. Mnchen: Hueber.
Bergmann, Jrg R.
1981
Ethnomethodologische Konversationsanalyse. In: Peter Schrder and Hugo
Steger (eds.), Dialogforschung. (Jahrbuch 1980 des Instituts fr deutsche
Sprache), 953. Dsseldorf: Schwann.
Bergmann, Jrg R.
1985
Flchtigkeit und methodische Fixierung sozialer Wirklichkeit. Aufzeichnungen als Daten der interpretativen Soziologie. In: Wolfgang Bon and Heinz
Hartmann (eds.), Entzauberte Wissenschaft. Zur Relativitt und Geltung soziologischer Forschung, 299321. Gttingen: Schwartz.
Bergmann, Jrg R.
1991
ber Erving Goffmans Soziologie des Gesprchs und seine ambivalente Beziehung zur Konversationsanalyse. In: Robert Hettlage and Karl Lenz (eds.),
Erving Goffman ein soziologischer Klassiker der zweiten Generation, 301
Bergmann, Jrg R.
1994
Ethnomethodologische Konversationsanalyse. In: Gerd Fritz and Franz
Hundsnurscher (eds.), Handbuch der Dialoganalyse, 316. Tbingen: Niemeyer.
Bergmann, Jrg R.
2000a
Harold Garfinkel und Harvey Sacks. In: Uwe Flick, Ernst von Kardorff and
Ines Steinke (eds.), Qualitative Forschung. Ein Handbuch, 5162. Reinbek:
Rowohlt.
Bergmann, Jrg R.
2000b
Ethnomethodologie. In: Uwe Flick, Ernst von Kardoff and Ines Steinke (eds.),
Qualitativeforschung. Ein Handbuch, 118135. Reinbek: Rowohlt.
Bergmann, Jrg R.
2001
Das Konzept der Konversationsanalyse. In: Brinker et al. (eds.), 919927.
Brinker, Klaus, Gerd Antos, Wolfgang Heinemann and Sven Frederik Sager (eds.)
2001
Text- und Gesprchslinguistik. 2nd vol. (Handbcher zur Sprach- und Kommunikationswissenschaft 16.) Berlin/New York: de Gruyter.
280

1987
Politeness. Some Universals in Language Usage. Cambridge: Cambridge University Press. First published in: Esther N. Goody (ed.), Questions and Politeness, 56289. Cambridge: Cambridge University Press 1978.
Bublitz, Wolfram
1991
Discourse Analysis: The State of the Art. In: Claus Uhlig and Rdiger
Zimmermann (eds.), Anglistentag 1990 Marburg. Proceedings, 259284. Tbingen: Niemeyer.
Bublitz, Wolfram
2009
Englische Pragmatik. Eine Einfhrung. 2nd ed. Berlin: Erich Schmidt Verlag.
Clayman, Steve and John Heritage
2002
The News Interview. Journalists and Public Figures on the Air. Cambridge:
Coulmas, Florian
1981
Routine im Gesprch. Zur pragmatischen Fundierung der Idiomatik. Wiesbaden: Athenaion.
Coulthard, Malcolm
1977
An Introduction to Discourse Analysis. London: Longman.
Coulthard, Malcolm and David Brazil
1979
Exchange Structure. Birmingham: Birmingham University.
Dausendschn-Gay, Ulrich and Ulrich Krafft
2002
Text und Krpergesten: Beobachtung zur holistischen Organisation der Kommunikation. Psychologie und Sozialwissenschaft. Zeitschrift fr Qualitative
Forschung. Themenheft: Sprechen vom Krper Sprechen mit dem Krper 4:
3060.
Deppermann, Arnulf
2007
Grammatik und Semantik aus gesprchsanalytischer Sicht. Berlin/New York:
de Gruyter.
Deppermann, Arnulf, Reinhard Fiehler and Thomas Spranz-Fogasy (eds.)
2006
Grammatik und Interaktion. Untersuchungen zum Zusammenhang zwischen
grammatischen Strukturen und Interaktionsprozessen. Radolfzell: Verlag fr
Gesprchsforschung.
Domke, Christine
2006
Besprechungen als organisationale Entscheidungskommunikation. Berlin/
New York: de Gruyter.
Drew, Paul and John Heritage (eds.)
1992
Talk at work: interaction in institutional settings. Cambridge: Cambridge University Press.
Drew, Paul and Anthony Wootton (eds.)
1988
Erving Goffman. Exploring the interaction order. Cambridge: Polity Press. Reprinted Cambridge: Polity Press [1996].
Emirbayer, Mustafa and Douglas Maynard
2006
Pragmatism and Ethnomethodology. Paper presented at the annual meeting of
the American Sociological Association, Montreal Convention Center. Montreal, Canada: Online <PDF>. 2009-05-24 from http://www.allacademic.com/
meta/p95497_index.html.
281
Fairclough, Norman
1989
Language and Power. London: Addison Wesley.
Fairclough, Norman
1995
Critical Discourse Analysis: The critical study of language. London: Longman.
Flader, Dieter and Thilo von Trotha
1988
ber den geheimen Positivismus und andere Eigentmlichkeiten der ethnomethodologischen Konversationsanalyse. Zeitschrift fr Sprachwissenschaft 7:
92115.
Fraser, Bruce
2001
The form and function of politeness in conversation. In: Brinker et al. (eds.),
14061425.
Fritz, Gerd and Franz Hundsnurscher (eds.)
1994
Handbuch der Dialoganalyse. Tbingen: Niemeyer.
Garfinkel, Harold
1952
The Perception of the Other: A Study in Social Order. Ph.D. Dissertation. Boston: Harvard University.
Garfinkel, Harold
1974
The Origins of the Term Etnomethodology. In: Turner, Roy (ed.), Ethnomethodology. Selected Readings, 1519. Harmondsworth: Penguin Education.
Garfinkel, Harold (ed.)
1986
Ethnomethodological studies of work. London: Routledge and Kegan Paul.
Garfinkel, Harold
1996
Studies in ethnomethodology. Cambridge: Polity Press. First published Englewood Cliffs, NJ: Prentice Hall [1967].
Gloning, Thomas
1994
Praktische Semantik und Linguistische Kommunikationsanalyse. In: Fritz et
al. (eds.), 113129.
Goffman, Erving
1952
On Cooling the Mark Out: Some Aspects of Adaption to Failure. Psychiatry 15
(4), 451463. Reprinted in: Charles Lemert and Ann Branaman (eds.), The
Goffman Reader, 420. Oxford: Blackwell [1997].
Goffman, Erving
1955
On Face-work: An Analysis of Ritual Elements in Social Interaction. Psychiatry 18 (3), 213231.
Goffman, Erving
1959
The Presentation of Self in Everyday Life. Garden City, NY: Doubleday, Anchor Books.
Goffman, Erving
1963a
Behaviour in Public Places: Notes on the Social Organization of Gatherings.
New York: The Free Press.
Goffman, Erving
1963b
Stigma: Notes on the Management of Spoiled Identity. Engelwood Cliffs, NJ:
Prentice Hall.
Goffman, Erving
1964
The Neglected Situation. American Anthropologist 66: 133136. Reprinted in
282
Goffman, Erving
1967
Interactional Ritual: Essays on Face-to-Face Behaviour. Garden City, NY:
Doubleday, Anchor Books.
Goffman, Erving
1974
Frame Analyses. An Essay on the Organization of Experience. London: Harper
and Row.
Goffman, Erving
1978
Response Cries. Language 54 (4): 787815. Reprinted in Erving Goffman,
Forms of talk, 78122. Philadelphia: University of Pennsylvania [1981].
Goffman, Erving
1979
Footing. Semiotica 25 (1/2): 129. Reprinted in Erving Goffman, Forms of
talk, 124159. Philadelphia: University of Pennsylvania [1981].
Goffman, Erving
1981
Forms of talk. Philadelphia: University of Pennsylvania Press.
Goffman, Erving
1983a
Felicitys Condition. American Journal of Sociology 89 (1): 151. Reprinted in
Goffman, Erving
1983b
The Interaction Order. American Sociological Review 48: 117. Reprinted in
Goodwin, Charles
2000
Practices of seeing: visual analysis an ethnomethodological approach. In:
Theo van Leeuwen and Carey Jewitt (eds.), Handbook of visual analysis,
Green, Georgia M.
1989
Erlbaum.
Grundy, Peter
1995
Doing Pragmatics. London: Arnold.
1972
Directions in Sociolinguistics: The Ethnography of Communication. New
York/London: Holt, Rinehart and Winston.
Gnthner, Susanne and Wolfgang Imo (eds.)
2006
Konstruktionen in der Interaktion. Berlin/New York: de Gruyter.
Hausendorf, Heiko
1992
Gesprch als System. Linguistische Aspekte einer Soziologie der Interaktion.
Hausendorf, Heiko
2001
Gesprchsanalyse im deutschsprachigen Raum. In: Brinker et al. (eds.), 971
979.
Have, Paul ten
2007
Doing Conversation Analysis. A Practical Guide. 2nd ed. Los Angeles: Sage.
Heath, Christian and Paul Luff
2000
Technology in action. Cambridge: Cambridge University Press.
283
Heritage, John
1992
Garfinkel and ethnomethodology. Cambridge: Polity Press. First published
Cambridge: Polity Press [1984].
Heritage, John
2001
Ethno-sciences and their significance for conversation linguistics. In: Brinker
et al. (eds), 908919.
Hindelang, Gtz
1994
Sprechakttheoretische Dialoganalyse. In: Fritz et al. (eds.), 95112.
Holly, Werner
1979
Imagearbeit in Gesprchen. Zur linguistischen Beschreibung des Beziehungsaspekts. Tbingen: Niemeyer
Holly, Werner
2001
Beziehungsmanagement und Imagearbeit. In: Brinker et al. (eds.), 13821393.
Holly, Werner
2007
Audiovisuelle Hermeneutik. Am Beispiel des TV-Spots der Kampagne Du
bist Deutschland. In: Fritz Hermanns and Werner Holly (eds.), Linguistische
Hermeneutik, 389428. Tbingen: Niemeyer.
Holly, Werner, Ulrich Pschel and Jrg Bergmann (eds.)
2001
Der sprechende Zuschauer. Wie wir uns Fernsehen kommunikativ aneignen.
Wiesbaden: Westdeutscher Verlag.
1995
Some Remarks on the Development of Dialogue Analysis. In: Franz Hundsnurscher and Edda Weigand (eds.), Future Perspectives of Dialogue Analysis,
7993. Tbingen, Niemeyer.
2001
Das Konzept der Dialoggrammatik. In: Brinker et al. (eds.), 945952.
Hutchby, Ian and Robin Wooffitt
1998
Conversation Analysis: Principles, Practices and Applications. Cambridge:
Polity Press.
Jefferson, Gail
1974
Error corrections as an international resource. Language in Society 2:
181199.
Jefferson, Gail
1984
On the organization of laughter in talk about troubles. In: John Maxwell Atkinson and John Hertitage (eds.), Structures of social action: studies in conversation analysis, 346369. Cambridge: Cambridge University Press.
Kallmeyer, Werner and Fritz Schtze
1976
Konversationsanalyse. Studium Linguistik 1: 128.
Kissmann, Ulrike Tikvah (ed.)
2009
Video Interaction analysis. Methods and Methodology. Frankfurt am Main:
Lang
Knoblauch, Hubert and Christian Heath
1999
Technologie, Interaktion und Organisation: Die Workplace Studies. Schweizerische Zeitschrift fr Soziologie 25 (2): 163181.
Kohrt, Manfred
1986
Dialoggrammatik und/oder Konversationsanalyse. In: Franz Hundsnurscher
and Edda Weigand (eds.), Dialoganalyse. Referate der 1. Arbeitstagung
Mnster 1986, 6982. Tbingen: Niemeyer.
284
Khn, Peter
1995
Mehrfachadressierung. Untersuchungen zur adressatenspezifischen Polyvalenz sprachlichen Handelns. Tbingen: Niemeyer.
Labov, William and David Fanshel
1977
Therapeutic Discourse: Psychotherapy as Conversation. New York: Academic Press.
Lemert, Charles and Ann Branaman (eds.)
1997
The Goffman reader. Ed. with introductory essays by Charles Lemert and Ann
Branaman. Oxford: Blackwell.
1983
1988
Putting Linguistics on a Proper Footing. In: Paul Drew and Anthony Wootton
(eds.), Erving Goffman. Exploring the interaction order, 161227. Cambridge:
Polity Press.
Luhmann, Niklas
1984
Soziale Systeme. Grundriss einer allgemeinen Theorie. Frankfurt am Main:
Suhrkamp.
Luhmann, Niklas
1997
Die Gesellschaft der Gesellschaft, Vols 1 and 2. Frankfurt am Main: Suhrkamp.
Meibauer, Jrg
1999
Pragmatik. Eine Einfhrung. Tbingen: Stauffenburg.
Mey, Jacob L.
1993
Moeschler, Jacques
1994
Das Genfer Modell der Gesprchsanalyse. In: Fritz et al. (eds.), 6994.
Moeschler, Jacques
2001
The Geneva School. In: Brinker et al. (eds.), 952957.
Ochs, Elinor, Emanuel A. Schegloff and Sandra Annear Thompson (eds.)
1997
Pomerantz, Anita
1984
Agreeing and disagreeing with assessments: some features of preferred/dispreffered turn shapes. In: John Maxwell Atkinson and John Heritage (eds.),
Structures of Social Action: Studies in Conversation Analysis, 57101. Cambridge: Cambridge University Press.
Psathas, George (ed.)
1979
Everyday language: studies in ethnomethodology. New York: Irvington.
Rauch, Elisabeth
1992
Sprachrituale in institutionellen und institutionalisierten Text- und Gesprchssorten. Frankfurt am Main: Lang.
Redder, Angelika
1990
Grammatiktheorie und sprachliches Handeln: denn und da. Tbingen:
Niemeyer.
Rehbein, Jochen
2001
Das Konzept der Diskursanalyse. In: Brinker et al. (eds.), 927945.
285
Sacks, Harvey
1972
An Initial Investigation of the Usability of Conversational Data for Doing
Sociology. In: David Sudnow (ed.), Studies in social interaction, 3174. New
York: The Free Press.
Sacks, Harvey
1974 [1972] On the Analysability of Stories by Children. In: Roy Turner (ed.), Ethnomethodology. Selected Readings, 216232. Harmondsworth: Penguin Education. First published in John Gumperz and Dell Hymes (eds.), Directions in
sociolinguistics: The ethnography of communication. New York: Holt, Rinehart, & Winston.
Sacks, Harvey
1984
On doing being ordinary. In: John Maxwell Atkinson and John Heritage
Sacks, Harvey
1989
Lectures 19641965. Ed. by Gail Jefferson. Dordrecht: Kluwer Academic Publishers.
Sacks, Harvey
1992
Lectures on Conversation. Vols I and II. Ed. by Gail Jefferson. With an introduction by Emanuel A. Schegloff. Oxford: Basil Blackwell.
1978
A simplest systematics for the conversation of turn taking for conversation. In:
Jim N. Schenkein (ed.), Studies in the organization of conversational interaction, 755. New York: Academic Press. First published Language 50: 696735
[1974].
Schank, Gerd and Johannes Schwitalla (eds.)
1987
Konflikte in Gesprchen. Tbingen: Narr.
1968 [1972] Sequencing in conversational openings. American Anthropologist 70:
10751095. Reprinted in John Gumperz and Dell Hymes (eds.), Directions
in sociolinguistics: the ethnography of communication, 346380. New York:
Rinehart & Winston.
1972
Notes on a Conversational Practice. Formulating Place. In: David Sudnow
(ed.), Studies in social interaction, 75119. New York: The Free Press.
1979
The Relevance of Repair to Syntax-for-conversation. In: Talmy Givn (ed.),
Syntax and Semantics 12: Discourse and Syntax, 261286. New York: Acadamic Press.
1987
Some sources of misunderstanding in talk-in-interaction. Linguistics 25:
201218.
1992
Introduction. In: Harvey Sacks: Lectures on Conversation. Vols I and II. Ed. by
Gail Jefferson, ixxii. Oxford: Basil Blackwell.
Schegloff, Emanuel A., Gail Jefferson and Harvey Sacks
1977
The Preference for Self-Correction in the Organization of Repair in Conversation. Language 53 (2): 361382.
286

1973
Schenkein, Jim N. (ed.)
1978
Studies in the organization of conversational interaction. New York: Academic Press.
Schmitt, Reinhold (ed.)
2007
Koordination. Analysen zur multimodalen Interaktion. Tbingen: Narr.
Schneider, Wolfgang Ludwig
1997
Die Analyse von Struktursicherungsoperationen als Kooperationsfeld von
Konversationsanalyse, objektiver Hermeneutik und Systemtheorie. In: Tilmann Sutter (ed.), Beobachtung verstehen, Verstehen beobachten. Perspektiven einer konstruktivistischen Hermeneutik, 164227. Opladen: Westdeutscher Verlag,
2001a
Gesprochene-Sprache-Forschung und ihre Entwicklung zu einer Gesprchsanalyse. In: Brinker et al. (eds.), 896903.
2001b
Konflikte und Verfahren ihrer Bearbeitung. In: Brinker et al. (eds.), 1374
1382.
2001c
Beteiligungsrollen im Gesprch. In: Brinker et al. (eds.), 13551361.
Selting, Margret and Elizabeth Couper-Kuhlen
2000
Argumente fr die Entwicklung einer interaktionalen Linguistik. Gesprchsforschung Online-Zeitschrift zur verbalen Interaktion 1: 7695
1975
Towards an analysis of discourse. London: Oxford University Press.
Streeck, Jrgen
1983
Konversationsanalyse. Ein Reparaturversuch. Zeitschrift fr Sprachwissenschaft 2: 72104.
Streeck, Jrgen
2009
Gesturecraft. The manufacture of meaning. Amsterdam: Benjamis.
Sudnow, David (ed.)
1972
Studies in social interaction. New York: The Free Press.
Tannen, Deborah
1984
Conversational Style: Analyzing Talk Among Friends. Norwood, NJ: Ablex.
Tracy, Karen
1990
The many faces of facework. In: Howard Giles and W. Peter Robinson (eds.),
Handbook of language and social psychology, 209226. Chichester: John
Wiley and Sons.
Turner, Roy
1974a
Introduction. In: Roy Turner (ed.), Ethnomethodology. Selected Readings,
712. Harmondsworth: Penguin Education.
Turner, Roy (ed.)
1974b
Ethnomethodology. Selected Readings. Harmondsworth: Penguin Education.
Uhmann, Susanne
1997
Grammatische Regeln und konversationelle Strategien. Fallstudien aus Syntax
und Phonologie. Tbingen: Niemeyer.
287
Van Dijk, Teun

1972
Some Aspects of Text Grammars. The Hague: Mouton.
Ventola, Eija
2001
Discourse studies in the English-speaking countries. In: Brinker et al. (eds.),
979994.
Werlen, Iwar
2001
Rituelle Muster in Gesprchen. In: Brinker et al. (eds.), 12631278.
Widmer, Jean
1991
Goffman und die Ethnomethodologie. In: Robert Hettlage and Karl Lenz
(eds.), Erving Goffman ein soziologischer Klassiker der zweiten Generation,
Wodak, Ruth, Johanna Pelikan, Peter Nowak, Helmut Gruber, Rudolf de Cillia and Richard
Mitten
1990
Wir sind alle unschuldige Tter!. Diskurshistorische Studien zum Nachkriegsantisemitismus. Frankfurt am Main: Suhrkamp.
10.

Maeve Cooke
1.
Introduction
Sociology, as the scientific or systematic study of society, emerged as an independent academic discipline only towards the end of the nineteenth century. From the
beginning it has been marked by a dispute as to the method of inquiry appropriate
to it. A central question is whether its methodology should model that of the natural sciences or that of humanistic and cultural studies. The first position is often attributed to Auguste Comte and nineteenth and twentieth century positivism; it
holds that social phenomena can be subjected to observable natural laws. The second position is associated with Wilhelm Dilthey and nineteenth and twentieth century historicism and hermeneutics; it asserts an essential difference between the
methodology of the natural sciences and humanistic and cultural studies, emphasizing the interpretative component of social scientific inquiry.
The first position adopts an external perspective on the production of social
order, viewing it as a quasi-natural process that can be observed in its empirical
regularities and explained by way of nomological hypotheses. The second position
adopts an internal perspective on the production of social order, viewing it as a
meaningfully structured reality for social agents that should be interpreted in terms
of intentional actions.
To be sure, the lines of division are by no means clear cut. Max Weber, for
example, defines sociology as a science that attempts the interpretative understanding of social action in order thereby to arrive at a causal explanation of its course
and effects (Weber 1969). In line with this, he holds that rational interpretation, in
the sense of objective or impartial appraisal, is possible, when the agent and her interpreter equally accept the standards of judgement as valid (Habermas 1988).
Moreover, Weber is ambivalent on the question of whether the methods of the
social sciences should model those of the natural sciences. In this regard we find
striking differences in attitude and emphasis in his writings (McCarthy 1978:
141145).
We cannot explore this debate further here. For our present purposes, it is sufficient to note that, today, there is widespread, if not universal, agreement that the
methods of the natural sciences cannot simply be mapped onto sociology. Furthermore, among those who accept this, a significant number agree that the difference is due in large part to the intentional component of human action.1 By this is
meant, first, that the concept of society is internally linked with the concept of
human action and, second, that in the case of such action it is always in principle
290
Maeve Cooke
possible to ask what the agent(s) intended meant when carrying it out. The implication here is that a description of a particular human action that failed to consider the intentions of the agent would lack an essential ingredient.2 This implies in
turn that intention, understood as subjective meaning, is a sociological primitive:
an irreducible element of sociological analysis. It should also be noted that even if
we accept the reference to subjective meaning as constituting an essential methodological difference between the natural sciences and sociology, a number of important questions remain open, for example, the questions of whether causal explanations of social phenomena are possible and the degree to which objectivity is
achievable in sociology. Thus, accepting the irreducibility of subjective meaning
does not commit the sociologist to a purely internal perspective on social phenomena and require her to abstain from empirical-analytic analyses and causal explanations. However, it presents her with the challenge of developing a theoretical
framework, and corresponding methodology, that negotiates the tensions between
the external and internal perspectives.
In the early 1970s, as part of an endeavour to meet this challenge, Jrgen Habermas, a philosopher and sociologist in the Frankfurt School tradition of critical
social theory, turned to pragmatics.3 The importance of pragmatics in contemporary
sociology is largely due to Habermas contribution to the study of society. This contribution is so significant that no sociologist today can ignore his social theory and,
by extension, the programme of pragmatics underpinning it. While Habermas owes
a significant debt to his colleague Karl-Otto Apel, who introduced him to pragmatics in the late 1960s, Apels work has been more influential in philosophy than
social theory (Apel 1980). Furthermore, unlike Habermas, Apel did not develop his
reflections on pragmatics into a systematic programme. For these reasons, we will
take Habermas as our reference point in the remainder of our discussion.
In turning to pragmatics, Habermas concerns were not only methodological.
He sought, in addition, to provide an answer to central questions of sociology, such
as how we should conceptualize social action, how social order is possible and how
social order is reproduced. A further, equally important, aim was to provide a basis
for the rational critique of society without appealing to metaphysical notions of
truth, justice or the good. The Theory of Communicative Action, first published in
two volumes in 1981, attempts to meet these aims. In the following, we will examine the role of pragmatics, first, from the point of view of method, where we will
consider the conceptual strategy he deems most appropriate for sociological inquiry; then, from the point of view of the interconnected questions of social action,
social order and social reproduction; finally, from the point of view of social
critique, where we will look at the role pragmatics plays in critique of social pathologies and in developing a postmetaphysical conception of reason, which is the
normative reference point for this critique. In the concluding section we will look
briefly at some challenges to Habermas appeal to pragmatics.
2.
291
Method
In the course of the 1970s Habermas develops a theory of pragmatics that, as indicated, is supposed to help him answer central sociological questions relating to social action, social order and social reproduction as well as provide a normative
basis for his critical theory of society. In a preliminary essay based on lectures delivered at Princeton University in 1971, he focuses specifically on methodology,
identifying a number of metatheoretical issues as centrally important in the endeavour to develop a conceptual strategy appropriate for sociology.4 In each case,
in line with his concern to avoid a simple decision for either explanation or interpretation, Habermas adopts a position that seeks to retain the most valuable elements in the apparently conflicting approaches. For example, in his reflections on
the debate between methodological individualism and holism, he does not straightforwardly endorse either of these approaches but attributes to them complementary
strengths and weaknesses that make each useful for different purposes and in different contexts (Habermas 1971: 1314). Most relevant in the present instance is
the methodological decision he makes in favour of subjective meaning as a sociological primitive (Habermas 1971: 310). This decision leads him to view society
as a continuous process of generating a meaningfully structured reality, whereby
meanings are produced according to an underlying system of abstract rules and
have a sociologically relevant, intentional component. Having generally endorsed
theories that take subjective meaning as an irreducible element of sociological
analysis, Habermas now makes a distinction within this set between constitutive
and communicative theories.5 It is at this point that pragmatics enters the scene.
In his account, constitutive theories base the process of the generation of meaning on a transcendental subject. Some, taking their lead from Husserls account of
the constitution of the everyday world of lived experience (the lifeworld), conceive of this transcendental subject as an intelligible ego modelled on the empirical
individual subject. Others, taking their lead from Hegel and Marx, conceive of it as
a species-subject that constitutes itself in history. In each case, they encounter the
problem of how to move from the level of experience to the level of society. More
precisely, they run up against the problem of how adequately to conceptualize intersubjective meanings within a subject-object model of cognition and action. In
models of this kind, knowledge and action are conceived monologically, as the result of a relationship between a subject and an object that is distinct from it. However, if meaning is conceived of essentially as the product of a relationship
between a subject and some distinct object, commonly shared, intersubjective
meanings appear to have a secondary status. From a social-theoretical point of
view this gives rise to considerable difficulties. One difficulty is that it makes it
hard to explain how society, as a process of generating intersubjective meanings,
can be granted the primordial status it has in phenomenological accounts. This
weakness of constitutive theories, together with other, related ones, prompts Ha-
292
Maeve Cooke
bermas to look to communicative theories, which construe meaning as intersubjectively generated from the beginning.
In this preliminary essay, Habermas finds himself in the unwelcome situation
of being unable to refer to an existing body of work that would support the kind of
communicative theory he is seeking; he admits that the theoretical approach that he
deems methodologically most appropriate for social scientific inquiry has not yet
been adequately developed. However, he is clear as to the direction it is necessary
to take. What is required in his view is a sociological theory that would start from
the premise that everyday communication, as a process of generating shared meanings, is the basis for social action (and socialization); the use of ordinary language
would be seen as constituting the intersubjective relations within which action
takes place (as well as the personality structures of the subjects of action). His
name for this project is universal pragmatics.
Habermas programmatic essay What is Universal Pragmatics? first appeared in German in 1976. The adjective universal indicates the difference between his reconstructive approach to the study of linguistic behaviour and empirically oriented analyses. In his account, empirical-pragmatic approaches set
themselves the task of describing the speech acts typical of a certain milieu, which
may then be analyzed from sociological, ethnological and psychological points of
view. Habermas favours a broader perspective. He aligns himself with general
pragmatic theory, which is concerned not with specific instances of language use,
but rather with formally reconstructing the rule system that underlies the ability of
a subject to utter sentences in any relevant situation whatsoever (Habermas 1998:
54). Accordingly, he characterizes the task of universal pragmatics as being to
identify and reconstruct universal conditions of possible Verstndigung, which we
can translate here as mutual understanding6 (Habermas 1998: 21). Soon after the
publication of this essay, however, he distanced himself from the name universal
pragmatics. In a footnote to the English translation, published three years later in
1979, he writes that he is no longer happy with his original terminology, suggesting
that his project instead be named formal pragmatics. In opting for the word formal he wishes to emphasize the continuity between his approach and formal semantics, notwithstanding his focus on the use of language in speech acts or utterances as opposed to the semanticists concern with the properties of isolated
sentences.7 His choice of the word formal also makes sense in the context of his
adoption of the procedure of rational reconstruction which, as indicated, is the
method he attributes to his universally oriented, pragmatic approach. By rational
reconstruction he means the systematic reconstruction of the intuitive knowledge
of competent subjects (Habermas 1998: 2941). Reconstructions relate to pretheoretical knowledge of a general sort, to universal capabilities and not merely to the
competences of particular groups or individuals. Thus, what begins as an explication of meaning ultimately aims at the reconstruction of species competences.
Habermas sees it as the great merit of Noam Chomsky to have developed this idea
293
in the case of grammatical theory (Chomsky 1969). He also commends Jean Piaget
and Lawrence Kohlberg for, respectively, their theories of cognitive and moral development, which, like Chomskys, rationally reconstruct the pretheoretical and
implicit knowledge and competences of speaking and acting subjects in general.
In distinguishing his formal-pragmatic approach from an empirical-pragmatic
one, Habermas does not want to deny the importance of empirical validation. As
indicated, reconstructive theories belong to the category of general theory in the
sense that they seek to reconstruct the rule system that underlies the ability of a
subject to generate well-formed sentences in any language whatsoever. At the
same time, they belong to the category of empirical theories. Their explications of
general structures and universal conditions have a merely hypothetical status and
must be subjected to the usual methods of testing. Thus, Chomskys and Habermas reconstructions of linguistic competence are open to checking against speakers intuitions, scattered across as broad a sociological spectrum as possible (Habermas 1984:138). As is appropriate for scientific investigations of this sort, the
checking procedure is circular: it moves in a circle between theory formation and a
more precise rendering of the object domain (Habermas 1998: 40).
From this it is clear that Habermas places reconstruction, as an explicative exercise, on the side of interpretation rather than explanation, thereby connecting it
with an internal rather than external perspective on social phenomena. The point of
reconstructions is neither to describe behaviour nor to systematize their findings in
law-like explanations, but rather to bring to consciousness implicit knowledge
through a maieutic method of interrogation, which involves the choice of suitable
examples and counterexamples, the use of contrast and similarity relations, translation, paraphrase and so on (Habermas 1998: 40). This is not to deny that the procedures employed in constructing and testing hypotheses, in appraising competing
reconstructive proposals and in gathering and selecting data, are not in many ways
like the procedures used in the nomological sciences (Habermas 1998: 46). Like
such sciences, reconstructive theories usually operate at a high level of abstraction,
seeking to explicate basic conceptual frameworks and to formalize assumptions
built into everyday behaviour. The key difference, to recall our earlier discussion,
is that reconstructive social theories take subjective meaning as an irreducible element of social scientific analysis whereas theories that adopt a purely external
perspective do not.
We have seen that Habermas turns to pragmatics in order to find a method of
sociological analysis that takes subjective meaning as a sociological primitive,
while avoiding the weaknesses of constitutive models which attribute the process
of generating a meaningfully structured reality to some sort of transcendental subject. His aim is to develop a communicative model that replaces the subject-object
paradigm underlying constitutive theories with an intersubjective paradigm in
which the generation of shared meanings is seen as fundamental in the constitution
of social reality. He sees the use theory of meaning developed by Ludwig Wittgen-
294
Maeve Cooke
stein in his later work as making an important contribution towards developing

such a model (Wittgenstein 2001). But for reasons that will become clearer in the
next section, he finds speech act theory as developed by Austin and by Searle (in
his early work) more fruitful for his purposes.
Speech act theory shares with reconstructive language analysis a concern to explicate the rules that a competent speaker must master in order to form grammatical sentences and to utter them in an acceptable way. However, by contrast with
Chomskian linguistics, instead of starting from the postulate of an implicit ability
to produce sentences, speech act theory starts from the assumption of an implicit
ability to employ sentences in speech acts (utterances). Thus, it is concerned with a
competent speakers ability to communicate as opposed to produce a grammatical
sentence. Accordingly, a general theory of speech acts explicates the system of
rules that competent speakers master when they fulfil the conditions for the successful employment of sentences in utterances, no matter to which particular languages the sentences may belong and in which random contexts the sentences may
be uttered (Habermas 1998: 47). The importance of speech act theory in the development of Habermas conceptual strategy is twofold. First, by virtue of its focus on
utterances as communicative actions, it enables him to show how the meaning of
speech acts is produced by way of an action involving speaker and hearer(s); this
provides support for his shift in paradigm from a subject-object based, monological model of meaning production to an intersubjective one. Second, it enables him
to show how the production of meaning is internally connected with a number of
different kinds of claims to validity. The connection between meaning and validity
in a number of dimensions emerges from speech act theorys analysis of the successful employment of sentences in utterances. Habermas contends that for an utterance or speech act to be successful it must count as true for the speaker and hearers, in the sense of representing something in the world, must count as normatively
right, in the sense of conforming to socially recognized expectations and must
count as truthful, in the sense of expressing something intended by the speaker. As
we shall now see, the connection between meaning and validity claims is of crucial
importance in Habermas accounts of social action, social order and social reproduction and in his proposal for a postmetaphysical conception of reason that permits social critique.
3.
Social action, social order and social reproduction
Formal pragmatics forms the basis for the kind of communicative theory that Habermas thinks necessary in order to avoid the weakness of constitutive social theories and those that share their subjectivist orientation. In his view, such theories
not only have difficulties adequately conceptualizing meanings shared in common;
they are also not well suited to the task of adequately conceptualizing social action,
295
social order and social reproduction. This is because they tend to privilege instrumental action, understood as action undertaken as a means to achieving an end that
is specified in advance of the undertaking. Accounts of social order and reproduction that take instrumental action as the fundamental mode of social action run up
against the problem of how to explain long-term social cohesion, when the preservation and reproduction of the social order in question does not evidently equally
enable everyone to obtain their particular ends (cf. Habermas 1998: 234). A central
aim of Habermas theory of communicative action is to show that instrumental action is not the primary mechanism in social cohesion and social reproduction. He
draws on formal-pragmatic analyses of everyday language in order to achieve this
aim.
As the background to our discussion in this section and the next, it is important
to note that in The Theory of Communicative Action Habermas offers an account of
the developmental path of modern societies in terms of two categorially distinct
modes of societal integration. In what he calls the lifeworld, co-ordination of action takes place primarily by way of communicative action, which relies on intentionally acting human subjects who establish validity-based relationships with one
another. In the system, by contrast, co-ordination of action takes place primarily
by way of the functional interconnection of action consequences, bypassing the intentions of the agents concerned. For Habermas, rationalization processes, by
which he means the development of the internal logic of a particular mode of societal integration, are possible and desirable in each case. The rationalization of the
system refers to increasing complexity and increasing capacity to take on steering
functions such as material reproduction and administration. By contrast, the
rationalization of the lifeworld is a matter of communicative rationalization: the
development of the internal logic of communicative action. It takes place in the domains of cultural tradition, social integration and socialization and entails the increasing independence of processes of justification from traditional normative contexts of validity, together with increasing reliance on action oriented to achieving
understanding (Verstndigung).
Communicative action is a form of social interaction in which the plans of various agents are co-ordinated through an exchange of speech acts. Drawing on the
speech act theory of Austin and the early Searle, Habermas maintains that the success of a speech act is not just a matter of comprehending the sentence embedded in
the utterance, but of responding in the appropriate manner to what the speaker does
in performing the action. In addition to commending these speech act theorists for
emphasizing the action component of linguistic understanding, he praises them for
directing attention to modes of language use other than the assertoric and descriptive modes. In this respect, too, they take their lead from the later Wittgenstein.
However, Habermas finds speech act theory more promising than Wittgensteins
use theory of language, due to its assertion of a connection between meaning and
validity in a context-transcending sense. For Habermas, truth and truth-analogous
296
Maeve Cooke
concepts such as moral rightness have a context-transcending aspect in the sense

that they assert a validity that transcends particular places and times. This is one
reason why he refers to them as universal validity claims.8 The connection between
meaning and universal validity disappears from view in Wittgensteins theory, since
validity is held to be relative to particular, historically specific, language games. To
be sure, Habermas finds Austins and Searles approaches to universal validity
claims unsatisfactory. He criticizes Austin for blurring the distinction between clear
cut universal validity claims such as truth and normative rightness and a host of
other evaluative criteria (Habermas 1998: 290291). He admonishes Searle for acknowledging only the representational function, together with the claim to truth
corresponding to it, as the only clear cut universal validity claim (Habermas 1998:
291292). Against Austin and Searle Habermas seeks to demonstrate that everyday
language use is connected with three clear cut universal validity claims.
This is one of the two main respects in which Habermas moves beyond speech
act theory. The other is his emphasis on the relationship of obligation into which
the speaker and hearer(s) enter with the performance of a speech act. Taking his initial lead from Austin, he finds it instructive to consider what a speaker does in saying something. He begins by discussing Austins distinction between locutions and
illocutions (Austin 1962). Initially Austin reserved the term locution for the
propositional content of what is said and the term illocution for the performative
act of uttering sentences with a propositional content. He later reformulated the
distinction as one between constatives and performatives, defining constatives as
speech acts in which we concentrate on the truth or falsity of what is said and performatives as speech acts in which we concentrate on the illocutionary force of the
utterance (Habermas 1998: 6681). Habermas finds this way of distinguishing between constative and performative speech acts unhelpful, proposing instead that
we distinguish between utterances on the basis of the kinds of validity claims they
raise. In his reformulated version of Austins classificatory schema, constatives involve the cognitive use of language and raise claims to truth (examples are assertions and descriptions); here we thematize the content of the utterance as a statement about what is or could be the case. Regulatives involve the interactive use of
language and raise claims to normative rightness (examples are promises, requests
and warnings); here we thematize the relationship of obligation into which speaker
and hearer enter with the speech act. Habermas adds a third category, expressives,
to these revisions of Austins schema; as the name indicates, these involve the expressive use of language and raise claims to truthfulness (examples are confessions
and disclosures); here we thematize the truthfulness or sincerity of the speakers
intentions. He proposes, further, that we understand the illocutionary force of the
utterance what the speaker does in performing any speech act as a rationally
binding force (bindende Kraft), in the twin sense of binding and compelling (Habermas 1998: 8188). This rationally binding force is held to be operative in all
three modes of language use: the constative, regulative and the expressive.
297
This, then, is Habermas answer to the question of what we do in performing a

speech act: we enter into a relationship of rationally-based obligation with the
hearers. The obligation is rationally-based since it arises from the speakers undertaking to support what she says with reasons. He conceptualizes this in terms of a
validity claim raised by the speaker. The speaker claims that what she says is true
or that she is normatively right to say it or that she is sincere in saying it, thus
raising a claim to validity that calls for a yes or no response from the hearer.
The success of the interaction depends on the speakers ability to supply reasons to
support the validity of the claim, if challenged, and on the hearers ability to provide reasons in support of her yes or no. For Habermas, therefore, the illocutionary force of an utterance is not simply an aspect of its meaning conveyed by a
performative verb; it is a co-ordinating power that is based on reciprocal obligations to provide reasons. This is what he means by the rationally binding force of
linguistic interaction.
Habermas holds, furthermore, that every speech act raises all three claims simultaneously. More precisely, he holds that with every speech act, the speaker raises
one of these three kinds of validity claims directly, while the other two are raised
indirectly. Thus, for example, if a speaker states she has no money to pay the rent,
she is raising a claim to truth that calls for a yes or no response from the hearer(s) or for an abstention on grounds of insufficient knowledge of the facts. At the
same time she raises a claim that her utterance is appropriate in the context in question (normatively right) and that she is sincere in uttering it. In this case, a claim
to truth is raised directly and claims to normative rightness and sincerity are raised
indirectly. If a speaker requests a hearer to lend her some money, she raises a claim
to normative rightness directly and claims to truth and truthfulness indirectly (Habermas has clarified that indirect claims to truth should be understood as claims
that certain existential presuppositions are met, 1998: 146). A speaker who confesses that she has spent all her money raises a claim to truthfulness directly and
claims to normative rightness and truth indirectly.
The thesis that every speech act raises three claims to validity simultaneously is
important for Habermas attempt to show that an orientation to a plural multi-dimensional conception of validity is built into everyday language use. As we shall
see, this plural conception of validity is the basis for his non-logocentric conception of communicative rationality, which does not privilege propositional truth but
conceives of reason as an interplay of truth, moral rightness and truthfulness/authenticity. The thesis of three simultaneous validity claims is also important from the
point of view of social integration and reproduction. Its purpose is to show that the
three validity claims raised with every speech act amount to three structural components of speech acts that correspond to the three functions of the communicative
use of language: (a) representing states of affairs (or something in the world that
confronts the speaker), (b) entering into a relationship with a hearer and (c) giving
expression to the speakers intentions or experiences. Referring to the communi-
298
Maeve Cooke
cation-theoretic approach expounded by the German psychologist Karl Bhler

(Bhler 1934), Habermas posits these as three mutually irreducible but internally
connected linguistic functions.9 His thesis is that these linguistic functions, carried
out with every speech act, serve to maintain and reproduce the lifeworld in its three
domains of cultural tradition, social integration and socialization.
For this, however, the thesis of three simultaneous validity claims is not sufficient. Even if successful (see Cooke 1994: 9091), it would show merely that the
communicative use of language fulfils functions of representation, of establishing
a relationship with the hearer and of expression; it would not show that the communicative use of language is of fundamental importance in maintaining and reproducing the lifeworld. Showing the latter is central to Habermas critical project.
We will recall his claim that the colonization of the communicatively regulated domains of the lifeworld by the functional rationality of the system is a deeply troubling pathology of modernity. If he were unable to show that communicative action
is indispensable for the purposes of establishing, maintaining and reproducing
social order, he would lack grounds for referring to the shrinking of the communicatively regulated domains of the lifeworld as a pathology and for adopting the
metaphor of colonization to describe the expansion of functional rationality into
the communicatively regulated domains. In order to meet this challenge, Habermas
seeks to demonstrate the primacy of communicative action over instrumental (and,
in particular, strategic) action. Once again, he employs a formal-pragmatic strategy. This entails showing the primacy of the communicative use of language over
other modes, specifically the strategic one.
Just as the communicative use of language amounts to communicative action,
the strategic use of language amounts to strategic action. Strategic actions are instrumental insofar as one agent uses another as a means of achieving his respective
goals; he treats other persons as though they were objects or entities in the physical
world. In contrast to communicative action, in which the success of the interaction
is not at the disposal of the individual agent but depends on all parties involved,
strategically acting agents are not dependent on the co-operation of others (Habermas 1998: 220227). While Habermas acknowledges that strategic action, like
communicative action, is a mode of action within the lifeworld, he insists that it is a
merely secondary mode of social integration. His position, in other words, is that
communicatively regulated interactions must predominate in the lifeworld if this is
to maintain and reproduce itself. He seeks to show the priority of communicative
action by demonstrating that the strategic use of language is parasitic on the communicative one.10
In The Theory of Communicative Action Habermas attempts to show that the
use of language oriented to reached understanding (Verstndigung) is the primary
or original mode of language use by drawing on Austins distinction between locutions and perlocutions (Habermas 1998: 122129) In Austins account, the locutionary act is the act of saying something, the illocutionary act is the action per-
299
formed by saying something and the perlocutionary act is the effect the speaker
produces upon the hearer. Initially, Habermas took this to mean that perlocutionary
acts are an indication of the integration of speech acts into contexts of strategic action (Habermas 1998: 126). This interpretation neglects the importance of distinguishing between different types of perlocutions. It is important, above all, to distinguish between the contingent, often unintended effects on the hearer(s) of
particular speech acts and the intended, but concealed effects of particular speech
acts. An example of the former is the pleasure a speaker gives to his hearers husband by lending him money for the rent (for the act might have given him displeasure); an example of the latter is the speakers intention to use his loan of
money to blackmail the hearer. Habermas has subsequently modified his position,
clarifying that only perlocutionary effects of the latter kind are relevant for his argument (Habermas 1998: 329333). He sees them as examples of latently strategic
action. As illustrated by the case where the speaker intends to blackmail the hearer
by lending her money, the success of latently strategic action depends on the speakers success in hiding his intentions from the hearer; he must successfully pretend
that he is using language with an orientation towards mutual understanding (Verstndigung), as opposed to strategically, using the interaction with the hearer as a
means to furthering his own particular ends.
The latently strategic mode of language use is by definition parasitic on the
communicative use (Cooke 1994: 24).11 Thus, demonstrating its secondary character will not be sufficient for Habermas purposes. The more interesting and difficult case is the manifestly strategic use of language. In this mode of language use,
the presupposition of an orientation towards reaching understanding is overtly suspended (Habermas 1998: 225226). A speaker who backs up his demand that a
hearer lend him some money with the threat of sanction (for example, by threatening to harm her daughter) illustrates such language use. In order to make good his
claim that the communicative use of language is the primary mode of language use,
Habermas has to show that this manifestly strategic use of language, too, is parasitic on the communicative one. This part of Habermas argument is underdeveloped (Cooke 1994: 2326).
In this section we have considered the importance of formal pragmatics in Habermas accounts of social action, social order and social reproduction. We have
seen that its role is threefold: to demonstrate that everyday linguistic communication has an in-built rationally binding force, to show that this binding force is operative in three dimensions corresponding to the three linguistic functions of representation, establishing a relationship with the hearer(s), and expression, and to
demonstrate that the communicative use of language is the primary mechanism in
establishing, maintaining and reproducing the social order. In the next section, we
will look at the importance of formal pragmatics for Habermas project of social
critique.
300
4.
Maeve Cooke
Social critique
Habermas is first and foremost a critical social theorist. His sociological concerns
are intimately bound up with his concern to diagnose and criticize the pathologies
of modernity. We have already mentioned the colonization of the lifeworld as one
of these pathologies. Habermas, drawing on Weber, sometimes refers to this as the
modern loss of freedom (Habermas 1987: 301302). The second principal pathology he diagnoses is a loss of meaning (here, too, he draws on Webers analysis of
modernity). By this he means the cultural impoverishment of the lifeworld that results from the differentiation of the cultural value spheres of science, law/morality,
and art under conditions of modernity; these value spheres become fields for the
specialized treatment of questions of truth, normative rightness and beauty/authenticity by experts. While these specialized discourses promote a growth in cognitive-instrumental, moral-practical and aesthetic-expressive knowledge, thus giving rise to significant learning processes, they are also a potential source of
pathology. This is because with progressive modernization, the gap between the
expert cultures and the general public becomes ever wider. Learning processes
within the expert cultures do not automatically flow back into the communicative
practices of the lifeworld, contributing to their renewal and regeneration, but remain cut off from them. Lacking revitalizing impulses from the specialist cultures
(or, indeed, from other sources12), everyday communicative practices threaten to
become impoverished.
His critique of these pathologies is guided by a corresponding normative vision. At the core of this vision is a conceptual triad that has defined his work from
his earliest major publication The Structural Transformation of the Public Sphere
(Habermas 1989): the concepts of the public sphere, discourse and reason. The interplay of these concepts yields a utopian vision of a communicatively rationalized
lifeworld (Habermas 1991: 6469). In the projected society, cultural traditions
would be reproduced through processes of intersubjective evaluation of validity
claims, legitimate orders would be dependent on critical and open argumentative
practices for justifying laws/norms and for making political decisions, and individual identities would be self-regulated through processes of critical reflection. We
must assume, furthermore, that in such a communicatively rationalized lifeworld
there would be ongoing processes of semantic renewal in the form of continuous
flows of knowledge from the specialized, expert cultures into the communicative
practices of everyday life.13 This utopian projection of a rationalized lifeworld is in
turn part of a vision of society in which the development of systems rationality and
communicative rationality would be evenly balanced: systems rationality would
no longer encroach into the communicatively regulated domains of the lifeworld.
Formal pragmatics contributes to Habermas critical project in two main respects. His formal-pragmatic investigations are intended, first, to provide a normative reference point for the critique of the two principal pathologies of modern
301
societies: the colonization of the communicatively regulated domains of the lifeworld by systems rationality and its cultural impoverishment due to the emergence
of specialized discourses among experts in the areas of science, law/morality and
art. They are intended, second, to provide a normative reference point for criticism
of the conduct and outcomes of actual practices of deliberation in everyday life. In
each case, the normative reference point is given the name communicative
rationality.
As we have seen, Habermas thesis of the colonization of the lifeworld makes
the claim that the uneven developmental path of modern societies is pathological.
What is required, instead, is an evenly balanced pattern of societal rationalization.
Habermas expresses the idea of a non-pathological, evenly balanced process of societal rationalization in terms of the harmonious, free interplay of the three cultural
value spheres of science, morality/law and art, in which no one sphere would predominate at the expense of the other (Habermas 1984: 240). His formal-pragmatic
investigations are used as a normative underpinning for this idea.
The relevant formal-pragmatic finding is that everyday linguistic communication fulfils three mutually irreducible functions: representation, establishing a relationship between speaker and hearer(s), and expression; as we have seen, this is
the conclusion Habermas draws from his thesis that the speaker raises three validity
claims simultaneously with every speech act. He asserts, in addition, that these
three linguistic functions can be correlated with the kind of knowledge defining
each of the three cultural value spheres: scientific knowledge, moral-practical
knowledge and aesthetic-expressive knowledge. He then makes a further correlation between the three cultural value spheres and the three moments of a once substantively unified concept of reason, which have become separated from one another under conditions of modernity;14 the image of harmonious, free interplay is
extended to these three moments of reason, evoking an idea of reason whose unity is
defined purely formally as opposed to substantively, and that is non-logocentric, expanding the traditional philosophical focus on theoretical reasoning and truth to include other dimensions of validity (such as justice and beauty) on an equal footing.
If the colonization thesis suggests as its opposite evenly balanced processes of
functional rationalization and communicative rationalization, the thesis of cultural
impoverishment suggests as its opposite the continuous feedback back into everyday communicative practices of the knowledge produced by experts in their various fields. However, whereas Habermas evidently wishes to use his formal-pragmatic analyses to provide empirically based, normative support for the idea of
evenly balanced processes of societal rationalization, he gives no indication of any
formal-pragmatic finding that would lend support to the idea of the free flow of
knowledge between expert cultures and the lifeworld. Nor is it easy to see which
part of his analyses could help him in this regard.
In addition to these two principal pathologies he diagnoses and attributes to
deficiencies in modern processes of societal rationalization, Habermas identifies a
302
Maeve Cooke
number of further potential pathologies (Habermas 1987: 142144); by contrast

with the pathology of colonization, which threatens it from the outside, these potential disturbances arise within the lifeworld (in this they are similar to the pathology of cultural impoverishment). In addition to the loss of meaning that results
from the encapsulation of specialist cultures, he mentions the anomie that results
from disturbances in the process of social integration and the mental illnesses (psychopathologies) that result from disturbances in the process of individual socialization. At least initially, Habermas held that formal pragmatics could play a role in
the critique of these disturbances. In The Theory of Communicative Action and in
the preliminary studies that led up to it, he suggested that formal-pragmatic analyses could help to establish a normative basis for criticizing the phenomena in
question as pathologies. The specific strategy he envisaged was a formal-pragmatic
account of systematically distorted communication (Habermas 1974). Despite occasional remarks in his writings up to the middle of the 1980s that indicate that he
had not yet abandoned this strategy, he has not since then made any serious attempt
to develop such an account. His key intuition seemed to be that systematically distorted communication violates what formal pragmatics shows to be the internal organization of speech, by undermining the three linguistic functions of representation, establishing a relationship between speaker and hearer(s) and expression.
The violation, which takes the form of disconnecting meaning from validity, speaking from acting and/or meaning from intention, is due to the overwhelming pressure exerted on the internal organization of speech by its external organization; by
the external organization of speech he means the regulation of the normative context in which the communicative interaction takes place: for example, the determination of who is allowed to take part in which discussion, who can initiate topics,
who can bring the discussion to a close, who can contribute and in which order, how
the topics are ordered, and so on. To be sure, in order to develop a convincing account of systematically distorted communication, Habermas would have had to
embed it within his critical account of societal rationalization as colonization (at
the time of publication of his 1974 essay, he had not yet provided such an account);
specifically, he would have had to show how the infiltration of functional rationality into the communicatively regulated domains of social life affects the external
organization of speech, contributing to a violation of its internal organization. Formal-pragmatic arguments would not be helpful here. Nor would they be able to
ground the normative force of the thesis implicit in Habermas account of systematically distorted communication: the thesis that the regulation of the context in
which speech takes place should conform to demanding normative standards of inclusivity, fairness and openness. In short, even a well-developed formal-pragmatic
account of systematically distorted communication would show only what the internal organization of speech and its disruption would look like; on its own it would
not be sufficient to provide the sought-for normative basis for critique of disturbances such as anomie and psychopathologies (Cooke 1994: 148150).
303
As indicated, the normative reference point for Habermas diagnosis and

critique of social pathologies is the concept of communicative rationality. Habermas sets himself the task of elaborating a concept of reason that responds to two
(closely connected) challenges facing contemporary critical social theorists:
maintaining a dialectics of immanence and transcendence and maintaining concepts of validity that transcend times and spaces while acknowledging the historical situatedness of all actual conceptions of context-transcending validity. The
first is specific to theorists in the Frankfurt School tradition who take seriously the
Hegelian thesis that reason is embedded within history itself; the second arises for
any social theorists today who seek to uphold an idea of reason while taking on
board the post-Enlightenment suspicion of transcendental ideas such as truth, justice and the good. Formal pragmatics plays a role in meeting each of these challenges.
The term Left-Hegelian is sometimes used to characterize modes of critical
social theory that, taking their lead from Marx and Lukcs, are concerned to re-articulate Hegelian speculative thinking in a materialist way. Most critical social theorists in the Frankfurt School tradition fall into this category. The aspect of Hegels
thinking most relevant here is his emphasis on reasons simultaneous immanence
and transcendence. Hegel endeavours to show a) that reason is immanent to human
history in the sense that the latter is a process in which reason actualizes itself by
way of the progressive overcoming of contradictions and b) that reason always
transcends its particular historical self-actualizations.15 The attempt to maintain a
dialectical relationship between immanence and transcendence is taken up explicitly by Theodor W. Adorno, Max Horkheimer, Herbert Marcuse and other first
generation critical social theorists (see, for example, Adorno 1981) and continues
to play a central role today in the writings of Axel Honneth, the most prominent
member of the third generation of critical social theorists (Honneth 1994, 2003).
Habermas, the most prominent member of the second generation is less inclined to
describe his project in explicitly Left-Hegelian terms; nonetheless, a concern to
sustain a dialectics of immanence and transcendence is certainly also one of its
most important elements (see, for example, Habermas 1996:125). In Habermas
work, this amounts to the elaboration of a conception of reason that is at once context-transcending, in the sense of transcending times and places, thus extending to
human beings in general, and immanent, in the sense of being anchored in practices
within everyday historical practices, specifically in everyday practices of using
language communicatively.
Habermas theory of communicative action is of crucial importance to his attempt to elaborate a conception of reason that is at once immanent and transcendent. We have seen that he uses a formal-pragmatic analysis of everyday language
use as the basis for an account of social action, order and reproduction that has a
normative as well as an empirical content. It is empirical in the sense that it uses
empirical evidence to reconstruct the relevant structural features of everyday lan-
304
Maeve Cooke
guage use; it is normative in the sense that the resulting account of language is used
as the basis for a normative idea of a communicatively rationalized lifeworld that
co-exists harmoniously with a well-functioning system. An analogous interplay of
the empirical and normative characterizes the concept of communicative rationality. This concept expresses the rational potential implicit in immanent to everyday communicative action. As a practice of raising and responding to validity
claims, communicative action is conceptually tied to more or less rudimentary
practices of argumentation. Habermas maintains that the validity claims raised in
communicative action have a context-transcending, critical force that derives from
idealizations built into the concept of argumentation (see Cooke 2006: 4753).
One set of idealizations refers to the procedure of argumentation: to the normative
promise contained in the general understanding of how argumentation should be
conducted. Examples here are the idealizing presuppositions that all participants
are motivated by a concern to find the single right answer, that no force is exerted
except that of the better argument, that all relevant arguments are considered, that
no one should be excluded from the deliberation on grounds such as race, class and
gender and that everyones voice should be given an equal hearing. Another set
refers to the outcome of argumentation: to the normative promise contained in the
general understanding of the aim of argumentation. An example here is the idealizing presupposition that a discursively reached agreement warrants the truth of
propositions and constitutes the justice of norms. In both cases, the tension between the normative promise contained in these idealizations and what actually
happens in everyday communicative practices provides a basis for criticism: in the
one case, they permit criticism of the ways in which the outcomes of argumentation are reached; in the other case, they permit criticism of the outcomes from the
point of view of an idea of context-transcending validity (e.g. truth or justice). The
critical power of communicative action resides primarily in the tension between an
idealized notion of argumentation and what happens in actual empirical practices.
It should be noted that this critical power is not restricted to a particular socio-cultural context. Since it is grounded in universal features of language use, it expresses a critical perspective with context-transcending force, in the sense that its
validity would have to be accepted by everyone, everywhere, irrespective of sociocultural context. It should be noted, too, that the critical power of the concept of
communicative rationality refers to an idealized practice of communication that
has the status of a methodological fiction. Habermas warns against essentialist
misunderstandings of the ideal speech situation, emphasising that it is merely an
idealizing projection and best understood as enabling a fruitful thought experiment
(Habermas 1996: 322323). In this way, his formal-pragmatic analyses of language enable Habermas to assert an idea of reason that is at once immanent to history, in the sense that it is built into everyday practices of linguistic communication, and transcendent, in the sense that it provides a reference point for critique
of all actually existing communicative practices.
305
As part of his engagement with the post-Enlightenment critique of reason, from

the Romantics, through Nietzsche and historicism, up to contemporary post-structuralism, Habermas emphasizes the postmetaphysical character of the concept of
communicative rationality. The role played by formal pragmatics here is very similar to its contribution to sustaining a dialectics of immanence and transcendence.
By postmetaphysical he means, principally, that the concept of communicative
rationality breaks with traditional views of the substantive unity of reason; refrains
from specifying the content of truth, justice and the good life; moves from a subject-object model of cognition to an intersubjective one; gives up the traditional
fixation on truth and theoretical reason; and renounces all dependency on otherworldly projections such as God or the Good. Our discussion in the foregoing has
cast light on how his formal-pragmatic investigations help him to achieve these
aims. The reference to reason built into everyday language use is a formal as opposed to substantive conception of reason, for it amounts only to a procedure for
determining what is true or right and leaves open the question of the respective
content of these concepts; it is non-subjectivist in the sense that it starts from interaction in communicative practices rather than an individual subjects relation to
an object independent of it; it is non-logocentric in the sense that it does not assert
the primacy of theoretical or propositional truth but comprises a plurality of dimensions of validity that interact with each other and relate to each other on an
equal footing; it is historically rooted in the sense that it is contained within actual
practices of using language; and it is innerworldly in the sense that the ideal reference point to which it appeals is not some otherworldly idea of God or the
Good but an idealized projection of a speech situation in which the idealizing presuppositions relating to the conduct of intersubjective deliberation would be realized in practice and in which the outcomes of such deliberation would contribute
constructively to validity in a context-transcending sense.
In this section we have considered the ways in which formal pragmatics enables Habermas to develop a concept of communicative rationality that is the indispensable normative reference point for his critical social theory, both with regard to the critique of social pathologies and with regard to criticism of the
shortcomings and distortions of actual practices of intersubjective deliberation. In
my final remarks I will outline the principal lines of objection that have been raised
against Habermas use of formal pragmatics as the normative foundation for his
critical social theory.
5.
Objections
There are three principal lines of objection to Habermas use of formal pragmatics
as the underlining for his critical project. The first casts doubt on the ability of formal pragmatics to provide a foundation for social critique that is sufficiently robust
306
Maeve Cooke
for his purposes. The second criticizes the specific shape of the conception of reason that emerges from his formal-pragmatic analyses. The third claims that his appeal to formal pragmatics, and the vision of a communicatively rationalized lifeworld that he derives from these analyses, amounts to a foreshortened idea of
human flourishing.
1) The first line of objection challenges the ability of formal pragmatics to provide an empirically based, normative foundation for social critique; this objection
can be further divided into three sub-objections: a) an objection to the universality
of the idealizations built into everyday communicative practices; b) an objection to
the idea of the harmonious, free interplay of the three cultural value spheres/moments of reason as a standard for criticizing the colonization of the lifeworld; c) an
objection that Habermas does not show the primacy of the communicative use of
language over the strategic one.
a) The criticism here is that the concept of communicative action that Habermas
extracts from his formal pragmatic analyses is too weak for the critical purposes he attributes to it (Benhabib 1992; Cooke 1994; Cooke 2006). Some of
the idealizations he identifies as necessary presuppositions of the communicative use of language are not a feature of language use in general; they orient
communicative practices only in certain socio-cultural contexts as a result of
specific historical developments; in fact, they seem to be features of language
use only under conditions of modernity, in societies in which there has been a
desacralization of knowledge, a secularization of authority and an internalization of the principle of universal moral respect. Since the critical force of the
concept of communicative rationality depends to a considerable degree on
idealizations relating to social inclusiveness and equality, it is particularly
worrying that they appear to be features only of certain kinds of modern society. This presents Habermas with the following dilemma: on the one hand, if
he wants to maintain a formal-pragmatic basis for his critical perspective, he
will have to acknowledge the socio-cultural specificity of the resulting concept
of communicative rationality, significantly restricting the scope of its critical
power; on the other hand, if he wants to offer a critical perspective that is not
restricted to particular socio-cultural contexts but rather universal in scope, he
will have to supplement his formal-pragmatic analyses with arguments of other
kinds (Cooke 2006: 5761).
b) This objection queries the usefulness of the metaphor of the harmonious, free
interplay of the three moments of reason for the purposes of critique of the colonization of the lifeworld by functionalist rationality. It points out that Habermas critique of colonization appeals to a normative idea of evenly balanced,
non-selective processes of societal rationalization that calls for a balanced relationship between communicative rationality and functional rationality. The
difficulty is that the balance evoked by the metaphor of three harmoniously interrelating moments of reason is not a balance between communicative
307
rationality and functional rationality, but rather a balance between cognitiveinstrumental, moral-practical and aesthetic-expressive rationality. For the
metaphor to work, we would have to be able to equate functional rationality
with cognitive-instrumental rationality. Habermas, however, insists that they
are not equivalent: whereas instrumental rationality is the reasoning employed
by social agents to achieve particular, intended ends, functional rationality bypasses the intentions of the agent (see Cooke 1994: 144146).
c) The nub of this objection is that in order to assert the primacy of the communicative use of language Habermas needs to show that manifestly strategic action, too, is parasitic on the communicative use. The objection is not just that
Habermas shows only the parasitic nature of latently strategic action; it is the
stronger point that it is impossible in principle to show which use of language is
most fundamental without appealing to controversial philosophical presuppositions concerning the rational use of language; indeed it is impossible to decide whether there is a fundamental use of language at all (Apel 1998). In other
words, formal-pragmatic arguments will never suffice to establish the primacy
of any mode of language use. A related objection is that even successful demonstration of the primacy of the communicative use of language would leave
open a number of important questions; these include questions such as how
much communicative action is necessary for the maintenance of social order
and social reproduction, the extent to which the erosion of communicative action can count as colonization and the degree of reflexivity, inclusiveness and
openness communicative action has to display in order for social reproduction
to count as successful (Cooke 1994: 2428).
2) The second line of objection relates to the specific shape of the concept of communicative rationality that emerges from Habermas formal-pragmatic analyses.
Critics have drawn attention, in particular, to his sharp separation between the use
of language oriented to mutual understanding and the use of language to disclose
new perspectives (Cooke 1994: 7484). This produces an impoverished account of
communicative reason that excludes modes of language use that bring us to see the
world in a new way from the category of validity-oriented and, hence, rational communication; ultimately, it leads to an inability to explain the power of critical social
thinking to engage the imagination as well as the reason of its addressees, thereby
motivating them to work towards social change for the better (Cooke 2006).
3) The final line of objection is directed at the very project of formal pragmatics. One of the strongest criticisms of Habermas in this regard comes from
within the Frankfurt School tradition of critical social theory. Honneth, described
above as a third generation Frankfurt School critical theorist, commends Habermas
for the paradigm shift he introduces into this tradition of thinking through replacing a subject-object model of cognition and action by an intersubjective one. However, he discerns certain biases in Habermas critical social theory, which he traces
308
Maeve Cooke
back to its reliance on formal pragmatics. The key point is that by reducing practical reason to a conception of rationality that is extracted from an analysis of
everyday linguistic practices, Habermas adopts a critical perspective on society
that is blind to forms of human suffering, and corresponding forms of human flourishing, that cannot be conceptualized as failures of linguistic communication. His
critical perspective has no room, for example, for critique of violations of bodily
integrity, and the corresponding idea of human flourishing as involving bodily selfconfidence, or for critique of the disaffection and alienation that result from lack of
recognition of the value of ones work, and corresponding idea of human flourishing as involving self-esteem based on ones achievements. For Honneth, this leads
to a regrettable narrowing of the scope of social criticism, permitting only normative deficits of certain kinds to come into view. The objection, in sum, is that
Habermas theory fails to exploit the full potentials of an intersubjective framework, leading to a form of critical social theory that is too thin and too weak to provide a basis for critique of the principal pathologies of contemporary capitalism.
His alternative proposal is for an intersubjective model that does not prioritize the
linguistic aspects of human flourishing but takes account of its bodily-affective and
creatively-productive ones as well.
Habermas has responded to some but not all of these points of criticism. Moreover, in those places where he appears to have acknowledged the force of the criticisms, his responses have often been indirect. They manifest themselves in a gradual
moving away from the strong reliance on formal pragmatics that informed his theory
of communicative action towards more critical-hermeneutic and historicist lines of
argument (Cooke 2006: 71). To be sure, Habermas continues to afrm the importance of formal-pragmatic arguments in his critical project (Habermas 1996: 109). We
may take this as evidence of his unwavering commitment to a conception of reason
that, though historically situated, transcends times and places. Historicist and hermeneutic approaches, although productive in many ways, are unable to account for context-transcending validity. Were Habermas to dispense with formal pragmatics, or
even signicantly to reduce his projects reliance on it, he would have to nd some
other way of accounting for the context-transcending power of rational critique.
Notes
1. Important exceptions include sociologists who embrace systems theory as inspired by
Niklas Luhmann and structuralism as inspired by Ferdinand de Saussure. Both systems
theory and structuralism conceive of the deep structures that generate society as subjectless systems of rules. This also holds for poststructuralists such as Michel Foucault.
2. As we shall see, this position by no means implies the need for intentionalist accounts of
linguistic behaviour, which privilege the speakers intentions and neglect both the importance of what is said and the intersubjective relationship established in every com-
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
309
municative use of language. Nor does it imply that the speakers account of her own intention is incontrovertible.
Habermas views the tension as potentially positive insofar as it serves as a reminder that
both perspectives have their strengths and weaknesses. Although, as we shall see, he
makes a methodological decision in favour of the internal perspective, his theory of
communicative action acknowledges the importance of the external perspective, most
obviously when it offers a causal explanation of the pathologies of modern societies in
terms of the colonization of the communicatively regulated domains of social life by the
functional rationality of the economic and administrative systems.
These questions are further expounded in his longer study, The Logic of the Social
Sciences (Habermas 1988), an earlier version of which appeared as an article in the Philosophische Rundschau in 1967.
I have summarized Habermas discussion here. In the essay in question, Habermas distinguishes between three kinds of generative theory of society because he includes theories that do not take subjective meaning as a sociological primitive. Thus, in addition to
the two I mention, he identifies system theories that lack a subject (Habermas 1971:
18). See note 4 above.
Verstndigung (n.): reaching understanding, mutual understanding or communication. The corresponding verb is sich verstndigen. The term is ambiguous. It embraces linguistic comprehension (Verstehen) but goes beyond this to refer to the process
of reaching understanding, in the sense of reaching an agreement with another person or
persons.
When Habermas refers to formal semantics he is thinking of the philosophical theories of meaning that take their inspiration from Gottlob Frege and Betrand Russell and
in the second half of the twentieth century were associated with thinkers such as Donald
Davidson and Michael Dummett; Habermas mentions Peter Strawson, but his place in
this tradition is not straightforwardly evident (Habermas 1998: 99, note 62).
Habermas sometimes describes such claims as universal in the sense of transcending
times and spaces (e.g. Habermas 1994: 42). However, when he refers to universal validity claims on other occasions, he seems to mean universal in the sense of present in
every speech act (e.g. Habermas 1998: 89).
On Bhler (and Jakobson) cf. Daalder and Musolff this volume.
Habermas also asserts the primacy of the communicative use of language over what he
calls the indirect use of language such as the figurative and the symbolic (Habermas
1998: 166). Since his argument to this effect is cursory, I will not discuss it in the following. His neglect of the figurative and the symbolic uses of language, and demotion of
them to a subordinate status, is symptomatic of his overly sharp distinction between reason and imagination (see Cooke 1994: 7484 and Cooke 2009).
This is true irrespective of whether the intention to deceive or manipulate is conscious
or unconscious. Habermas distinguishes between conscious and unconscious latently
strategic action. Unconscious latently strategic action is correlated with what he calls
systematically distorted communication (Habermas 1998: 169). I cannot adequately
discuss the latter topic in the present context; I return to it briefly in the next section.
See note 13 below.
For twenty years subsequent to the publication of The Theory of Communicative Action,
Habermas had little to say about the problem of cultural impoverishment; in his most recent writings, however, he directly addresses the related question of semantic renewal,
310
Maeve Cooke
drawing attention to the potential semantic productivity of the beliefs and traditions of
the major world religions (Habermas 2008).
14. Habermas sees Immanuel Kant as the first philosopher to take seriously the differentiation of reason into its moments (Habermas 1990: 120, esp. 23).
15. I leave open the question of whether Hegel thought this process of self-actualization
would ever be concluded.
References
Adorno, Theodor W.
1981
Cultural Criticism and Society. In: Theodor W. Adorno Prisms. Trans. by
Samuel and Shierry Weber. Cambridge, MA: MIT Press.
Apel, Karl-Otto
1998
Openly Strategic Uses of Language: A Transcendental-Pragmatic Perspective.
In: Peter Dews (ed.), Habermas: A Critical Reader, 272290. Oxford: Blackwell.
Apel, Karl-Otto
1980
Towards a Transformation of Philosophy. Trans. Glin Adey and David Frisby.
London: Routledge and Kegan Paul.
Austin, John L.
1962
How to Do Things with Words. Oxford: Oxford University Press.
Benhabib, Seyla
1992
Bhler, Karl
1934
Sprachtheorie. Jena: Fischer.
Chomsky, Noam
1969
Aspects of the Theory of Syntax. Cambridge, MA: MIT Press.
Comte, August
1893
The Positive Philosophy of Auguste Comte. Trans. H. Martineau, 3rd edition.
London: Kegan Paul, Trench, Trbner and co.
Benhabib, Seyla
1992
Cooke, Maeve
1994
Language and Reason. A Study of Habermass Pragmatics. Cambridge, MA:
MIT Press.
Cooke, Maeve
2006
Re-Presenting the Good Society. Cambridge, MA: MIT Press.
Cooke, Maeve
2009
Zur Rationalitt der Gesellschaftskritik. In: Rahel Jaeggi and Tilo Wesche (eds.),
Was ist Kritik?, 117133. Frankfurt/Main: Suhrkamp.
Dilthey, Wilhelm
1996
Hermeneutics and the Study of History. Edited by Rudolf Makkreel and Frithjof Rodi. Princeton, NJ: Princeton University Press.
Fraser, Nancy and Axel Honneth
2003
Redistribution or Recognition? A Political-Philosophical Exchange. London:
Verso.
311
Habermas, Jrgen
1971
Reflections on the Linguistic Foundation of Sociology: The Christian Gauss
Lectures (Princeton University, February-March 1971). In: Habermas 2001,
1104.
Habermas, Jrgen
1974
Reflections on Communicative Pathology. In: Habermas 2001, 129170.
Habermas, Jrgen
1984
Habermas, Jrgen
1986
A Reply. In: Axel Honneth and Hans Joas (eds.), 214264.
Habermas, Jrgen
1987
Habermas, Jrgen
1988
On the Logic of the Social Sciences. Trans. Shierry Weber Nicholsen and Jerry
A. Stark. Cambridge, MA: MIT Press.
Habermas, Jrgen
1989
The Structural Transformation of the Public Sphere. Trans. Thomas Burger
with the assistance of Frederick Lawrence. Cambridge, MA: MIT Press.
Habermas, Jrgen
1990
Moral Consciousness and Communicative Action. Trans. Christian Lenhardt
and Shierry Weber Nicholsen. Cambridge, MA: MIT Press.
Habermas, Jrgen
1991
The New Conservatism: Cultural Criticism and the Historians Debate. Trans.
Shierry Weber Nicholsen, Cambridge, MA: MIT Press.
Habermas, Jrgen
1992
Postmetaphysical Thinking. Philosophical Essays. Trans. William Mark Hohengarten, Cambridge, MA: MIT Press.
Habermas, Jrgen
1996
Between Facts and Norms. Trans. William Rehg, MIT Press, Cambridge, MA:
MIT Press.
Habermas, Jrgen
1998
On the Pragmatics of Communication. Trans. and edited by Maeve Cooke.
Habermas, Jrgen
2001
On the Pragmatics of Social Interaction. Preliminary Studies in the Theory of
Communicative Action. Edited by Barbara Fultner. Cambridge, MA: MIT
Press.
Habermas, Jrgen
2008
Between Naturalism and Religion. Trans. Ciaran Cronin. Cambridge, MA:
MIT Press.
Honneth, Axel
1994
The Social Dynamics of Disrespect: On the Location of Critical Theory Today.
Constellations 1(2): 255269.
Honneth, Axel
2003
The Point of Recognition. A Rejoinder to the Rejoinder. In: Nancy Fraser and
Axel Honneth, 237247.
312
Maeve Cooke
Honneth, Axel and Joas, Hans (eds.)

1986
Communicative Action. Cambridge: Polity.
McCarthy, Thomas
1978
The Critical Theory of Jrgen Habermas. Cambridge: MA: MIT Press.
Weber, Max
1969
Basic Concepts in Sociology. Trans. H. P. Secher. Westport, Conn: Greenwood
Press.
2001
Philosophical Investigations. Trans. Gertrude E. M. Anscombe. Oxford:
Blackwell.
Part III
Key topics in pragmatic description
11.

William F. Hanks
1.
Introduction
The study of deixis is central to pragmatics, because deictic systems define points
of intersection between linguistic structure and the social settings in which speech
takes place. It is standard in contemporary English language sources to distinguish
Spatial (here, there), Temporal (now, then, tense), Person (pronominals), Discourse
(coreference or reference to prior talk) and Social Deixis (honorification and any
indicators of social identity or status relations among participants and contexts).1
Any one of these functional foci may be formally marked in utterances morphologically, lexically, or by construction type. Deictic phenomena inevitably interact
with gesture (conventionalized or not) and with many other parts of the grammar,
including evidentiality, status, modality, aspect, noun classification, possession,
spatial and temporal description. This chapter focuses on those forms whose primary function is to individuate objects of reference (including events, material
things, talk itself or any individuated concept). Referential deixis is found in all
human languages and includes at least demonstratives, person markers, locative,
directional and temporal markers, but excludes much social deixis such as honorification (where social status is indexed but usually not singled out for comment)
and standard sociolinguistic markers (where social factors are indexed but not
singled out for comment). By starting from the distinctive structure of referential
deixis, one can better characterize its relation to grammar and to other varieties of
indexicality. The main grammatical question is how deictic functions are encoded
in languages. The challenge for pragmatics is to determine the constant capacity of
deictic types to contribute to token acts of situated referring, and the relation between such referring and other aspects of utterance context.
Although they appear in various grammatical categories, deictics are all part of
a single field, which I will call the deictic field (Bhler 1934). The deictic field is
a psycho-social construct composed of: (1) the roles of communicative agents
relative to the participant frameworks they occupy, that is, the relations between
Speaker (Spr), Addressee (Adr), and Others, as defined by the language and communicative practices of its speakers; (2) the positions occupied by objects of reference, that is, the kind of object and its role in the activity in which the deictic
utterance is produced; and (3) the multiple dimensions whereby participants have
cognitive access to objects, such as perceptibility, relative proximity, memory, anticipation and prior talk. When a Spr performs an act of deictic reference, as in
thats John right over there (punctual point), we came from over there (vague
316
William F. Hanks
point), or heres your coffee (extending in hand), she produces a social relation
between herself, her interlocutor and the object. The deictic field is a template
for the elementary social relations on which conversation rests. The remainder of
the paper is organized into two parts, theoretical approaches and typological approaches.
2.
Theoretical approaches to deixis
Deictics are what Sacks (1992) called indicator terms and what linguists and
philosophers have called variously indexicals,2 shifters3 or, as in the present
paper, deictics (Fillmore 1997, Hanks 1983, 1990). In his theory of semiotics (cf.
Nth this volume), CS Peirce (1955) distinguished indexes from icons and symbols, according to the relation between the perceptible sign vehicle (or representamen) and its object (that for which it stands). Icons relate to their objects by way
of similarity, symbols by way of sheer convention, and indexicals by way of dynamical contiguity. As Peirce (1955:107) put it, the index stands for dynamical
(including spatial) connection [of the deictic form] both with the individual object,
on the one hand, and with the senses or memory of the person for whom it serves as
sign, on the other. Certain indexes are in a causal relation to their object (smoke
indexes fire, a weather vane indexes wind direction) and others are conventionalized (this indexes objects close at hand, there places at a remove from the location of utterance, road signs index road conditions). In the prototypical cases, the
index does more than stand for its object, it directs an addressees attention to it,
and this directive force is especially salient in certain types of deixis (e.g. presentatives, such as Here it is (take it!), There he goes! (look!)). Such expressions do
more than individuate an object; they direct the addressee to it, and the proper uptake is usually a gesture or a shift in attention focus.
Morris (1946) simplified the Peircean framework, distinguishing characterizing signs (icons and symbols, both of which characterize their object) from indices
(which are devoid of descriptive information about their object) and universal
signs (anyone, anywhere, etc., which are applicable to any item in a class). From
this perspective, what is noteworthy about deictics is that they contribute to individuated referring without in any way describing their objects. As Sacks (1992:
520) put it, such terms, which he called indicators, are altogether abstract.
[] capable of invoking the sheer fact of the setting, without in any way formulating it.
The definition of deixis in terms of individuation without meaning is common
in philosophical literature (Searle 1969:80, Nunberg 1993, Bar Hillel 1954), but
has less prominence in the linguistic literature, where cross-language comparison
and detailed description reveals many conventionalized features in deictic systems
(see below). While Morriss framework played a pivotal role in the development of
317
linguistics (cf. Nth this volume), his definition of indexicality is too oversimplified to describe natural language deixis. By contrast, Peirce insisted that his three
sign types designate relations which typically combine in actual signs. The conventionality of any language guarantees that the pairing of form with meaning in
the language is conventional, no less so for deictics than for other kinds of expression. Moreover, this conventionality encompasses deictic features such as
punctate vs. regional space, centripetal vs. centrifugal movement relative to utterance context, kinds and degrees of perceptual access to the objects denoted by demonstratives, the presence or absence of gender, animacy and noun class in demonstratives, and any special presentative forms in a language (see French voici,
voil, Yucatec Maya hebe there it is (listen!).The result is that actual deictic
forms and arrays always combine sheer indexical functions with other kinds of
information (indexical, iconic or symbolic). If one maintains the more austere
philosophical definition of deixis as pure indexicality, we are led to conclude that
natural language systems are always impure blends between indexical and non-indexical functions. A more useful and realistic alternative is to conclude that deixis
in natural language requires a more robust Peircean definition of the phenomenon,
and challenges researchers to explain why certain bundles of functions cluster
around indexicality in language after language. This point is all the more clear
when we consider that deictics are typically organized in closed class paradigms,
where obviously indexical functions are bundled with a wide array of other information.
Natural language deictics are what Bloomfield (1933) called category representatives, that is, they bear features for morphosyntactic category and function as
syntactic heads. Some but not all deictics may be used either exophorically, for reference to extra-linguistic context, or endophorically, for anaphoric co-reference
with prior talk (A: I prefer this one, B: Oh, I like that too) or cataphoric co-reference with subsequent talk (the two statements have this in common, they are
terse) (see Halliday and Hasan 1976). While there are borderline cases where this
distinction is difficult to draw, there are also regular patterns in which deictic types
are specialized for one or the other. Anaphora, for instance, is often associated with
forms which when used exophorically indicate remoteness of the object, while spatial adverbial paradigms often include one or more forms that can be used only exophorically. They also differ as to whether they may be lexically expanded, as in
here vs. here in my room, this vs. this blue pencil, and so forth. It is common
to describe the simple forms as pro-locative, pronominal, and so forth, and the expanded forms as adlocative, adnominal and so forth. Moreover, deictics in many
languages may be combined in distinctive n-tuples of coreferential or mutually reinforcing forms (this here book, Yucatec leti e hela < the one the here.it.is =
this here one, te waya < right.there here = right here [point within a region]).
It is also common to find distinctions in spatial deixis according to the regional
scope of the place referred to, as in Yucatec tela right here (punctual) vs. waye
318
William F. Hanks
here (regional). By contrast, English here may denote a point on the speakers
nose, the room, house, neighborhood or country in which it is uttered, depending on
context, and similar extensions apply to there, now, then, and others. In short,
when one examines the pragmatics of actual deictic systems, many features arise
that are elided in the austere definition of indexicality as semantic emptiness.
In his classic work Sprachtheorie, Bhler (1934) initiated what would become
another line of thought concerning deixis in language. Bhler distinguished three
dimensions in the elementary speech event, the Speaker, the Addressee and the Object talked about. Intersecting all three poles is the division between what he called
the Zeigfeld, the demonstrative field in which objects can be shown, and the Symbolfeld, the discourse field to which description of the object, and other discourse
belongs. This relatively simple schema already generates six values (three in each
field) and provides the elementary framework for a functional analysis of deixis:
deixis and demonstration belong to the field of exophoric reference to objects
available in the immediate perceptual field. Anaphora, reference to prior talk and
to remembered objects are treated as blends of the Zeigfeld and Symbolfeld.
Jakobson (1960, 1971) expanded and refashioned Bhlers distinctions into the
division between the speech event (Es) in which talk occurs and the narrated event
(En) talked about. Jakobson eschews the term deixis, but treats all referential indexicals under the rubric of shifters (a term coined by Jespersen 1965: 290).
Shifters are linguistic categories whose distinctive feature is that they individuate a
referent (in the En) in relation to the situation of speech (Es). Thus if Ben tells Katherine I saw him right here yesterday, every item but the verb see is a shifter
because each one establishes a relation between the utterance and the event narrated: I equates the current speaker with the seer, the past tense of the verb places
the En in a time prior to the moment of utterance (reinforced by yesterday, the
day prior to the day of utterance) and him indicates that the person seen is neither
the current Spr nor Adr. Jespersen adopted the label shifter in light of the distinctive behavior of these grammatical categories in reported speech. If Katherine subsequently reports Bens utterance to Madeleine, she says Ben told me he saw him
there yesterday. In reported speech the shifters shift to accommodate the new
speech event frame of the quoted speech, while maintaining sameness of reference
to the event of seeing and its participants. Inversely, in direct quotation, the linguistic forms remain constant, but the reference is dislocated. If Katherine directly
quotes Ben, she says Ben said I saw him here yesterday. Notice that despite the
I, here in this utterance, Katherine does not therein refer to herself or to the location in which she reports the speech. In effect, the introductory phrase Ben said
projects a prior Es in which the utterance was produced, and it is that prior event
that serves as the ground of reference for the deictics. Hence there is no shift in the
forms, but the speech event relative to which they are interpeted has been transposed. Silverstein (1976) follows Jakobsons shifters model and expands it to a
wide array of indexical effects, only some of which contribute to reference. Essen-
319
tially the same set of distinctions underlies the still more elaborate analysis of the
speech event in the ethnography of speaking.4
Phenomenologists were centrally concerned with indexicality (Husserl 1978,
Schutz 1970, Merleau-Ponty 1967) as evidence of the unavoidable situatedness
of speech in the immediate experiential settings of its production. While phenomenology has little impact in English language linguistics, it is fundamental to ethnomethodology (Garfinkel 1967, 1972) in its claim that all meaning is (indexically) produced in the ad hoc processes of interaction, and similarly to the
approaches to interaction pioneered by Cicourel (1964, 1974, 1985, 1992), Schegloff (1972, 1982), Gumperz (1982) and Goodwin (1981, 2003). From a psycholinguistic perspective, Rommetveit (1974) combines phenomenology with linguistics
to argue that linguistic production relies not on abstract semantics, but on the
experiential worlds of speakers. Conversation analysis as developed by Sacks,
Schegloff and others carries forth the basic premise that speech is pervasively indexical and grounded in the interaction between co-participants. Developing the
structural linguistics of Meillet (1926), Martinet (1960) and the Prague School,
Benveniste (1965, 1974) incorporated phenomenology in his landmark papers on
subjectivity in language.
This range of approaches to indexicality raises a number of issues. There is
widespread agreement that indexicality designates the context dependency between utterances and speech contexts, and that deictic types are specialized in individuating single objects or groups of objects in relation to the participants in the
event of speaking. We can schematize this as in Figure 1, where the Indexical
ground is the zero-point, or pivot relative to which the deictic points to the object
denoted. Thus for example, in an utterance like its sunny here, the locative here
projects a relation of relative immediacy to or inclusion of the Speakers location at
the moment of utterance.
Relation
Indexical ground
Object
Figure 1. Relational structure of referential deixis
This three-way distinction raises questions on which there is considerable debate

in the literature. First, what kinds of relational features are encoded in deictic
types? The spatialist view holds that relative contiguity (this = proximal, that =
non-proximal and so forth) is fundamental, and other pragmatic effects are incidental to deixis or derived by metaphorical extension (Anderson and Keenan1985)
or contextual inference (Levinson 1983). For careful assessment of these alternatives, argued with data from Lao, see Enfield (2003, 2009). By combining a spatialist definition with Gricean implicature (Grice 1967, 1978) and conversational
organization, the enriched values of deictic tokens can be calculated as implica-
320
William F. Hanks
tures. By contrast, recent non-spatialist views reject the a priori primacy of space,
and substitute for it something closer to accessibility, that is, how the participants
have or gain access to the object.5 Spatial contiguity sometimes provides a support
for accessibility, but is only one possible parameter alongside others, including
perception (via any of the senses), prior talk and memory, salience (cognitive or
perceptual), evaluative stance, relative familiarity with the object, and in some
cases ownership of the object by a participant. Deictics may be relatively lean in
terms of descriptive information, but the pragmatics of deixis nonetheless responds
systematically to the features of actual objects. Where the spatialist approach presumes that space is basic across all contexts, the accessibility approach looks to
situational relevance, the interactive move made by the deictic utterance, the social
definition of the object and other features of context (Hanks 2005a).
The second major question bears on how one defines the indexical ground.
Traditional treatments assume that the zero-point of deixis is the speaker, as in
here means a place close to me. One of the earliest statements of what can be
called the egocentric approach is Russells description of egocentric particulars
(1940), which cover the range of deictics (compare Evans 1982: Chapter 6, Gale
1968). Most traditional grammars and typologies (eg. Anderson and Keenan 1985)
assume egocentricity, and Benvenistes (1966, 1974) anchoring of indexical reference in subjectivity makes the same move (cf. Lyons 1982: 121). With the exception of Schutz, whose emphasis was intersubjectivity, most phenomenology also
proceeds from the ego-subject (albeit with different inflections). Egocentricity is
consistent with the commonsense idea that the solitary speaker is the one who produces her utterance, and the minimal requirement for utterance production is a
speaker. The challenge for this position is to explain the many cases in which a
deictic responds not to speaker accessibility, but to the addressees access to the
object, or to the relative symmetry of access between the co-participants. Moreover, while the idealization of the solitary speaker has been productive for certain
purposes, Conversation Analysis played a leading role in showing it to be patently
inadequate as a basis for pragmatics.6
There are various alternatives to egocentricity. Reichenbach (1947) described
indexicals under the heading of token-reflexivity, by-passing the problem of the
Speaker in favor of the utterance itself as zero-point. The utterance here denotes
a place immediate to the utterance token here, just as I denotes the individual
who produces the token I. This removes the problem of the zero-point from the
realm of interaction, stating in effect that these forms denote in relation to instances of themselves. Close description of actual deictic usage indicates that elements such as common ground,7 collaboration between participants (Goodwin
1981, 2003), and the relative symmetry of participant access (whether speaker and
addressee have roughly the same or significantly different access to the object in
situ; Hanks 1990, 1992) all impact on the use and understanding of deictic tokens.
Any of these factors may be conventionalized in the semantics of deictic types.
321
This suggests that the zero-point of indexicality is best conceived as the participation framework in which talk is produced. Just as space may be a key kind of access for some uses, the speaking ego may be the ground for some uses. But neither
space nor egocentricity is a priori privileged.
These different approaches to deixis imply different background pictures of utterance context. The egocentric spatial picture holds that deictic acts occur when
the Spr, the Adr and the Object are physically copresent and mutually perceptible.
The zero-point of the deictic utterance is the Sprs body and the utterance orients
the Adrs attention to a specific sector of the broader sphere of spatial proximity,
not unlike a pointing gesture (which may indeed be part of the utterance). To
understand a deictic is not to interpret it but simply to recognize by observation
what it singles out in the physical situation of utterance. The space of dialogue
may be interperspectival, but it is the Spr who produces the utterance and his or
her body that serves as the indexical anchor-point: a self-contained individual,
oriented in physical space and endowed with a sphere of contiguity and sensory access. This picture is dominant in much of the English language literature on deixis.
It is found in most modern linguistic descriptions, especially typological and psycholinguistic work, for which it seems to offer a controlled basis of cross-linguistic
comparison.8 Despite its appeal, however, this picture omits critical features of
deictic practice, including the mutual orientation of interactants, all non-perceptual
modes of access such as background knowledge, memory or anticipation, and all
extra-physical aspects of social settings. Non-spatial aspects of deictic speech are
either ignored, treated as metaphorical extensions from space, or derived from
other, non-deictic principles.
There is a second, widely attested background picture for deixis, according to
which the core of context is social interaction.9 In this view, deictic utterances involve at least the following factors.
(i) The reciprocal perceptual fields of the parties, especially reciprocal vision in the
face-to-face (Schutz 1970, Goodwin 1981, 1994, 2000, Kendon 1992).
(ii) The broader reciprocity of perspectives whereby each party assumes that the
other has a perspective and that if (s)he adopted that perspective, the world would look
to her as it does to the other (Schutz 1970). This is precisely reflected in the alternation
of I, you, here, there among Spr and Adr: each party to talk is an I, but only to himself, and each one is a you, but only to the other, my here is your there and so forth.
(iii) Participants bodies are expressive through gestures, as well as receptive through
the senses, and gesture is a key aspect of deixis.10
(iv) Any interaction has sequential organization, perhaps the best studied feature of
conversation. Sequential organization is embodied in turn taking, adjacency, contiguity
and proximateness.11
(v) The interaction has a motivational structure: the Spr speaks in response to something and in order to achieve something else; the Adr responds in the light of the aim of
the first utterance, in order to achieve his or her own aim, which then serves as the because motive of the next turn, and so on back and forth (Schutz 1970).
322
William F. Hanks
(vi) Any utterance has a relevancy structure according to which certain things matter
and others dont.12
(vii) Utterance meaning must be negotiated or worked out by the co-engaged parties.
It is not given in advance, nor is it fixed by the intentions of the Spr (Clark, Schreuder
and Buttrick 1983). For deixis, this means that, for instance, the reading of a deictic
token as spatial proximity, perceptibility, cognitive salience, egocentric or altercentric
can vary by context and must be worked out by the participants.
(viii) Participants display to one another their sense of the current situation, the relevancy structure they are assuming, and their current relation to their own speech. Deixis is a primary resource for such display precisely because it points into the situation
and thereby positions the Spr.
In the Interactionist picture, there is an emergent space of interaction, but it is not

the kind of egocentric, physical space presumed by the Spatialist view (Kendon
1992), and to study it, we must attend to sequential organization of talk, to situational variation and to the micro-ethnography of everyday usage. In the course of
deictic practice, interactants must jointly establish and display the relevance of
spatial perceptual, discursive or other contextual frames. Moreover, in much ordinary referring, sequential placement of a deictic utterance in talk combines with
spatial location of the object in the course of identifying the denotatum. Ordinary
practice therefore blends endophoric with exophoric functions.
For the purpose of deictic analysis, situational relevance has two main sources:
(1) local relevance turns on the immediate circumstances of utterance, including
sequential placement of the utterance, the location of the object and the positions
of the interactants in the interactive field, and (2) global relevance turns on the embedding of the deictic field in a larger social field. The broader social setting can
define the boundaries of here for the purposes of talk, and factors such as social
identity or hierarchy and activity spaces may license extensions, transpositions, or
they may constrain usage. For instance, if a foreman on a work site says I am
installing the electric outlet here, the utterance can be fully normative even if
the foreman does none of the actual work. The hierarchical role of the foreman
licenses first person singular usage for tasks of which he is the leader but not literally the one who does the work. Similarly, spatial references in the extended
households of Yucatec speakers respond systematically to the social organization
of space, so that here may refer to the entire household or exclusively to the portion occupied by the Spr. Shamans and doctors who use imaging technologies
(divining crystals, x-ray images and the like) routinely make deferred ostensions
in which pointing to the image is pragmatically equivalent to pointing to its bodily
counterpart. In cases like these, the activity space in which the deictic practice
takes place provides specific affordances and relevancies that help motivate practice.
The Spatialist and Interactionist picures of utterance context can be combined
so long as we treat them as complementary: spatialism is a claim about the re-
323
lational features of deixis, and interactionism is a claim about the indexical ground.
We can grant interaction as the ordinary context of utterances, while still claiming
that the semantics of deixis is egocentric and spatial. In this case, something akin to
Gricean implicature is invoked to derive contextual effects from what is literally
spatial reference. At the same time, conventionalized non-spatial features, such
as perceptual access, are taken to be non-deictic. Recent work by a variety of
scholars has combined grammar and interaction, thereby opening the possibility of
a blended account of deixis.13
While some combination of linguistic and interactional analysis is undoubtedly
necessary to describe the pragmatics of deixis, it is not without cost. If the interactionist grants that verbal deixis is fundamentally spatial in its semantics, then interaction becomes an epiphenomenon of grammar; it is merely how people ordinarily use linguistic forms whose meanings are fixed apart from context. This is
surely contradictory to the work of ethnomethodologists such as Garfinkel (1967),
who took the strong position that all meaning is produced in the course of ordinary
talk, and it also contradicts the more recent position that talk is inherently situated.
Similarly, the spatialist position is reduced to an idealization if one grants that
neither egocentricity nor spatial relations are necessarily salient in ordinary interaction. Without recourse to the counter-evidence of actual usage, and protected by
the vagueness of Gricean implicature, it is hard to see what evidence could be adduced either for or against spatialism.
What the two background pictures have in common is that they take the conditioning circumstances of pragmatics to be local to the utterance: in the spatial view,
it is the field of proximity to the speaker, and for the interactionist, it is the immediate sequential context, augmented by those features of context to which participants are oriented (Schegloff 1987). Linguistic anthropologists have questioned
the localism of both approaches, in an effort to understand the how broader social
values and definition of the occasions of talk systematically affect deictic practice
(Hanks 1990, 2005a, Enfield 2003, Goodwin 2006, Silverstein 1976). Cross-cultural descriptions of talk suggest that there is no value-neutral way to define the
dynamical contiguity that is at the core of deixis. Starting from the linguistic
structure of deictic systems, to the local circumstances of deictic acts, this third approach proceeds to the social fields in which the deictic field is embedded (Hanks
2005a, 2006a). This includes the identities of participants, the genres of practice of
which deixis is a part, the social definition of place and time and the values attaching to objects of reference. The resulting framework is less elegant than formal accounts, but more empirically precise to the varieties of deictic practice that we find
in actual talk.
324
3.
William F. Hanks
Typological approaches
Deictic systems can be compared along a number of axes according to their grammatical and functional dimensions. The first point of comparison is an inventory of
the grammatical categories for which the forms are marked: Noun (pronouns, articles, demonstratives, e.g. I, the, this, that), Adverbs of space (here, there), time
(now, then), manner (like this, like that), Presentative forms (French voici, voil,
Maya hela here it is (giving), helo there it is (pointing), hebe there it is
(sniffing, cocking ear), Verbs (especially motion verbs like come, go, see Fillmore 1997), and various elements that attach to major categories (directionals, as
in Mayan verb stems; clitics and utterance or phrase initial or final particles).
Prototypical deictics are category representatives that can substitute for, or be expanded by lexical descriptions (this vs. this red pen). In the former case, the
deictics is pro-nominal, and in the latter it is ad-nominal. In some languages the
pronominal forms and the adnominal qualifying usages correspond to different
series, as in Mulau, a Daic language, and Japanese (Diessel 1999: 59). Diessel
(1999: 58) reports that in twenty-four of the eighty-five languages he sampled, pronominal and adnominal usages are marked by different categories, and in languages such as Turkish and Lezgian, in which the same deictic roots are used in
both functions, the pronominal forms inflect for case, whereas the adnominal ones
do not.14
The grammatical classification of deixis raises various questions, of which two
deserve mention here. The first is whether all or part of certain category-specific
subsystems are derived from forms in other categories. For instance, adverbial
deictics in some languages are derived from nominal ones, as in Quechua (kay
this > kay-pi here, chay that [medial] > chay-pi there, haqay that [dist] >
haqay-pi there [dist] (Calvo Perez 1999, Lasater 2009). Temporal deictics are
often derived from spatial ones, as in Yucatec tol-akhak-o back then [remote
past] < tol-o out there [remote], or from non-deictic lexemes (cf. today, yesterday, aujourdhui, Maya behelae today (lit. road.here.it.is).
The second question is the degree to which the different subsystems in any language are similar or different in structure. Do we find the same series of oppositions in different subsystems, as in French ceci, cela (nominal), ici, l (spatial)
voici, voil (presentative) or English this, that, here, there, now, then? The apparent proportionality of deictics across catgories has encouraged many scholars to
propose paraphrase relations among deictics, such as this is the one here, that
is the one there, here is the place occupied by I, and so forth.15 Both the derivational relations between subsystems and the relative proportionality of oppositions go to the larger question of the degree to which deixis in any language is a
single system unified by shared forms, regular formations or recurrent arrays of
distinctions. Comparative research has demonstrated that languages differ on this
axis.
325
A closely related axis is the degree of compositionality of deictic forms. Languages like English and Hausa have monomorphemic deictics.
nn
nan
cn
can
Prox to Spr (immediate field)

Distal to SPr/Prox to Adr
Prox to Spr
(non-immediate field)
Distal to Spr (Hill 1982, Jaggar and Buba 1994)
By contrast languages such as Yucatec show bimorphemic stems in all categories,

with a number of more complex stems among the temporal adverbs (Hanks 1990):
lel-a
tel-a hel-a
bey-a
lel-o
tel-o hel-o
bey-o
wal-ak-hak-a at this time of day in the past
tol-ak-hak-o at that time in the remote past
In Quechua, nominal demonstratives mark a three way contrast between Prox,

Medial, Dist with monomorphemic forms, and the spatial adverbs are derived from
the demonstratives by suffixation of [-pi], as in:
kay
chay
haqay
this
that
that
kay-pi here
chay-pi there
haqay-pi
over there
Similarly, Mandarin demonstratives are monomorphemic, whereas spatial adverbs

are derived by suffixation (Tao 1999, Yang 2006).
zh
z
this, hi accessibility
that, low accessibility
zhr, zhli
nr, nli
here
there
At the high end of compositionality are languages like Nungubuyu, Santali and
Malagasy, where deictic words may be composed of five or more morphemes, each
of which contributes to the meaning of the whole. For Nungubuyu, Heath (1980)
reports that four basic roots distinguish Proximal, Accessible (to Adr), Distal and
Anaphoric in both demonstrative and adverbial categories, with the adverbs
derived by suffixation. These are cross-cut by + Concrete (precise vs. vague),
+ Kinetic (trajectory vs. location), + Centric (where +Centric is subdivided
into Centripetal Centrifugal, and Centric is used for transverse relative to the
Sprs field of vision). The nominal forms are further subdivided by + Absolute
(which Heath glosses definiteness), gender (masc, fem), number (sg, dual, pl) for
human referents, and an additional five noun classes for non-humans. Each of these
distinctions is marked by morphological exponents, resulting in relatively compositional forms.
In Santali all deictics are composed of minimally three morphemes, but potentially more (Zide 1972). There are three bases (proximal, medial, distal) that combine with three regions (immediate, medial, remote) to yield nine different spatial
values. The resulting forms may be further marked as Lateral (for objects at the
edge of the perceptual field, like Nungubuyu non-Centric), Intensive (for referen-
326
William F. Hanks
tial precision, like Nungubuyu Concrete), Visible Invisible (two distinct morphemes), Animate Inanimate, and a three way number distinction. Denny (1982:
371372) reports that in Inuktitut, there are twelve deictic roots, distinguishing
Out of field In-field and within In-field, Away (non-prox) At (prox), and
within Away, Bounded (Exterior Interior) Vertical (InferiorSuperior) Horizontal. The In-field categories are further divided by a binary division between
Restricted (equidimensional, precise) Extended (elongate, mobile, regional).
Further modifications (by inflection or derivation) include grammatical category
(nominal, adverbial, predicative), transposition of the indexical ground from Spr to
Adr or Other, case, number, existence in a place and motion yield over six
hundred and eighty-five deictic words.
In general, the greater the degree of compositionality, the greater the functional
transparency of the composed forms and the more likely it is that some deictic morphemes (root or affix) will recur across subsystems, resulting in a greater degree
of proportionality (although non-compositional languages, like English, can be
highly proportional). Grammatical categories, derivational relations, proportionality of oppositions and compositionality jointly make up the paradigmatic skeleton of deixis in any language and provide a basis for more detailed comparison
of functional structures in different subsystems and across languages. Are there recurrent gaps in the paradigms, markedness asymmetries, or domains in which
functions are hyperdifferentiated? Individual subsystems in any language and
across languages differ widely in the relative delicacy of the distinctions they
mark.
Many linguists have sought to organize this variability in typologies based on
the number of opposed terms given deictic systems display. Frei (1944) first formulated this approach, noting that the most commonly attested systems are two- or
three-term systems, often augmented by an additional term that is neutral as to distance. Anderson and Keenan (1985) build on Freis schema, as does Fillmore
(1997) and more recent typological work done by scholars at the Max-Planck Institute for Psycholinguistics. Among languages with four or more terms, Anderson
and Keenan (1985) cite Malagasy, Quileute and Sre, to which we can add Yucatec
Maya (Hanks 1990), Tlingit (Boas 1917, Swanton 1911), Crow (Graczyk 1986)
and Santali (Zide 1972).
There are several challenges facing this kind of typology. First, how to distinguish between multi-term systems in which each term indicates a degree of spatial
remove vs. ones in which an erstwhile medial is actually neutral or marks a shift in
the indexical ground. The former are treated as distance-centric, the latter as person-centric (Anderson and Keenan 1985). A second challenge is to decide on
which forms belong in the inventory. In highly compositional systems or ones in
which multiple deictics are combined in relatively fixed phrases, the line between
basic deictics, on which the typology is focused, and non-basic ones can be difficult to draw (a difficulty pointed out by Frei 1944: 114, 120). A language that ap-
327
pears two-term from a restrictive definition of the paradigm may in fact systematically mark three or more distinctions at the level of conventionalized phrases.
Inversely, a system with three forms may resolve to two binary distinctions. Hence
the number of formal distinctions need not correspond to the number of functional
distinctions. Finally, the number of terms in one subsystem of a language need not
be identical to the number in another. In Maya for instance, the Manner deictics
show a binary split, plus a neutral term, whereas the Spatial deictics make a fiveway distinction and the nominal deictics a four-way split (Hanks 2005a).16
Diessel (1999: 51) presents a useful summary of what he treats as semantic,
syntactic and pragmatic features of deixis. Under deictic semantics he lists Distance (prox, medial, dist), Visibility (vs. Invisibility), Elevation (up, down relative
to origo), Geography (uphill, downhill, upriver, downriver) and Movement (toward, away from, across the visual field of the Spr). Under quality semantics he
cites Ontology (location, object/person), Animacy, Humanness, Sex, Number and
Boundedness (bound, unbound). The difference is that the deictic features apply to
the relation between the ground or origo and the denotatum, whereas qualitative
features characterize the denotatum itself. Grammatical categories, case marking
and agreement are grouped under Syntax, while Pragmatic features consist of Use
(exophoric, anaphoric, discourse deictic, recognitional) and Reference (emphasis,
contrast, precision). While this way of arraying the features is heuristically useful,
it leaves out features of the indexical ground (origo as speaker, addressee, common
ground or other) and draws an artifical distinction between semantics and pragmatics (since conventionalized indexicality is both semantic, because encoded,
and thoroughly pragmatic, because linked to usage). Furthermore, by splitting
deictic from use features, the typology obscures the fact that spatial location,
vision, endophoric and recognitional features denote alternative modes of access
(perceptual and cognitive) to the referent. For our purposes, it is more useful to
classify features in terms of the three-part relation between the indexical ground,
the relational features and the object denoted (see Figure 1).
4.
The relational values
The relational values of deictics are widely discussed in the literature. They include Relative Immediacy (in space or time), Interiority (inside, outside, lateral),
Location vs. Trajectory, Perception (visual or other), and several varieties of Cognitive Access (anaphora, reference to discourse, relative salience, perspective
[prospective, retrospective] and recognitional). Temporal deixis poses numerous
problems beyond the scope of this essay, including tense, aspect, mode and the
status of quasi-deictic shifters like today, tomorrow, yesterday, ago. Like pronoun
systems, temporal shifters are functionally, and sometimes formally related to
other kinds of deixis, but sufficiently different to merit separate treatment.
328
4.1.
William F. Hanks
Spatial immediacy
Frei (1944) noted that most deictic systems distinguish two or three degrees
of relative proximity between the ground and the object. Examples are English
(here there), Vietnamese (Close vs. Far; Diessel 1999: 36), West Greenlandic
(Prox, Medial, Dist; Fortescue 1984 cited by Diessel 1999: 46) and Turkish (Prox,
Med, Dist; see Diessel 1999: 60). Among three-term systems, Anderson and Keenan (1985: 282285) and Diessel (1999: 50) distinguish two kinds: (1) distanceoriented systems, in which the medial term marks mid-range from Spr and the Distal term marks remote from both Spr and Adr, and (2) person-oriented systems in
which the medial marks Prox or Identifiable to Adr. The relation between deictic
distinctions and participants was addressed by Brugmann (1904), Boas (1911),
Bhler (1934), Frei (1944), Kurylowicz (1972) and Fillmore (1982: 4748). In the
current framework, the so-called person oriented systems illustrate not distinct relational features, but distinct indexical grounds, a point to which we return below.
According to Anderson and Keenan (1985: 286289), numerous languages (including Malagasy, Santali, Sre, Quileute, Tlingit) distinguish four or more degrees
of distance from ground. Both Fillmore (1982: 4849) and Diessel (1999: 4042)
argue that no language distinguishes more than three degrees of distance, and such
expanded systems always involve some other dimensions (such as perception, Inside Outside or a shift in the origo). Debates over the status of the medial term
and hyper-differentiated distance values both illustrate the difficulty of isolating
purely spatial distinctions in the deictic field.
In addition to degrees of remove from the indexical origo, some languages distinguish values on a dimension that can be called interiority, that is, whether the
referent is inside, outside or lateral to the indexical origo. West Greenlandic for instance makes a three-way spatial distinction in the simple demonstratives, but also
has a special deictic root for reference to objects outside the utterance space, as
well as four separate roots that combine distance with verticality: medial.up, medial.down, distal.up, distal.down relative to origo (Fortescue 1984: 259262). Denny
(1982: 361) reports a five-way distinction for the distal locative there in Inuktitut:
up.there, down.there, in.there, out.there, over.there (horizontal). In the spatial
deictics in Yucatec Maya, there is no verticality, but regional forms (as opposed to
punctual ones) distinguish Inside (inclusive of origo) from Outside (exclusive of
origo)17.
In a number of languages, including Nungubuyu, Inuktitut, Santali and Kiowa,
there are specialized deictics for reference to objects in motion, for which the origo
serves as source space for centrifugual motion, the goal space for centripedal, or
the ground for lateral motion (transverse or along the edge). Heath (1980: 152) describes the Nungubuyu forms as Kinetic, and shows the three way distinction between [-ala-] approaching, [-ali-] going away and [-waj-] across. In Inuktitut,
deictics may combine with a motion infix [-qa-] plus case marking to distinguish
329
source, goal and path readings (Denny 1982: 373). For Crow, Graczyk (1986) cites
the postposition [-ss(ee)] which marks centripedal trajectory. Zides (1972:
267268) description of Santali deixis includes the infix [-Vh-], which marks that
the referent is off to the side (lateral) of the origo space. Diessel (1999: 46) cites a
similar distinction in Kiowa between centrifugal and centripedal deixis. In other
languages, such as English and many Mayan languages, directionality is encoded
in the verb complex, either lexically or by way of affixes (see Fillmore 1975 on
English come, go). Dixon (1972: 48) reports an uphill, downhill distinction in
Dyirbal deictics, and Matisoff (1973: 110111) sites a similar one in Lahu.
4.2.
Perceptual access to the object
In theory, perceptual access to the referent may play a role in any exophoric deixis,
and vision is a key source in spatial deixis. While noting the importance of visual
access in some deixis, Fillmore (1982: 51) adopted a restrictive definition in which
such perceptual conditioning is not deictic per se. By contrast, Bhler (1934) starts
from the premise that sound is critical, because to identify the referent of here one
need only follow the speaking voice to its source. In some languages, such features
are conventionalized: certain deictics or bound morphemes indicate the presence
or absence of perceptual access to the object. Despite the variation in perceptual
coding, it appears that in all languages, if perceptual access is encoded, it is access
for the Spr (and not the Adr or some Other) that is in play (Diessel 1999: 4142). In
other words, no language specifically marks that the Object is perceptually available to the Adr, even though there are many cases in which spatial contiguity is
computed relative to the Adr (see next section). Diessel cites seven languages in
his sample in which Visibility is overtly marked, and there are many more cases.
In some languages, both Visible and Invisible are coded in separate roots or
stems. For example, Quileute shows a root distinction between Visible and Invisible, where Visible is further subdivided into Prox (Spr), Prox (Adr), Medial (Spr +
Adr) and Distant (Andrade 1933). Chinook makes a basic distinction between
Visible (also used for Present) and Invisible (also used for Past) (Boas 1911:
617623). Kwakwala shows a similar distinction (Anderson and Keenan 1985:
290). Crow (Graczyk 1986) shows a slight variant in the four-way distinction between Immediate, Medial, Distal (+ Vis), Distal (-Vis). These oppositions between
presence and absence of visual access are all equipollent in that both terms are
overtly marked.
By contrast, many languages show privative oppositions, in which one term is
marked and the other unmarked. Santali appears to show a three-way split between
stems unmarked for perception, ones marked for lateral visual and ones explicitly marked for unseen objects (Garvey et.al. 1964: 115, Zide 1972: 267). West
Greenlandic has a special root [-im] for Invisible objects (Fortescue 1984: 259
262, Diessel 1999: 46), as do the Malagasy locative deictics (Anderson and Kee-
330
William F. Hanks
nan 1985: 293). Diessel (1999: 42) cites Ute to illustrate what he considers the
most common alignment: Prox Distal Invisible. It appears that in any language
with a privative opposition on the parameter of vision, it is the Invisible form that
gets special marking. No language has a special form marking Visible unless it also
has one for Invisible. The implication is that visual access is the normal state of affairs, and the use of simple proximal or distal forms in a language with an overt Invisible, pragmatically implies that the object is within sight. This is consistent with
the fact that vision plays a central role in locating objects in space.
The close association between visual and spatial identification is reflected in
the widespread tendency among linguists to overlook the role of other senses in
identifying objects of deictic reference. With the exception of Bhler (1934) and
Hanks (1983, 1990, 1993a, 2005a) none of the works cited in this chapter describes
auditory, olfactory, tactile or gustatory access of deictic referents. While such distinctions are evidently rarely encoded in deictics, they frequently part of the phenomenal conditions in which deictic acts occur. But in Yucaec Maya presentative
deictics, there is a three way division between Tactual (for objects within touch of
the Spr), Visual (objects visible to both participants) vs. Peripheral sensory (objects perceived by hearing, smell or taste) (Hanks 1993a, 2007b).18
4.3.
Cognitive access
Deictics in any language may individuate objects that are cognitively accessible,
but neither perceivable nor locatable in the current utterance space. This includes
for instance, cognitive salience, as in the Hausa distinction between High and Low
focus deictics (Anderson and Keenan 1985: 289, cf. Jaggar and Buba 1994). It also
includes oppositions between Prospective (here we go, Im going this way) and
Retrospective (so thats my story [resumptive conclusion of narrative], I came
that way). Prospect and retrospect are closely related to future and past respectively, and when deployed in Discourse, they merge with cataphora and anaphora.
In many languages, such as English, non-proximal forms are used for anaphora (recall) and proximal ones for cataphora (anticipation), but in some, there are special
anaphoric deictics.
Heath (1980) cites a special Anaphoric base in Nungubuyu and Hanks (1990)
shows that Yucatec spatial deictics have a form unmarked for distance or interiority, but used for reference to places known (either anaphoric or recognitional). The so called recognitional usage, identified by Schegloff (1972) and cited
by Diessel (1999), denotes usages of names, descriptions and deictics in which the
Spr assumes that the Adr will be able to identify the referent without further specification. Essentially, this is a species of cognitive access based on common
ground. Hanks (1990) and Diessel (1999: 103, 113) distinguish anaphora proper
(coreference with a prior expression in the discourse) from reference to prior discourse (as in English Thats an outrageous claim, said in response position). In
331
any spatialist account, these uses are treated either as secondary, non-deictic uses
(as in Diessel 1999), or as metaphorical extensions from space (as in Anderson and
Keenan 1985). In an interaction-based approach in which the core function of deixis is to signal participant access and direct the Adrs attention to referents, they
demonstrate that cognitive access functions alongside spatial and perceptual access. Terminological differences aside, any language in which indexical reference
were limited to the exophoric field of utterance would be too constrained to meet
the demands of ordinary talk. This is obvious in any language with dedicated forms
for cognitive access, but is a factor of the pragmatics of deixis in any language.
4.4.
Features of the indexical origo
Among the highly significant contributions made by Bhler (1934) was his discussion of what we have called deictic transpositions. These are uses in which the
origo of deictic reference is shifted so that forms normally understood in relation to
the Spr are grounded on the Adr, some other object, or a narrative space. An easy
example is provided by talk to young children, in which a caregiver speaks for the
child e.g. a mother saying to a grimacing infant Oh Mommy, I dont like that. As
in quoted speech, the transposed deictics are grounded in the origo of another, in
this case the addressee. A narrator can project herself into a narrative space, as
when one retells an event from the past, saying So here I am walking down the
street and this guy comes up to me and asks me my name. The use of the proximal
deictic and the verb come are grounded in the narrative, not in the current situation of speech. A related transposition takes place whenever a distant place or object is imaginatively projected into the current indexical ground of utterance by
using unexpected proximal forms: So this is Paris, this is Berlin and this is Strasbourg (all on a table-top). Similarly, even very small utterance spaces can be
treated as if they were large, thus licensing use of normally remote forms for objects actually within reach: youve left your queen way over there (in reference to
a chess piece on a board between the interlocutors). What these examples have in
common is that deictic usage in effect manipulates the indexical ground of utterance. In the examples cited so far, these are purely pragmatic effects with little or
no grammatical conventionalization.
By contrast, many languages have special deictic forms whose function is to
signal shifts in the indexical ground. Anaphora and recognitional uses can be seen
in this light if we consider that they function only when both Spr and Adr share
cognitive access to the object. Thus anaphoric forms are symmetric in that they
presuppose that both participants can recover the antecedent. From a pragmatic
perspective, symmetry of access is the prototypical condition for deixis. In the
face-to-face, what is close to one party is also close to the other, and their perceptual fields are convergent (overlapping or reciprocal). Key aspects of cognitive access either rest on common ground or are co-constructed. By contrast, actual ego-
332
William F. Hanks
centric or altercentric uses are asymmetric because the origo attaches selectively to
one party but not the other. Whenever participants are spatially, perceptually or
cognitively separated, the ground is de facto asymmetric. Languages have various
ways of encoding different types of origo, as is clear in the so-called person-centered systems.
According to Frei (1944), Bisaya marks a six-way distinction between Distal
and Proximal relative to three distinct origos: Spr, Adr and Spr+Adr. Anderson and
Keenan (1985:287) cite demonstratives in Bemba, which mark a five-way distinction: Immediate (Spr), more Immediate to Spr than to Adr, Immediate (Spr + Adr),
Immediate (Adr) and Distal (Spr + Adr). These two languages appear to be maximal in that they distinguish both Proximal and Distal relative to all three origos.
Bemba is evidently unique in conventionally distinguishing degrees of relative
symmetry of access to the object. Chinook (Boas 1911) and Kwakwala (Anderson
and Keenan 1985: 290) are similar in that they lack a distal category, but mark
Proximal relative to three distinct grounds: Spr, Adr and Other. Japanese (Diessel
1999: 59), Ewondo (Diessel 1999: 40), Spe (Anderson and Keenan 1985: 287) and
Quileute (Andrade 1933) are nearly identical in distinguishing Prox (Spr) vs. Prox
(Adr) vs Medial or Distal (Spr+Adr), and Diessel cites several other languages
with a similar paradigm. Inuktitut (Denny 1982: 362) illustrates yet another solution, with a special affix marking transpositions from the Speaker-ground to
Other (including Adr or third).
There are several reasons that it is challenging to actually demonstrate these different values for the indexical ground. Detailed observation of ordinary usage is
requisite, and most of the published literature simply fails to provide the necessary
evidence. In any language lacking dedicated symmetric forms (for which the common ground is the origo), one must infer or guess at how the attested forms are used
when access to the object is roughly the same for both participants. In languages
with a three-way split among SprAdrOther, either the Other category covers
common access, or the use of either egocentric or altercentric forms pragmatically
implies common ground. Similarly, in languages like Japanese, the question is how
one denotes an object proximal to both parties, since shared access is apparently
signalled in the distal form only. Heath (1980) does not mention the ground for
either Proximal or Distal deictics in Nungubuyu, whereas separate forms mark Prox
(Adr) and Anaphoric (which implies shared cognitive access). Presumably the Prox
and Distal categories may be either egocentric (Spr) or symmetric (Spr+Adr).
If we take the published descriptions at face value, the following generalizations emerge: (1) No language has a distinctive category for the Spr-origo, unless
it also has a form for Adr-origo. In the absence of an Adr-origo form, the attested
forms are simply unmarked and likely to be used for egocentric, altercentric and
shared ground. Adr-centered deixis is more pragmatically marked than Spr-centeric deixis, and more likely to be overtly signalled. (2) Medial and Distal categories tend strongly to relate to symmetric origo (Spr+Adr). Even if Speaker-cen-
333
tric and Addressee-centric forms are distinguished in the Proximal deictics, Distal
forms are likely to be symmetric (Spr+Adr). (3) In languages lacking a dedicated
Anaphoric form, the non-Proximal forms are most likely to subserve this function
(because they share the indexical symmetry of Distal categories). (4) Immediate
forms mark higher salience than their non-Immediate counterparts, and are more
likely to be used for new information, high-focus and prospective cognitive access.
There are many more subdivisions among deictics in the published literature,
and long-term fieldwork in any language demonstrates the complexity of decitic
practices, in which all of the foregoing factors may be simultaneously in play. In
numerous languages for instance, there are special forms to distinguish precise or
punctate from vague or regional reference. Compare Punctual in Yucatec Maya
(Hanks 1990), Concrete in Nungubuyu (Heath 1980), Precise in Tsimshian (Boas
1911, Dunn 1979), Intensive in Santali (Zide 1972) and Restriced in Inuktitut
(Denny 1982: 372). Insofar as these categories serve to mark the relative precision
with which the object is individuated, as in this very one vs. this one, they nicely
display one of the functional requisites of deictic reference, that is, the resolution
of reference. Such resolution relies on the co-operation of spatial, perceptual and
cognitive access, along with the range of indexical origos to which they are anchored.
Language key to items in bibliography

N ATIVE N ORTH A MERICA
Tsimshian
Tlingit
Chinook
Wasco-Wishram
Bella Coola
Crow
Mohawk
Oneida
Inuktitut
Aleut
Quileute
Greenlandic
M ESOAMERICA
Tlapanec
Yucatec
Toba
Quechua
AND
A RCTIC
Boas 1911b, Dunn 1979
Boas 1917, Swanton 1911
Boas 1911a
Silverstein 1978
Davis and Saunders 1975, 1976
Graczyk 1986
Bonvillain 1981
Abbott 1981
Denny 1978, 1982
Bergsland 1951
Andrade (1933)
Fortescue 1984
AND
L ATIN A MERICA
Wichmann 1993
Hanks 19832009 (various)
Manelis Klein 1981
Calvo Perez 1999
334
William F. Hanks
I NDO -E UROPEAN
French
Germanic
Russian
A FRICA
Hausa
Swahili
Malagasy
Morel and Danon-Boileau, eds. 1992

Brugmann 1904, Bhler 1934
Friedrich 1979
Hill 1982, Jaggar and Buba 1994

Leonard 1982, 1985
Cousins 1894, Garvey 1964, Dez 1980, Anderson
and Keenan 1985, Hanks 1986
A SIA (S OUTH , S OUTHEAST ), P APUA N EW G UINEA

Mandarin
Tao 1999
Japanese
Kuno 1973
Santali (Munda)
Zide 1972
Belhare (Sino-Tibet) Bickel 2000
Lao
Enfield 2003, n.d.
Lahu
Matisoff (1973)
Jahai
Burenhult 2003, n.d.
Yl Dnye
Levinson n.d. (to appear)
A USTRALIA
Nunggubuyu
Ngandi
Mara
Guugu Yimithirr
Dyirbal
Wardaman
Kayardild
Heath 1980
Heath 1978
Heath 1981
Haviland 1993
Dixon 1972
Merlan 1994
Evans 1995
Notes
1. Cf. Fillmore 1971, Silverstein 1976, Hanks 1984, 1990, 2005a, Diessel 1999, Enfield
2003.
2. Cf. Morris 1946, Peirce 1955, Eco et al. 1988, Husserl 1978, Benveniste 1974, cf. Searle
1969.
3. Cf. Jespersen 1965 [1924], Jakobson 1971 [1957], Silverstein 1976.
4. Cf. Hymes 1974, Gumperz and Hymes 1964, Sherzer 1983, 1991.
5. Cf. Hanks 1990, 2005a, Himmelmann 1996, Janssen 2000, 2002, Leonard 1985, Mondana 2005.
6. See Garfinkel 1967, Schegloff 1972, Enfield 2009, Sidnell 2010.
7. See Clark and Marshall 1981, Clark, Schreuder and Buttrick 1983, Enfield 2006.
335
8. Cf. Frei 1944, Anderson and Keenan 1985, Diessel 1999: 158, Lyons 1977, Senft and
Smits 2000: 6580, Levinson 2003.
9. See Sacks 1992, Goodwin 1981, Schegloff 1984, Heritage 1984, Sidnell 1997, 2005.
10. Cf. Haviland 1993, 2001, 2003, Goodwin 1994, 2000, Kita 2003, Enfield 2009.
11. Cf. Sacks, Schegloff and Jefferson 1974, Schegloff 1987, 1992, Sidnell 2005, 2010,
Stivers 2007.
12. Cf. Schutz 1970, Goodwin 1994, Goodwin and Heritage 1990, Sacks 1992, Schegloff
1972, 1992.
13. See Haviland 1996, 2003, Goodwin 1981, 2003, Sidnell 1997, 2005 and Ochs, Schegloff and Thompson 1996.
14. See Diessel 2005 for similar generalizations over a larger sample of languages.
15. Compare Russell 1940, Reichenbach 1947, Frei 1944, Kurylowicz 1972, Lyons 1982,
Fillmore 1997.
16. For recent discussions of deictic typology, see Himmelmann 1996, Fillmore 1997 and
Diessel 1999.
17. Cf. Hanks 1990, where I note that Maya and Inuktitut illustrate interiority as well as
does Greenlandic; Santali indicates lateral to origo, which I suggest is related to interiority. There is something like a family of effects around interiority, inclusiveness,
edge-ness (at boundary of origo). My use of interiority, perhaps too vague, is meant to
cover this range.
18. From the viewpoint of ordinary usage and native metalinguistic commentaries, it is
clear that Yucatec speakers distinguish a three-way split between Tactual, Visual and
Peripheral Sensory access, although the comparative evidence suggests that the Peripheral Sensory category may be better analyzed as [+ Perceived, +Invisible], similar to the
systems in other languages. In any event it is clear that vision is the most commonly
marked modality and it is the absence of visual access that usually gets special marking.
References
Abbott, Clifford
1981
Here and there in Oneida. International Journal of American Linguistics 47(1):
5057.
Anderson, Stephen and Edward Keenan
1985
Deixis. In: Timothy Shopen (ed.), Language Typology and Syntactic Description, vol. 3: Grammatical Categories and the Lexicon, 259308. Cambridge:
Andrade, Manuel J.
1933
Quileute. Handbook of American Indian Languages 3, 151292. New York:
Columbia University Press.
Bar Hillel, Yehoshua
1954
Indexical expressions. Mind LXIII: 359379.
Benveniste, Emile
1965
Language and Human Experience. Diogenes 51: 112.
Benveniste, Emile
1966
De la subjectivit dans le langage. In: Problmes de Linguistique Gnrale,
336
William F. Hanks
Benveniste, Emile
1974
Le langage et lexprience humaine. In: Problmes de Linguistique Gnrale,
Benveniste, Emile
1974
Problmes de Linguistique Gnrale, vol. 2. Paris: Gallimard.
Bergsland, Knut
1951
Kleinschmidt Centennial IV: Aleut demonstratives and the Aleut-Eskimo relationship. International Journal of American Linguistics 17(3): 167179.
Bickel, Balthazar
2000
Deictic transposition and referential practice in Belhare. Journal of Linguistic
Anthropology 10(2): 224247.
Bloomfield, Leonard
1933
Language. New York: Holt.
Boas, Franz
1911a
Chinook. In: Handbook of American Indian Languages. Smithsonian Institution, Bureau of American Ethnology, Bulletin 40, ed. by Franz Boas. 2 vols, I:
Boas, Franz
1911b
Tsimshian. In: Handbook of American Indian Languages. Smithsonian Institution, Bureau of American Ethnology, Bulletin 40, ed. by Franz Boas. 2 vols. I:
Boas, Franz
1917
Grammatical Notes on the Language of the Tlingit Indians. Vol. VIII, No. 1.
Philadelphia: Anthropological Publications of the University Museum.
Bonvillain, Nancy
1981
Locative semantics in Mohawk: Time and space. International Journal of
Brugmann, Karl
1904
Demonstrativpronomina in den indogermanischen Sprachen. Leipzig: Teubner.
Bhler, Karl
1934
Sprachtheorie. Jena: Springer.
Burenhult, Niclas
2003
Attention, accessibility and the addressee: The case of the Jahai demonstrative
ton. Pragmatics 13(3): 363379.
Burenhult, Niclas
To appear Distance undone: new light on the Jahai multi-term demonstrative system.
In: Michael Dunn and Srgio Meira (eds.). Demonstratives in cross-linguistic
perspective.
Calvo Perez, Julio
1999
Pragmatica y gramatica del Quechua Cuzqueo. Cuzco: Centro de Estudios
Regionales Andinos Bartolome de las Casas.
Chapman, Mackenzie
2009
A brief introduction to demonstratives in Mandarin Chinese. (Unpublished
paper, Summer Institute of Linguistics of the Linguistic Society of America,
7pp).
337
Cicourel, Aaron
1964
Method and Measurement in Sociology. New York: Free Press.
Cicourel, Aaron
1974
Cognitive Sociology: Language and meaning in social interaction. New York:
Free Press.
Cicourel, Aaron
1985
Text and discourse. Annual Reviews of Anthropology 14:159185.
Cicourel, Aaron
1992
The interpenetration of Communicative contexts. In: Alessandro Duranti and
Charles Goodwin (eds.), Rethinking Context: Language as an interactive phenomenon, 291310. Cambridge: Cambridge University Press.
1981
Webber and Ivan A. Sag (eds.), Elements of Discourse Understanding, 1063.
1983
Common ground and the understanding of demonstrative reference. Journal of
Verbal Learning and Verbal Behavior 22: 245258.
Cousins, William E.
1894
A Concise Introduction to the Study of the Malagasy Language as Spoken in
Imerina. 3rd ed. Antananarivo: The Press of I.M.S.
1975
Bella Coola Nominal Deixis. Language 51(4): 845858.
1976
Bella Coola Deictic Roots. International Journal of American Linguistics
42(4): 319330.
Denny, J. Peter
1978
Locating the universals in lexical systems for spatial deixis. In: Donka Farkas,
Wesley M. Jacobsen, and Koral W. Todrys (eds.), Papers from the Parasession
on the Lexicon, 7184. Chicago: Chicago Linguistic Society.
Denny, J. Peter
1982
Semantics of the Inuktitut (Eskimo) spatial deictics. International Journal of
Dez, Jacques
1980
Structures de la Language Malgache, lments de grammaire lusage des
francophones. Paris: Publications Orientalistes de France.
Diessel, Holgar
1999
Demonstratives: Form, function and grammaticalization. Amsterdam: Benjamins.
Diessel, Holger
2005
Distance contrasts in demonstratives. In: Martin Haspelmath, Matthew Dryer,
David Gil, and Bernard Comrie (eds.), World Atlas of Language Structures,
Dixon, Robert M.W.
1972
The Dyirbal Language of Northern Queensland. Cambridge: Cambridge University Press.
338
William F. Hanks
Dixon, Robert M.W.

1980
The Languages of Australia. Cambridge: Cambridge University Press.
Dunn, John A.
1979
A Reference Grammar for the Coast Tsimshian Language. Canadian Ethnology Service, Paper 55. Ottawa: National Museum of Man, Mercury Series.
Eco, Umberto, Marco Santambrogio and Patrizia Violi
1988
Meaning and mental representations. Bloomington, IN: Indiana University
Press.
2003
Demonstratives in space and interaction: Data from Lao speakers and Implications for semantic analysis. Language 79(1): 82117.
2006
Social consequences of common ground. In: Nicholas Enfield and Stephen Levinson (eds.), Roots of Human Sociality, Culture, Cognition and Interaction,
2009
The Anatomy of Meaning, speech, gesture and composite utterances. Cambridge: Cambridge University Press.
Enfield, Nicholas J. and Tanya Stivers (eds.)
2007
Person Reference in Interaction, Linguistic, Cultural and Social Perspectives.
Evans, Gareth
1982
The Varieties of Reference. New York: Oxford University Press.
Evans, Nicholas
1995
A grammar of Kayardild. Berlin: Mouton de Gruyter.
Evans, Nicholas
2003
Bininj Gun-wok: a pan-dialectal grammar of Mayali, Kunwinjku and Kune.
2 vols., xxix + 746. Canberra: Pacific Linguistics.
1971
Santa Cruz Lectures on Deixis. Bloomington, IN: Indiana University Linguistics Club.
Fillmore, Charles
1982
Towards a linguistic framework for spatial deixis. In: Robert Jarvella and Wolfgang Klein (eds.), Speech, Place and Action, 3159. Chichester: John Wiley.
Fillmore, Charles
1997
Lectures on Deixis. Stanford: CSLI publications.
Fortescue, Michael
1984
West Greenlandic. London: Croom Helm.
Frei, Henri
1944
Systmes de dictiques. Acta Linguistica 4: 111129.
Friedrich, Paul
1979
Language, Context and the Imagination. Stanford: Stanford University Press.
Gale, Richard M.
1968
Is it now now? Mind 73: 97105.
Gale, Richard M.
1968
Indexical expressions, egocentric particulars and token-reflexive words. In:
P. Edwards (ed.), Encyclopedia of Philosophy. Vol IV, 151155. New York:
Collier Macmillan.
339
Garfinkel, Harold
1967
Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice-Hall.
Garfinkel, Harold
1972
Studies in the routine grounds of everyday activities. In: David Sudnow (ed.),
Garvey, Catherine
1964
Malagasy Introductory Course. Washington DC: Center for Applied Linguistics.
Goodwin, Charles
1981
Conversational Organization: Interaction between speakers and hearers. New
York: Academic Press.
Goodwin, Charles
1994
Professional Vision. American Anthropologist 96(3): 606633.
Goodwin, Charles
2000
Practices of Seeing, Visual Analysis: An Ethnomethodological Approach. In:
Theo van Leeuwen and Carey Jewitt (eds.), Handbook of Visual Analysis,
Goodwin, Charles
2003
Pointing as situated practice. In: S. Kita (ed.), Pointing: Where language, culture and cognition meet, 217242. Hillsdale, NJ: Erlbaum.
Goodwin, Charles
2006
Human sociality as mutual orientation in a rich interactive environment: Multimodal utterances and pointing in aphasia. In: Nicholas Enfield and Stephen Levinson (eds.), Roots of Human Sociality, Culture, Cognition and Interaction,
Goodwin, Charles and John Heritage
1990
Conversation Analysis. Annual Review of Anthropology 19: 283307.
Graczyk, Randolph
1986
Crow deixis. Unpublished Ms., Department of Linguistics, University of Chicago.
Green, Keith (ed.)
1995
New Essays in Deixis. Amsterdam: Rodopi.
Grice, H. Paul
1967
Logic and conversation. William James Lectures, Cambridge, MA: Harvard
University.
Grice, H. Paul
1978
Further notes on logic and conversation. In: Peter Cole (ed.), Syntax and
Gumperz, John
1982
1964
The Ethnography of Communication. American Anthropologist, Special Publication 66, Number 6, Part 2.
1976
Hanks, William F.
1983
Deixis and the Organization of Interactive Context. Unpublished PhD thesis.
University of Chicago, Department of Linguistics, Department of Anthropology.
340
William F. Hanks
Hanks, William F.
1984
The evidential core of deixis in Yucatec Maya. In: Papers from the Fifteenth
Regional Meeting of the Chicago Linguistic Society, 154173. Chicago: Chicago Linguistic Society.
Hanks, William F.
1986
Markedness and category interactions in the Malagasy deictic system. University of Chicago, Department of Linguistics Working Papers. 20pp.
Hanks, William F.
1990
Referential Practice: Language and lived space among the Maya. Chicago:
Hanks, William F.
1992
The indexical ground of deictic reference. In: Alessandro Duranti and Charles
Goodwin (eds.), Rethinking Context: Language as an interactive phenomenon,
4377. Cambridge: Cambridge University Press. [Reprinted from Papers from
the Parasession on Language in Context. Chicago Linguistic Society, 1989].
Hanks, William F.
1993a
Metalanguage and pragmatics of deixis. In: John Lucy (ed.), Reflexive language: Reported speech and metapragmatics, 127157. Cambridge: Cambridge University Press.
Hanks, William F.
1993b
Notes on semantics in linguistic practice. In: Craig Calhoun and Moise Postone (eds.), Towards a reflexive sociology: The social theory of Pierre Bourdieu, 139155. Oxford: Basil Blackwell.
Hanks, William F.
1995
Language and Communicative Practices. Critical Essays in Anthropology.
Edited by Maurice Bloch, Pierre Bourdieu, and John L. Comaroff, Boulder:
Westview Press.
Hanks, William F.
1996
Language form and communicative practices. In: John Gumperz and Stephen
Levinson (eds.), Rethinking Linguistic Relativity, 232270. Cambridge: Cambridge University Press.
Hanks, William F.
1999
Intertexts: Writings on Language, Utterance and Context. Denver: Rowman
and Littlefield.
Hanks, William F.
2001
Indexicality. In: Alessandro Duranti (ed.), Key terms in language and culture,
119121. Oxford: Blackwell.
Hanks, William F.
2005a. Explorations in the deictic field. Current Anthropology 46(2): 191220.
Hanks, William F.
2005b
Pierre Bourdieu and the Practices of Language. Annual Review of Anthropology 34: 6783.
Hanks, William F.
2006a
Context. In: Encyclopedia of Language and Linguistics, 2nd edition. Ed. by
Keith Brown. Oxford: Elsevier.
Hanks, William F.
2006b
Deixis. In: Encyclopedia of Literature. London: Routledge.
341
Hanks, William F.
2006c
Spatial frames of reference in language and thought. Review article of Space in
Language and Cognition, by Stephen Levinson. Language in Society 35(2):
285296.
Hanks, William F.
2007
Construal of persons in Maya conversation. In: Tanya Stivers and Nicholas Enfield (eds.), Person Reference in Interaction: Linguistic, cultural and social
Hanks, William F.
2007b
The evidential core of deixis in Yucatec Maya. In: Zlatka Guerncheva (ed.),
Lnonciation mdiatise. Tome II, 311334. Paris-Louvain: Peeters.
Hanks, William F.
2008
Lngua como prtica social: das relaes entre lngua, cultura e sociedade a
partir de Bourdieu e Bakhtin. Cortez Editora: So Paulo, Brasil.
Hanks, William F.
2009
Fieldwork on Deixis. Journal of Pragmatics 41: 1024. Special issue on
Emancipatory Pragmatics. Ed. by William Hanks, Sachiko Ide, Yasuhiro Katagiri.
Haviland, John B.
1993
Anchoring, iconicity and orientation in Gugu Yimithirr pointing gestures.
Journal of Linguistic Anthropology 3(1): 345.
Haviland, John B.
1996
Projections, transpositions and relativity. In: John Gumperz and Stephen Levinson (eds.), Rethinking Linguistic Relativity, 271323. Cambridge: Cambridge University Press.
Haviland, John B.
2001
Gesture. In: Alessandro Duranti (ed.), Key Terms in Language and Culture,
8386. Malden, MA: Blackwell.
Haviland, John B.
2003
How to point in Zinacantn. In: Sotaro Kita (ed.), Pointing: Where language,
culture and cognition meet, 139170. Mahwah, NJ: Erlbaum.
Heath, Jeffrey
1978
Ngandi Grammar, Texts and Dictionary. Australia Institute of Aboriginal
Studies, Canberra. Atlantic Highlands, NJ: Humanities Press.
Heath, Jeffrey
1980
Nunggubuyu deixis, anaphora and culture. In: Jody Kreiman and Almerindo
Ojeda (eds.), Papers from the Parasession on Pronouns and Anaphora,
151165. Chicago: Chicago Linguistic Society.
Heath, Jeffrey
1981
Basic Materials in Mara: Grammar, Texts and Dictionary. (Pacific Linguistics,
Series C, No. 60.) Department of Linguistics, Research School of Pacific
Studies. The Australian National University.
Heath, Jeffrey
2004
Person. In: Geert Booij, Christian Lehmann, Joachim Mugdan and Stavros
Skopeteas (eds.), Morphologie: Ein internationales Handbuch zur Flexion und
Wortbildung / Morphology: An International Handbook on Inflection and
Word-Formation. Vol. 2, 9981015. Berlin: de Gruyter.
342
William F. Hanks
Heritage, John
1984
A change-of-state token and aspects of its sequential placement. In: Maxwell
Atkinson and John Heritage (eds.), Structures of social action: Studies in conversation analysis, 225246. Cambridge: Cambridge University Press.
Hill, Clifford
1982
Up/down, front/back, left/right: A contrastive study of Hausa and English. In:
Jrgen Weissenborn and Wolfgang Klein (eds.), Here and There: Cross-linguistic Studies on Deixis and Demonstration, 1342. Amsterdam: Benjamins.
Himmelmann, Nicholas P.
1996
Demonstratives in narrative discourse: A taxonomy of universal uses. In: Barbara Fox (ed.), Studies in Anaphora. 205254. Amsterdam: Benjamins.
Husserl, Edmund
1978
The origin of geometry. In: Thomas Luckmann (ed.), Phenomenology and
Sociology: Selected Readings, 4270. New York: Penguin.
Hymes, Dell
1974
Jaggar, Phillip and Malami Buba
1994
The space and time adverbials NAN/CAN in Hausa: cracking the deictic code.
Language Sciences 16: 387421.
Jakobson, Roman
1960
Closing statement on linguistics and poetics. In: Thomas Sebeok, (ed.), Style in
Language, 350377. Cambridge, MA: MIT Press.
Jakobson, Roman
1971 [1957] Shifters, verbal categories and the Russian verb. In: Roman Jakobson: Selected Writings. Vol. 2: Word and Language, 130147. The Hague: Mouton.
Janssen, Theodorus
2000
Deixis and reference. In: Geert Booij, Christian Lehmann, Joachim Mugdan
and Stavros Skopeteas (eds.), Morphologie: Ein internationales Handbuch zur
Flexion und Wortbildung / Morphology: An International Handbook on Inflection and Word-Formation, 983998. Berlin: de Gruyter.
Janssen, Theodorus
2002
Deictic principles of pronominal, demonstrative and tense deictics. In: Frank
Brisard (ed.), Grounding: The epistemic footing of deixis and reference,
Jespersen, Otto
1965 [1924] The Philosophy of Grammar. New York: WW Norton.
Jungbluth, Konstanze
2005
Pragmatik der Demostrativpronomina in den iberoromanischen Sprachen.
Tbingen: Niemeyer.
Kendon, Adam
1992
The negotiation of context in face-to face interaction. In: Alessandro Duranti
and Charles Goodwin (eds.), Rethinking context, 323334. Cambridge: Cambridge University Press.
Kita, Sotaro (ed.)
2003
Pointing: Where language, culture, and cognition meet. Mahwah, NJ: Erlbaum.
343
Kita, Sotaro and James Essegbey

2001
Pointing left in Ghana: How a taboo on the use of the left hand influences gestural practice. Gesture 1(1): 7395.
Kuno, Susumo
1973
The Structure of the Japanese Language. Cambridge, MA: MIT Press.
Kurylowicz, Jerzy
1972
Universaux linguistiques. In: Luigi Heilman (ed.), Proceedings of the Eleventh
Congress of Linguists, 4773. Bologna.
Lasater, Ann
2009
Notes on Quechua deixis. (Unpublished Ms., Summer Institute of Linguistics
of the Linguistic Society of America, Berkeley, 9 pp.)
Leonard, Robert
1982
The semantic system of deixis in Swahili. PhD thesis, Columbia University.
Leonard, Robert
1985
Swahili demonstratives: Evaluating the validity of competing semantic hypotheses. Studies in African Linguistics 16(3): 281293.
Levinson, Stephen
1983
Levinson, Stephen
2003
Space in language and cognition: Explorations in cognitive diversity. Cambridge: Cambridge University Press.
Levinson, Stephen
To appear Demonstratives in Yl Dnye, the language of Rossel Island Papua New
Guinea. In: Michael Dunn and Srgio Meira (eds.), Demonstratives in crosslinguistic perspective. 29 pp.
Levinson, Stephen et al.
n.d
Demonstratives: Towards a semantic typology. MPI Nijmegen working paper
(Table, 1 page).
Linell, Per
2001
Approaching Dialogue, talk, interaction and contexts in dialogical perspectives. Philadelphia: Benjamins.
Lyons, John
1977
Semantics. Cambridge: Cambridge University Press.
Lyons, John
1982
Deixis and subjectivity: Loquor ergo sum? In: Robert Jarvella and Wolfgang
Klein (eds.), Speech, Place and Action: Studies in Deixis and Related Topics,
101124. New York: Wiley.
Manelis Klein, Harriet E.
1981
Location and Direction in Toba: Verbal morphology. International Journal of
Martinet, Andr
1960
Elements de Linguistique Gnrale. Paris: A. Colin.
Matisoff, James
1973
The Grammar of Lahu. Berkeley: University of California Press.
Meillet, Antoine
1926
Linguistique historique et Linguistique Gnrale. Paris: Champion.
344
William F. Hanks
Merlan, Francesca
1994
A Grammar of Wardaman: A Language of the Northern Territory of Australia.
Berlin: de Gruyter.
1967 [1945] The Phenomenology of Perception. Paris: Gallimard.
Mondana, Lorenza
2005
La constitution de lorigo dictique comme travail interactionnel des participants: une approche praxologique de la rfrence spatiale. Intellectica 4142:
75100.
Morel, Mary-Annick and Laurent Danon-Boileau (eds.)
1992
La Dixis, Colloque en Sorbonne (89 juin 1990). Paris: Presses Universitaires
de France.
Morris, Charles
1946
Signs, language, and behavior. New York: George Braziller.
Nunberg, Geoffrey
1993
Indexicality and deixis. Linguistics and Philosophy 16(1): 143.
Ochs, Elinor, Emanuel Schegloff and Sandra Thompson (eds.)
1996
Interaction and Grammar. Cambridge: Cambridge University Press.
Ozyrek, Asli and Kita Sotaro
n.d.
Joint attention and distance in the semantics of Turkish and Japanese demonstrative systems. (Ms)
Peirce, Charles S.
1955
Philosophical writings of Peirce. Ed. by Justus Buchler. New York: Dover
Publications.
Reichenbach, Hans
1947
Elements of Symbolic Logic. New York: Macmillan.
Rommetveit, Ragnar
1974
On Message Structure. London: Wiley.
Russell, Bertrand
1940
An Inquiry into Meaning and Truth. London: Allen and Unwin.
Sacks, Harvey
1992
Lectures on conversation. 2 volumes. Ed. by Gail Jefferson. Oxford: Blackwell.
1974
A Simplest Systematics for the Organization of Turn-Taking in Conversation.
Language 50(4): 696735.
Sadock, Jerrold
1978
Schegloff, Emanuel
1972
Note on a conversational practice: Formulating place. In: David Sudnow (ed.),
Schegloff, Emanuel
1982
Discourse as an interactional achievement. In: Deborah Tannen (ed.), Analyzing Discourse: Text and Talk. Georgetown University Round Table on Languages and Linguistics 1981, 7194. Washington: Georgetown University
Press.
345
Schegloff, Emanuel
1984
On some questions and ambiguities in conversation. In: J. Maxwell Atkinson
and John Heritage (eds.), Structures of social action: Studies in conversation
analysis. 2852. Cambridge: Cambridge University Press.
Schegloff, Emanuel
1987
Between micro and macro: Context and other connections. In: Jeffrey Alexander, Bernhard Giesen, Richard Mnch and Neil J. Smelser (eds.), The MicroMacro Link, 207234. Berkeley: University of California Press.
Schegloff, Emanuel
1992
In another context. In: Alessandro Duranti and Charles Goodwin (eds.), Rethinking context, 191288. Cambridge: Cambridge University Press.
Schutz, Alfred
1970
Phenomenology and Social Relations. Ed. by Helmut R. Wagner. Chicago:
Searle, John R.
1969
Speech Acts. Cambridge: University Press.
Senft, Gunter and Roel Smits (eds.)
2000
Annual report 2000. Nijmegen: Max Planck Institute for Psycholinguistics.
Sherzer, Joel
1983
Kuna Ways of Speaking. Austin: University of Texas Press.
Sherzer, Joel
1991
Verbal Art in San Blas. Cambridge: Cambridge University Press.
Sidnell, Jack
1997
Organizing Social and Spatial Location: Elicitations in Indo-Guyanese Village
Talk. Journal of Linguistic Anthropology 7(2): 143165.
Sidnell, Jack
2005
Talk and Practical Epistemology: The Social Life of Knowledge in a Caribbean Community. Amsterdam: Benjamins.
Sidnell, Jack
2010
Conversation Analysis, Comparative Perspectives. Cambridge: Cambridge
University Press.
1976
Shifters, linguistic categories, and cultural description. In: Keith Basso and
Henry Selby (eds.), Meaning in Anthropology, 1155. Albuquerque, NM: University of New Mexico Press.
1978
Deixis and deducibility in a Wasco-Wishram passive of evidence. In: Proceedings of the Fourth Annual Meeting of the Berkeley Linguistics Society,
238253. Berkeley, CA: Berkeley Linguistics Society.
Stivers, Tanya
2007
Alternative recognitionals in person reference. In: Tanya Stivers and Nicholas
Enfield (eds.), Person Reference in Interaction: Linguistic, cultural and social
Swanton, John
1911
Tlingit. In: Handbook of American Indian Languages. Smithsonian Institution,
Bureau of American Ethnology, Bulletin 40, ed. by Franz Boas. 2 vols., I:
346
William F. Hanks
Tao, Hongyin
1999
The grammar of demonstratives in Mandarin conversational discourse: A case
study. Journal of Chinese Linguistics 27(1): 69103
Weissenborn, Jrgen and Wolfgang Klein (eds.)
1982
Here and There: Crosslinguistic Studies on Deixis and Demonstration. Amsterdam: Benjamins.
Wichmann, Sren
1993
Spatial deixis in Azoy Tlapanec. Sprachtypologie und Universalienforschung
(STUF), Berlin, 46(2): 113132.
Wilkins, David
1999
Demonstrative questionnaire: THIS and THAT in comparative perspective. In: David Wilkins (ed.), Manual for the 1999 field session. Nijmegen:
Max-Planck Institute for Psycholinguistics.
Yang, Yuling
2006
A study on the asymmetry of individual zhe and individual na in contexts [in
Chinese]. Chinese Teaching in the World 4: 3341.
Zide, Norman
1972
A Munda demonstrative system: Santali. In: Jacqueline M.C. Thomas and Lucien Bernot (eds.), Langues et Techniques, Nature et Socits I. Approche Linguistique, 267274. Paris: Editions Klincksieck.
12.

Introduction
This chapter will provide an overview of pragmatic research on how people use
and comprehend referring expressions. A referring expression is typically defined
as one which is produced in order to identify or re-identify (in case of anaphora)
an object or a set of objects for a listener or reader in a relevant domain of text or
discourse. Thus, reference and anaphora are concepts inherently pragmatic in nature. Without considering pragmatic notions (such as focus of information, world
knowledge, contextual influence) their description in linguistic theory would remain insufficient. We will focus on past and current research on reference and anaphora. One line of this research is to delineate the mental processing of referring
expressions with the focus on text comprehension.
1.
Notions of reference
Reference (-theory) is the area of scientific research which deals with the ancient
problem of how language relates to the world. Reference is a term that has undergone major changes during the last five decades of research. In this section, the development from a static notion of reference as a relation independent from language users (1.1) to a dynamic notion of reference as a cognitive process (1.2) is
outlined.
1.1.
Static notions of reference
In the tradition of philosophical logics and formal semantics, reference has been
defined as a static relationship between expressions and the things they denote. The
basis for this notion is an objective view of the world. For many philosophers of
language, the abstract relationship between words and things is primary, irrespective of intentions, contextual influence and conceptual knowledge.
In this view, the function of language is to depict reality; each word stands for a
certain class of objects. Linguistics concentrated on the denotation of different
kinds of referential expressions. The distinction between sense and reference, formulated by Frege a century ago (1892), had been in the focus of semantic research
for quite some time. A semantic theory had to specify for each expression in a language what semantic information that expression conveyed. In the tradition of
348
Russell (1905), who has defined referring as asserting the existence and the
uniqueness of an object, the referential potential of definite descriptions and proper
names became a field of intensive linguistic research (see Thrane 1980). Apart
from the question whether reference is regarded as a static relation or as a process,
it has been discussed which kinds of noun phrase use are referential at all: Donnellan (1966) excluded attributive use of definite NPs from his notion of reference.
Thus, in his well-known example (1) the NP Smiths murderer is used referentially
only in case the speaker knows who Smiths murderer is and intends to identify
this individual for the hearer. Involving hearers roles in reference, Donnellan
approaches a dynamic notion of the term (see 1.2). Still, it crucially depends on
truth conditionality, and there is no notion of reference to mental concepts like
W HOEVER HAS MURDERED S MITH let alone hypothetical referents.
(1) Smiths murderer must be insane.
The main topics in reference theory were naming, truth and extension. By focussing on the semantic features of definite noun phrases and reference to real physical
objects, however, the actual use of referential expressions was largely neglected.
According to formal semantics, reference is not related to an intersubjective, mental or perceptual world, but to a formal model that functions as a substitute for reality (Dowty 1979). In model-theoretic semantics, the intension of an expression is a
function from possible worlds to truth values, and an extension (i.e. the set of possible referents) is the truth value for a particular world. Thus, reference theory in
modern linguistics has been dominated by a formal and not an empirical approach.
A reference theory of truth conditions, however, disregards the mental apparatus of
language users. This semantic, static notion of reference is still (albeit implicitly)
prevailing in some contemporary formal semantic approaches like Discourse Representation Theory.1
1.2.
Dynamic notions of reference
Dynamic notions of reference crucially arose in the framework of pragmatic approaches like the speech act theory; however, some semanticists made attempts to
integrate procedural aspects of reference: Strawson (1950) claimed (at least implicitly) that speakers have to account for the recipients capability of identifying a
referent and that an act of reference has to be accomplished by the recipient. Reference as a dynamic notion is thus a collaborative activity of speakers and hearers
who activate extra-linguistic mental representations by using language. It is based
on a common ground, i.e. knowledge shared by speakers and recipients (Clark and
Marshall 1981; Clark, Schreuder, and Buttrick 1983). Linguists have realized that
it is a mistake to think that the referring expression itself can guarantee uniqueness
in discourse. We assign referential distinctness not simply on the basis of lexical
or grammatical features such as definiteness but also on some vaguer notions of
349
plausibility and assumptions about the speakers knowledge of the distinctive

properties of the domain (Stenning 1978).
In the 1970s interest became centered on the pragmatic aspects of reference.
Reference, it has been argued, is not an abstract semantic relation, but part of a
speech act. This is reflected in Donnellans (1978) terms of speakers reference
versus semantic reference (the latter term being equivalent to meaning). With
Searle (1969), a reference act is part of a propositional act, which consists of the
reference act and a predication about the referent, propositional act and illocutionary act being the core aspects of a speech act. However, pragmatics did not from
the outset pay equal attention to the speakers and the recipients roles in reference
acts: reference is merely defined by the speakers intention to make the recipient
recognize a referent (Donnellan 1966, 1978); accordingly, a reference act is felicitous in case the referent exists, regardless of the hearers reception. Similarly,
Searle (1969) defines successful reference from the speakers point of view, but
he accounts for recipients who, in this model, will ask for more specification until a
state of fully consummated reference is achieved.
In the late 1970s, interest shifted towards a more cognitive based explanation
of reference. Emphasis was now placed on the mental representations and operations involved in referential communication. How recipients identify referents
became part of the broader question of how recipients extract the novel information conveyed by an utterance and incorporate this information into memory
(Clark and Sengul 1979). How do we process referential expressions in on-line
understanding? became one of the main questions. Important work has been done
in psycholinguistics and cognitive psychology dealing with the constructive nature
of referential interpretation. Bransford, Barclay, and Franks (1972) and others have
demonstrated that memory representations of a sentence combine information provided by the meaning of the sentence with information drawn from real-world
knowledge. Empirical studies have shown that language understanding is a constructive process. In on-line-processing studies it has been shown that recipients
instantaneously use pragmatic information (real world knowledge as well as cotextual information) in resolving referential ambiguities and in interpreting anaphors (Marslen-Wilson and Tyler 1982; McDonald and MacWhinney 1995). Accordingly, the structure of semantic memory and the representation of concepts as
part of world knowledge has become the focus especially of neuropsychological
studies relevant for reference (Damasio et al. 1996; Caramazza and Mahon 2006;
Caramazza and Mahon 2009).
In keeping with this mentalistic framework, the concept of world has changed
as well: It is no longer presumed that reference implies the existence of an object in
the physical world (see 1.1.). In reference theory, the traditional realistic approach
to perception is being challenged. We cannot make statements about the world independent of our perception and cognition. We can only talk about the world we
live in, i.e., the world structured by our perceptual and cognitive mechanisms. Ref-
350
erence is concerned with the projected world which is a mental construct of our
mind/brain. Thus, the notion of world is explained as a mental phenomenon
(Jackendoff 1983, 2002; Schwarz 1996).
Reference research has also benefited a lot from the development of cognitive
science, an interdisciplinary approach to the study of mental faculties which has so
strongly been influencing recent research. According to cognitive science, the
mind is too complex to be seen clearly from the perspective of a single discipline.
Thus, the scientific understanding of cognition requires a synthesis of psychology,
linguistics, AI, philosophy, anthropology (and most recently neuroscience), the result of which is an increased interdisciplinary awareness (Schwarz 2008; for methodological reflections and applications Derry, Gernsbacher, and Schunn 2005).
Thus, interesting questions concern the nature of our mental knowledge, how we
represent this knowledge in memory, and how we put this knowledge to use. The
following working hypothesis holds true for cognitive science: The human mind
uses mental representations to store knowledge about the world. Human beings are
viewed as language processors that make use of this body of stored knowledge to
operate on temporary structures.
To sum up, the area of reference theory has radically changed. The main
achievement is that reference is no longer seen as a static relation, but as the result
of a complex mental process which involves all types of knowledge (essentially
perceptual, conceptual and linguistic knowledge).
A modern theory of pragmatic/cognitive reference must take into account a
series of connected topics which until recently have been dealt with in separate
fields: pragmatics, language understanding, perceptual processing, and conceptual
categorization. The rules and principles which underlie referential communication
cannot be explained adequately without the knowledge of how we store and access
both meanings and concepts in memory.2 Thus, cognitive reference theory deals
with three main questions: What is the knowledge basis of referentiality in long
term memory? How do we activate this knowledge in the process of referring?
How does this activation interact with more general procedures of cognition (perception and conceptual categorization)?
1.3.
Text, mind, and world: basic assumptions on referential representation

and underspecification
Current reference theory does not merely describe the function of referential expressions but attempts to explain the production and comprehension of referential
items as part of natural language communication in text and discourse grounded in
pragmatic and cognitive principles. To fully understand a text, a recipient must integrate information in one sentence with information in another. Explaining how
people do so is one of the main tasks of modern text linguistics. On the grammatical surface of the text, particular expressions (discourse markers) or grammatical
351
patterns serve as explicit cohesive ties to relate different parts of the text.3 The text,
as a consequence, is perceived as a coherent structure, rather than as an assembly
of disjointed referential information.
One of the central assumptions in modern text linguistics, however, is that the
explicit grammatical structure of the text is to be regarded as just one part or level
of textual coherence. Accordingly, recent developments in text research suggest
that the implicit conceptual relations in text are of vital importance and deserve
particular attention.4 At the same time, many researchers have realized that no
serious analysis of coherence can be carried out without appealing to its cognitive
aspects.5 Completing the propositional form by cognitive operations in order to get
to the underlying referential representation becomes necessary quite frequently
and often involves some kind of semantic-thematic decomposition as can be seen
in (2):
(2) She dug a hole into the frozen ground.
In order to get to the complete conceptual representation, some slot-filling or referent-recreating operation involving WITH AN INSTRUMENT, TYPICALLY A SHOVEL
has to be performed.6 In (3) it can be seen that these default elaborations play an
important role in establishing and accepting coherence:
(3) She opened the door reluctantly. (a) When she turned the key / (b) ?? When
she turned the spoon / (c) ?? When she turned the shovel
The continuation the key in (a) is accepted as coherent without any cognitive effort,
whereas in (b) and (c) the recipient would have to make an extra inference to establish some kind of (strange or unusual) continuity relation.
Referential underspecification is an essential feature of the relation between
verbal expressions and the conceptual representations they are intended to express.
Therefore, underspecification and implicitness in text should by no means be regarded as some kind of textual deviance.7 On the contrary, systematic underspecification is to be regarded as default. It is the result of applying the principle of relevance to the production process by leaving out superfluous information (see
Sperber and Wilson 1986/21995). In accordance with some principle of cognitive
economy, underspecification is achieved by selecting only the really relevant information and thereby avoiding redundancy (Schwarz 2000a: 83; Sanders and
Spooren 2001: 4). In order to elaborate the linguistically encoded meaning of a text
(which is based on the grammatical and lexical text structure), recipients automatically construct a mental text-world model; i.e. recipients enrich the text base by incorporating both information from the text and information activated through conceptual instantiation and inferential processing.
352
1.4.
Text-world models and reference construction
The text-world model represents a referential constellation of states-of-affairs in

working and in episodic memory which is (in most cases) more complex and elaborated than the semantic text basis, i.e. a propositional level derivable from the text
surface. Accordingly, we have to distinguish three levels: the text surface comprising grammatical structure and cohesive means, the semantic level, which is a level
of lexical meaning and compositionally derivable from the text, and the text-world
model, which is not a text level but a mental level of referential structures
(Schwarz 2001: 18).
From a procedural perspective, semantic enrichment and conceptual elaboration (as part of text-world model building) occur spontaneously and have to be
considered as part of our textual competence. The recipients elaborative capacity
is based on their memory structures and the ability to activate parts of these knowledge structures or to draw inferences. Hence, text comprehension implies both decoding the linguistically encoded meaning of the text and at the same time constructing a mental text-world model which is a cognitive conceptualization of the
world depicted in the text. The term text-world model refers to the assumption that
recipients create a mental model of the world described in a specific text and store it
in episodic memory (Schwarz 2001). The general idea behind this is that verbal expressions serve as mental processing instructions for the recipient (Givon 1992,
1995; Fauconnier 1994). As already pointed out, the text-world model of a text is
based on the information of the text itself but elaborated by the recipients cognitive
processing. It is the result of applying knowledge representations to the text base.
Building up a text-world model is an automatic process which always takes place in
the comprehension process.8 Once a text-world model has been developed during
reading, all subsequently encoded information will be integrated into the configurations of the model. These configurations or mental files, which are stored in working memory, are constantly being updated as the reading process progresses. For
instance, the conceptual file of the referent introduced in the first sentence of
(4) includes the information (YOUNG MAN IN C OLOGNE ) in short-term memory.
(4) Kln. Ein junger Mann wurde gestern Abend Opfer eines brutalen berfalls.
Als der 25jhrige Vizedirektor des Kaufparks in Nippes die Tr ffnete Joachim H. sah pltzlich eine Kanone vor seinen Augen Der Vater von 2 kleinen
Kindern (Newspaper article, Klner Stadtanzeiger, 11 March 2004)
Cologne. A young man was brutally assaulted yesterday evening. When the
25-year-old assistant director of the shopping center in Nippes opened the
door Joachim H. saw a gun right before his eyes The father of 2 little
kids
With each sentence new information is added to the conceptual knot (which functions as a file address in working memory), and the end result of the referential
353
movement is a file which includes the following information: (Y OUNG M AN

J OACHIM H. IS 25 YEARS OLD , HAS 2 LITTLE CHILDREN , WORKS IN C OLOGNE AS ASSISTANT DIRECTOR ).
Building up coherence and resolving anaphoric expressions in text comprehension are two processes which go hand in hand. Focussing on the role of the
text-world model and the leading principle of conceptual plausibility in understanding and constructing coherent representations of texts, the interface between
verbal structures and memory representations has been illustrated. Coherence has
been described as conceptual continuity, in the sense that the states-of-affairs expressed by parts of the text can be connected and integrated into the text-world
model automatically and without effort on the base of our linguistic and conceptual competence. The maintenance of continuity can best be described as a process
of mapping referential representations onto conceptual representations evoked by
prior text.
1.5.
Accessibility of referents
One of the most important aspects of explaining referential coherence in text involves the division between old (or given) and new information. Different cognitive accounts concerning the information distribution and thematic development
have been proposed in recent years.9 Old or given information corresponds to
entities assumed to be accessible in text, new information corresponds to entities
to be assumed inaccessible to recipients.
Considering the resolution of (definite) referential expressions which connect
sentences and serve as signals for topic continuity leads to the question of how recipients access and recover the referents of these expressions in text comprehension and what information has to be available in order to establish continuity. The
notion of accessibility is of particular interest here. This notion refers to the degree
of activation of information in long- or short-term memory (see Givon 2005 on
grounding). It is assumed that highly accessible mental entities (those which are
most active in consciousness) require less processing effort to be retrieved than do
entities of low accessibility.10 Ariel (1990) mentions four factors that affect accessibility: recency of mention (the more recent the last mention of an entity, the
more accessible it will be); saliency (either physical or textual salience); competition (salience of an entity compared to other entities of the same type that may also
be present in the context); unity (whether an antecedent is within the same paragraph or point of view as an anaphor). One of the main functions of referring expressions is to indicate the degree of accessibility of the mental representations of
their referents in a given context (Ariel 1990: 16; see von Heusinger 2000, 2003 for
formal aspects of accessibility theory).
A problem with these approaches arises from the one-to-one correlation that is
assumed to hold between grammatical categories of referring expressions and the
354
accessibility of referents (such as zero-pronouns indicating highly accessible referents, i.e. in case of anaphoric chains, versus demonstratives indicating the lowest
degree of accessibility, i.e. in case of deixis.) Firstly, such a matching seems to be
empirically inadequate (i.e. for demonstrative anaphors expressing the speakers
emotive attitude to the referent, see Averintseva-Klisch and Consten 2007). Secondly, a mere relationship between grammatical forms and referent status, regardless of language users perspectives, runs the risk of being a circular explanation
(Consten 2004; Consten and Loll 2009): from the speakers point of view, properties of referents in a given context seem to be a precondition for using certain referential means. Speakers choose grammatical features (such as definite vs. indefinite, demonstrative vs. non-demonstrative, lexical vs. pronominal NP) according
to the status of the referent in discourse (such as identifiability, first-mention vs. resumption). From the recipients point of view, grammatical features of the utterance evoke expectations about the referent in question. Grammatical features of
referential means are the base for mental operations of reactivating referents retrievable from memory or establishing new referents. Linking accessibility and givenness within the account of our text-world model proposal, referential movement in text comprehension can best be described as a sequence of the following
three transitional states: Activation, re-activation and deactivation.11
Activation of referents takes place when a new mental file (formally represented as a knot in a network) is opened. The referent is activated and stored as a
conceptual label. Hence, incoming information about the same referent may be
filed under this label. This process is to be characterized as re-activation, since the
already established knot in the text-world model is once again activated in working
memory. If a new referent is mentioned in the text, the referent in current focus is
deactivated while at the same time a file for the new referent is opened and stored
as an additional conceptual knot in the text-world model (Schwarz 2000b;
Schwarz-Friesel 2007a). In this regard, given means that the referent is mentally
accessible in text-world model, new means that this is not the case. The distributive flow of information in text comprehension, then, is to be described as an interplay of retaining continuity and progressive development.
Traditionally, definite NPs serving as anaphors are regarded as the main grammatical means to indicate continuity by referring to accessible, given entities. In
the following sections, however, we shall present a number of definite anaphoric
NPs whose referents are seen as accessible even though they would procedurally
be viewed as constituting new instead of given information.
2.
Anaphora and coreference
Anaphorical functions are closely related to the grammatical feature definiteness.

Definiteness is characterized as a grammatical instruction to the recipient to access
355
or construct the referents needed in order to establish a coherent representation.12

Definiteness itself does not spell out the precise connections that have to be constructed in the progress of resolving any individual NP. A unified theory of anaphora has to specify the conditions under which the various uses of anaphors can
be resolved and how the resolution depends on the meaning of the antecedents/anchors.13 In order to provide a full account of how several kinds of anaphors are processed, a cognitive theory of reference resolution must include both representational and procedural factors of cognitive processing and pragmatic elaboration.
2.1.
Topic continuity and antecedent-search-models
The traditional view in text linguistics has been that anaphors are used to continue
a pre-established reference in text by pointing back to a specific antecedent,
usually an NP. Thereby, they sustain the current focus of attention and at least
with longer anaphoric chains make their referent discourse topic of a text segment. Thus, the main anaphoric function is topic continuity. The anaphoric procedure is a linguistic instrument for having the hearer continue (sustain) a previously established focus (Ehlich 1982: 325).14 In this respect, anaphoric
elements tell the recipient where to search for the topic in the existing storage
structure of episodic memory (Givon 1992: 5). Hence, the classical and prototypical anaphoric case can be seen in example (5):
(5) A man (referential expression 11) entered the room. Obviously, he/this man
(referential expression 21) was drunk.
The referential relationship between anaphor and antecedent is based on coreference (indicated by the subscript1), and the link between the expressions denoting
coreference can be described in terms of grammatical congruence and semantic
compatibility. In this respect, anaphors have been described as retrieval cues or
echoes of their antecedents (Hintzman 1987: 412; Gernsbacher 1989: 100). According to this view, anaphora resolution is seen as some kind of search procedure
which aims at discovering the best matching antecedent in textual structure. This
mental checking in resolving anaphors in text comprehension always leads to the
conclusion that the referential expression 2 refers to the same extra linguistic item
as the referential expression 1. In on-line processing, the recipient must decide in
each case of reference whether there is referential identity or not. Considering (6),
one can conclude that the referent of R2 in (a) is identical with the referent of R1,
whereas in (b) this is not the case (which becomes evident to the recipient only
when the relative clause in R2 is processed).
(6a) Peter G. (R11) has beaten his wife. The man (R21) is quite unscrupulous.
(6b) Peter G.(R11) has beaten his wife. The man who lives across the street (R22)
observed this.
356
Anaphors which establish topic continuity by merely reactivating the information

knot of their antecedent involve pronouns, NPs with the same head noun, hyperonyms, synonyms and certain general expressions. The analysis of anaphors, however, calls for a more flexible description. Establishing the relationship of identity
(R EFERENT OF R2 IS IDENTICAL WITH REFERENT OF R1) which is necessary for successful anaphoric processing very often demands more than just checking semantic
compatibility. Reference chains in natural language texts reveal a lot more types of
anaphora which do not fall into the category echo of their antecedent. The establishment of coreference and continuity relies much more on both the context and
constructive mental activity of the recipient than on the grammatical devices in the
text. Hence, establishing coherence in text is best to be described as the interaction
of linguistic indicators and cognitive processes. The text itself gives only some information potential which offers readers certain (necessary) clues to the mental
construction of conceptual representations of what is communicated.
2.2.
Progressive anaphors
We are going to discuss a number of anaphors with explicit antecedents which not
only reactivate old files in text-world model constellations, but activate at the same
time new information about the referent in focus. It should be noted that these
types of anaphors occur quite frequently in natural language texts and are by no
means to be evaluated as some kind of textual deviance. To illustrate these progressive or moving anaphors, consider example (7) which depicts specifying
anaphors.15 The G RANDDAUGHTER -referent, who functions as the protagonist, once
it is introduced with a lexical NP, is referred to with personal- or zero pronouns,
whereas the G RANDMOTHER -referent, as the main theme of the story, is more
closely specified with lexical anaphors.
(7) Rentnerin erwrgt
Hamburg Eine Frau aus Hamburg-Bergedorf hatte ihre Gromutter seit drei
Tagen nicht gesehen und machte sich Sorgen. Als sie die gehbehinderte alte
Dame besuchen wollte, entdeckte sie die 81-Jhrige tot in ihrer Wohnung.
(beginning of a newspaper article, Express, 30 June 2006)
Pensioner [fem.]1 strangled
Hamburg A woman from Hamburg-Bergedorf2 had not met her2 grandmother1 for three days and 2 worried about her1. When she2 was going to visit
the physically handicapped lady1, she2 found the 81-year-old1 dead in her1
flat.
In cases like (7), the anaphoric expressions are much more than a mere semantic
echo of their antecedent. Several new information chunks are filed successively in
the same conceptual node. Hence, continued elaboration through activation processes of the current open file is the case. In on-line comprehension, specifying an-
357
aphors include both re-activation and activation in working memory. In cognitive

terms, their informational status is both given and new, since the text-world model
constellation changes (see Schwarz 2000b). Further, anaphors (with explicit
antecedents) may not only introduce new information about a referent, they may
even trigger new referents into the text-world model.
In accordance with our conception of coherence, felicity of anaphors depends
crucially on the plausibility criterion which is dependent both on long-term memory knowledge and on the text-world model at hand. In (8) we see a combining anaphor (Kombinationsanapher in German, Schwarz 2000b: 122) combing entities into a new referent with different ontological status.16
(8) Give the powder1 into a glass of water2 and drink down the dilution1+2.
Here, a new referent (DILUTION ) is mentally constructed on the basis of old information. Resolving this anaphor means activating a new file for a new referent in
the text-world model. If the lexical anaphor was replaced by a pronoun, the result
would be awkward since lexical meaning has to be provided in order to specify the
new referent.
2.3.
Complex anaphors: condensing and evaluating abstract referents
Referent-creating mechanisms are a major part in the interpretation of so-called

complex anaphors as well, as in (9).
(9) Unbestritten ist, da die Zahl der Arbeitspltze in der Industrie geringer wird,
whrend gleichzeitig das Gewicht des Dienstleistungssektors zunimmt.
a) Dieser Proze ist auch noch lngst nicht abgeschlossen.
(Newspaper article cit. from TiGer Korpus 181381813917; b) and c) are constructed versions)
b) Diese Tatsache prgt die europische Volkswirtschaft seit Mitte des 20. Jahrhunderts.
c) Diese Behauptung ignoriert allerdings Statistiken, die noch Raum fr Zweifel
lassen.
[It is indisputable that [the number of jobs is decreasing in the industrial sector
with the importance of the service sector growing at the same time.] process ] state
a) This process process has not finished yet.
b) This fact fact has shaped European economy since the middle of the 20th century.
c) This claim proposition, however, ignores statistics still leaving doubts.
In these examples, it is not an NP functioning as antecedent, but a whole clause.
There are heterogeneous terms for these cases, see Halliday and Hasan (1976) for
extended reference and reference to fact, Fraurud (1992) and Dahl and Hellmann (1995) for situational anaphora, Schmid (2000) for shell nouns and
358
Asher (1993) for abstract object anaphora, Schwarz (2000a) and Consten,
Knees, and Schwarz-Friesel (2007) for complex anaphora. Furthermore, the
phenomenon is discussed as discourse deixis (Webber 1991), see Lenz (1997)
and Consten (2004) for differentiation between anaphora and discourse deixis.
Referents of the complex anaphoric NPs are not individual entities (although
coded by an NP) but second-order or third-order entities (see Lyons 1989 for aspects of semantic ascent) like events, processes, states, facts, and propositions (in
(9) indicated by the indices p = process, s = state, f = fact). In (9), the referential
structure denoted by the preceding that-clause is picked up as a mere process (i.e.,
concerning its temporal structure, see (a)) or with regard to its truth-conditional
factual state (b), or the whole sentence including the main clause it is indisputable
is regarded as the antecedent (thus, the respective referent is the STATE OF BEING INDISPUTABLE ) (see (c)). This referent becomes subject to a kind of metalinguistic
statement and, as a result, is ontologically changed into an abstract object unbound
to a truth value. With (b) and (c), a higher level in the semantic hierarchy of the text
is produced. Functioning as a paraphrase of a preceding part in the text, a hypostatization takes place denoted by the anaphoric expression. On the basis of old information, a new referent is created in the dynamics of the text. Further, complex
anaphors do not only include a categorization (e.g. the incident, this assumption)
but often evaluate the referent by their lexical content (e.g. this tragedy, this outrageous imputation), see Schwarz-Friesel (2007b: 215216, 2008: 199201).
Hence, two main functions can be made out: Complex anaphors give a classification (in the sense (X Y); THIS EVENT IS A DISASTER ) and a new conceptual file
node (DISASTER ) is established in the text-world model. In cognitive terms of the
given-new-distinction, both re-activation and activation take place at the same
time (for further discussion of complex anaphors see Consten, Knees, and
Schwarz-Friesel 2007, 2009; Consten and Knees 2008).
So far, we have seen that in many cases the processing of anaphoric devices is
not only constrained by their lexical content, but by the activation of conceptual
knowledge representations and the performance of referent-creating operations.
Besides, the examples discussed show that the resolution of many direct anaphors
involves much more than a simple search-and-match-procedure. Their full interpretation requires a more complex cognitive process. Resolving anaphors may involve specification, combination, classification, and/or evaluation mechanisms.
From a procedural perspective, many anaphors combine given and new information for the recipient and therefore can be characterized as given-and-newentities in the text.
2.4.
359
Reference and inference: indirect anaphora referring to mental

representations
Definite NPs may introduce new referents into the text-world model and at the
same time signal topic continuity, which can be most clearly illustrated with the
phenomenon of indirect anaphora which by no means is a marginal use of definite
NPs. A study by Fraurud (1990, 1992) has shown that in a corpus of written Swedish, over 60 % of all definite NPs occur as first mentions of a referent but are, in
fact, necessarily related to the preceding text (see Schwarz 2000a and Consten
2004 for a detailed account of indirect anaphora as a form of domain-bound reference). The passage in (10) is perfectly coherent although there is no textual
antecedent for the definite NPs the rails, the conductor and the engine.
(10) We arrived late since our train stood still at the rails for one hour. The conductor said, There are problems with the engine. (constructed example)
Indirect anaphora fall under the heading of anaphora because the interpretation of
the definite NPs depends on the existence of an adequate textual element in the
preceding text adopting the terminology of Fraurud (1990, 1992), we will call
such elements anchors and the process of linking indirect anaphors to these expressions anchoring18 (see (10) where train functions as anchor). Further, indirect anaphors establish referential continuity just like direct anaphora.
In text comprehension, the resolution of indirect anaphors confronts the cognitive system of the recipient with the problem of identifying and localizing the referent to which the definite description refers in the text. The definite article signals
accessibility of the referent, but in the text structure no such referent can be found.
Thus, we are concerned with the accessibility of implicit referents in the textworld. The prevailing assumption is that there is an associative relationship between trigger and anaphor which helps to establish some link between the two expressions. Thus, these referential instances are often called associative anaphors
(e.g. Hawkins 1978; Heim 1991). However, since only a small number of indirect
reference relations can be explained by the activation of associations, we prefer the
more general term indirect anaphor (see Erk and Gundel 1987).19 According to
most theories, the relation between anchor and indirect anaphor can best be explained in terms of associability, and the anchoring always involves some kind of
inferential linking or bridging (see Clark 1977; Fraurud 1992; Garrod and Sanford
1994; Cornish 1999).
According to another popular view, indirect anaphors are explained as frameevoked entities in implicit focus. In this respect, some authors distinguish direct
and indirect anaphors with the help of explicit and implicit focus.20 Explicit focus
is the current focus of attention, the highly activated part of workspace where incoming information is temporarily held, that is short-term memory which contains
representations of active referents explicitly introduced. Implicit focus is the rep-
360
resentation of entities evoked implicitly as they form part of a frame retrieved from
memory. Whereas the referents of direct anaphora are accessible in explicit focus,
the referents of indirect anaphora have to be accessible in implicit focus, that is as
default values of a frame or script.
However, not all indirect anaphors in a text refer to particular parts of a frame.
Besides, we have seen that the resolution of many direct anaphora demands more
than a search-and-match in short-term memory. Here, too, mental configurations
which exceed the explicit information in the text are built up (by the reader).
Therefore, we will argue for a more complex cognitive approach. Taking a procedural perspective, we want to demonstrate that there are different types of indirect anaphora. A distinction will be drawn between indirect anaphora which are
based on the activation of default knowledge in long-term memory and indirect anaphora which require a more refined cognitive process for their resolution. Further,
we are going to explain why all kinds of indirect anaphora have to be seen as
given-and-new-entities.
Different types of indirect anaphora and their resolution can be gradually distinguished. The criteria established for distinguishing these types are based on the
linguistic expressions used as anchor and on the different knowledge structures involved in interpreting the mental relationship between anchor and indirect anaphor. These conditions of anchoring do not sufficiently cover the whole range of
the complex phenomenon but show some fundamental procedures of interpreting
referring expressions which pragmatically depend on the preceding text. The
examples (which are taken from natural language texts in German) concern the
most frequent and prototypical types of textual indirect anaphora.
(11) Zu den Neuheiten gehrt eine elektronische Promille-Wegfahrsperre fr Berufskraftfahrer []. Mit der Zndung eines Autos verbunden, gibt die Barriere den Motor erst dann frei, wenn der Fahrer zuvor einen Alkoholtest per
Atemprobe abgegeben hat. (Newspaper article cit. from TiGer Korpus,
611561116, modified)
One of the novelties is an electronic blood alcohol immobilizer for professional drivers. Connected with the ignition circuit of a car, the barrier deblocks the engine only when the driver has absolved a field sobriety test by
breath check.
This type of indirect anaphor is established by meronymic relation between the anchor a car and the indirect anaphors the engine and, in a broader sense of meronymy, the driver. There is a part-whole-relationship between the referents ENGINE
and CAR, and an affiliation-relationship between the DRIVER-referent and the CARreferent.
Example (12) shows an indirect anaphor based on the thematic roles of the verb
park in the preceding sentence: the thematic structure of parking involves an agent
(that is addressed directly by you) and an object that is parked. Unlike (11), the
361
close anchoring allows for an pronominal indirect anaphor referring to the expected CAR-referent.
(12) Ich wre wunschlos glcklich, wenn Sie nicht immer auf dem Lehrerparkplatz parken wrden. Das nchste mal lasse ich ihn abschleppen. (Beverly
Hills 90210, German version of the TV-soap, cit. from Consten 2004: 93)
Id be completely happy if you didnt always park on the teachers parking
lot. Next time I will have it towed away.
In example (13), a script-based type is seen. The anchoring of the indirect anaphor
is based on the activation of the appropriate default values in long-term memory.
(13) In jenen letzten Augusttagen des Jahres 1961 erwacht in einem kleinen Krankenhauszimmer das Mdchen Rita Seidel Die Krankenschwester tritt an
das Bett als morgens der Arzt (Christa Wolf, 1973, Der geteilte Himmel, Mnchen: Deutscher Taschenbuchverlag. p.9).
In those last days of August 1961, a girl named Rita Seidel awakens in a hospital room. The nurse comes to the bed the doctor
Establishing a coherent link between the sentences in (13) is to treat the nurse and
the doctor as fulfilling typical roles or default values of the script H OSPITAL in
long-term memory which is activated in a top-down-process by the anchor-expression hospital room. Thus, resolving indirect anaphora involves an elaboration
which extends the explicit representation of the text. Comprehension depends on
the extension of the current representation, taking into account more than textual
information.
It is a common view that interpretation of all indirect anaphors is based on
some kind of inference or linking. This claim, however, neglects the difference between indirect anaphors which are based on a cognitive process involving the activation of semantic roles or default values and indirect anaphors which are truly
based on an inference process. The term inference should not be used for all kinds
of memory processes involved in text comprehension in the sense of any piece of
information that is not explicitly stated in the text (McKoon and Ratcliff 1992:
440), but only for problem-solving activities, that is the constructive use of world
knowledge. Otherwise, it is impossible to describe cognitive processes more precisely and in explicit detail. Lexical decomposition, slot-filling and spreading activation represent types of semantic activation processes which should not be
equated with using knowledge in inferential problem-solving activities. An indirect anaphor which is resolved not merely by activating a semantic role or default
value, but by drawing an inference is exemplified in (14), which is a quite macabre
example:
362
(14) Das Urteil ist somit rechtskrftig und wurde bereits vollstreckt. Die Urne
kann im Krematorium Dresden in Empfang genommen werden.
(From a letter the East German secret service addressed to a detainees wife
who had enquired about her missed husband. Published at a historical exhibition at Erfurt, 2000, cit. from Consten 2004: 98)
Thus, the judgement is non-appealable and has been enforced already. You
can collect the funeral urn at the Dresden crematorium.
The recipient did not yet know that her husband had been sentenced to death, and
this fact is not mentioned in the preceding text of (14) either. Thus, she had to draw
the inference THE ADJUDGEMENT WAS A DEATH SENTENCE in order to integrate the
definite NP the funeral urn coherently. The process of inferential anchoring does
not necessarily involve the activation of one specific frame or script. However, a
lot of inferences are based on script knowledge (see Schwarz 2008: 198f). Besides,
inferential processing does occur within the configuration (and constraints) of the
text-world model, that is, referents have to fulfill requirements of conceptual
plausibility.
The examples so far discussed show that within the group of indirect anaphors
different types can be made out which rely both on different knowledge structures
and processing operations for their anchoring (Schwarz 2000a and 2007a; Consten
2004). Indirect anaphors which rely on the activation of semantic or conceptual
knowledge are resolved within the immediate mental scope of the anchor expression comprising the prototypical information of the cognitive domain. Resolving inferable indirect anaphors involves the expansion of this scope. In general, anchoring conditions of referents for indirect anaphors can be formulated in the
following way: The referent of an indirect anaphor must be either an identifiable
part of the implicit semantic structure of the preceding sentences, or it must be a
default value of a specific frame or script activated by the anchor expression, or it
must be inferable on the basis of cognitive plausibility determined by prototypical
world knowledge.
The definite NP denoting an indirect anaphor refers to an entity being introduced into the text for the very first time (for the function of definiteness, see the
beginning of 2.). In this respect, indirect anaphors carry new information into the
text-world model. Even more, a new node for the new referent has to be established
in an activation process in working memory. On the other hand, at the semantic
level of the text the definite article signals that (given) information is easily accessible and, thus, demands a re-activation process (for indefinite indirect anaphors,
see Schwarz 2000a: 67). Seen as grammatical means with processing instructions,
indirect anaphors signal accessibility of the discourse referent while a new referent
is introduced to the text-world model.
From a given-new-perspective, thus, indirect anaphors combine both referential continuity and elaborative development in information processing. An acti-
363
vation process is taking place by establishing a new file in short-term memory,

while at the same time we have re-activation in long-term memory. More precisely,
the cognitive domain of the anchor expression has been in the status of semi-activation (for aspects of activation levels and the status of semi-activation in longterm memory representations of the anchors, see Schwarz 2000a: 136139).
Hence, re-activation of the semi-activated domain accounts for the ease of processing and the acceptability of indirect anaphors as continuity markers. In this respect,
indirect anaphors serve as means to establish coherence in the same way as direct
anaphors (Schwarz-Friesel 2007a).
To sum up: The anchoring of indirect anaphors is neither solely determined by
the principle of associability nor by the activation of a script. It turns out that both
associability and inferrability are restricted by the conceptual constraint of plausibility. A strong semantic link between anchor and indirect anaphor does not per se
guarantee proper and successful anchoring. Rather, the plausibility within the specific text-world model governs the anchoring process. The immediate resolution of
indirect anaphors comes about through an interaction between the information in
the anchor and the anaphor, the mental accessibility of the implied referent within a
specific context and the focus state of the prior and current textual representation.
From a perspective of referential movement, indirect anaphors have to be regarded
as given-and-new-entities. Combining both easy accessibility and the establishment of new nodes in text-world models, they serve as progressive continuity
markers.
To establish coherence, a mental representation for the textual referent must be
either available in the conceptual text representation or inferable from memory at
the time the anaphoric item is processed. Given this similarity between direct and
indirect anaphors, the interpretation of direct and indirect anaphors seems to obey
similar rules (see Schwarz 2000a; Consten 2003, 2004; Schwarz-Friesel 2007a). In
both cases, the semantics of these expressions only provides a minimal framework,
which needs to be complemented by different cognitive mechanisms at the conceptual level of the text-world model. In this respect, explaining anaphora means at the
same time finding an explanation for the interaction between language structures
and cognitive processes. Thus, anaphoric reference exemplifies prototypically the
pragmatic dimension of reference.
Notes
1. DRT; fundamentals by Kamp (1981) and Heim (1988); for an important application to the
theory of anaphora see Asher 1993.
2. In some theories of cognitive semantics a distinction is drawn between concepts (i.e.
mental representations that store all kinds of knowledge about entities of the world) and
meanings (i.e. concepts bound to linguistic forms). There is strong empirical evidence
from aphasia and lg. acquisition for such a distinction (see Schwarz 2008: 5974).
364
3. See Andersen and Fretheim (2000); Ariel (1998); Fraser (1999); Lenk (1998a and
1998b); Spooren and Risselada (1998).
4. See Gernsbacher and Givon (1995); Bublitz (1999); Spooren (1999); Sanders and
Spooren (2001).
5. See Hobbs (1979); Rickheit and Habel (1995); Sanford and Moxey (1995); Cornish
(1999); Schwarz (2000a); Zelinsky-Wibbelt (2003).
6. See Singer (1994); van den Broek (1994) for a detailed discussion of the procedural aspects of these elaborations; Recanati (2004) for problems that the necessity of conceptual elaborations entails for a notion of literal meaning.
7. This is implied in many accounts, i.e. with respect to underspecified anaphors Reinhart
(1980); Prince (1981); Heim (1988); Ariel (1990); see Schwarz (2000a: 78).
8. For slightly different conceptions of text-worlds or mental models, see de Beaugrande
and Dressler (1981); Van Dijk and Kintsch (1983); Johnson-Laird (1983); Garnham and
Oakhill (1992); Givon (1995); Cornish (1999).
9. See Prince (1981); Grosz (1981); Givon (1992); Gordon, Grosz, and Gilliom (1993);
Chafe (1994); Lambrecht (1994); Schwarz (2000a, b); Hasselgrd et al. (2002) for language specific phenomena at the grammar-discourse-interface.
10. See Gundel et al. (1993); Ariel (1988); Walker, Joshi, and Prince (1998); Byron, BrownSchmidt, and Tanenhaus (2008).
11. See Givon (1992) and Chafe (1994); similar states have been postulated by centering
theory, i.e. continuing, retaining and shifting; see Grosz (1981); Gordon, Grosz, and
Gilliom (1993); Maes (1997); Walker, Joshi, and Prince (1998); Kibble and Power
(1999).
12. For a survey of definite reference, see Consten (2004), for critical remarks on the interpretation of definite determiners, see Consten and Loll (2009).
13. See 2.4 for our notion of anchor.
14. See also Clark and Haviland 1977; Murphy 1985; Carter 1987; Bosch 1988; Givon
1992.
15. They are also called labelling anaphors, Francis (1986: 37); Spezifikationsanaphern
in German, see Schwarz (2000a: 6264.)
16. For similar examples, see Brown and Yule (1983); Stenning (1992); Freitas and Lopes
(1996).
17. The TiGer Korpus is a German newspaper corpus released in 2003, see
www.ims.uni-stuttgart.de/projekte/TIGER. The numbers indicate the sentence numbers
in the original corpus.
18. Hawkins (1978) calls them trigger, Cornish (1999) antecedent trigger.
19. Indirect anaphora has also been called bridging reference or inferrables. We prefer
the term indirect anaphora since it includes all types of anaphora that occur as firstmention definites in text and need some kind of anchoring but do not necessarily depend
on a bridging inference or rely on an associative relation; see Schwarz (2000: 4950).
20. Sanford and Garrod (1981, 1994a, b); Garrod (1995), for critical remarks, see Cornish
(2005) and Cornish et al. (2005).
365
References
Andersen, Gisle and Thorstein Fretheim (eds.)
2000
Ariel, Mira
1988
Referring and accessibility. Journal of Linguistics 24: 6587.
Ariel, Mira
1990
Accessing noun-phrase antecedents. London: Routledge.
Ariel, Mira
1998
Discourse Markers and Form-Function Correlations. In: Andreas H. Jucker
and Yael Ziv (eds.), Discourse Markers: Descriptions and Theory, 223259.
Ariel, Mira
2001
Accessibility theory: an overview. In: Ted Sanders, Joost Schilperoord and
Wilbert Spooren (eds.), Text Representation: linguistic and psycholinguistic
processes, 2987. Amsterdam: Benjamins.
Asher, Nicholas
1993
Reference to Abstract Objects in Discourse. Dordrecht: Kluver.
Averintseva-Klisch, Maria and Manfred Consten
2007
The role of discourse topic and proximity for demonstratives in German and
Russian. In: Bergljot Behrens, Cathrine Fabricius-Hansen, Hilde Hasselgrd
and Stig Johansson (eds.), Language in Contrast Volume 7/Issue 2, 219240.
Beaugrande, Robert-Alain de and Wolfgang U. Dressler
1981
Einfhrung in die Textlinguistik. Tbingen: Niemeyer. English translation: Robert-Alain de Beaugrande and Wolfgang U. Dressler
1994
Introduction to Text Linguistics. London: Longman.
Bosch, Peter
1988
Representing and accessing focussed referents. Language and Cognitive Processes 3: 207231.
Bransford, John D., Richard R. Barclay and Jefferey J.Franks
1972
Sentence memory: a constructive versus interpretive approach. Cognitive Psychology 3: 193209
Broek, Paul van den
1994
Comprehension and memory of narrative texts. In: Morton Ann Gernsbacher
(ed.), Handbook of Psycholinguistics, 539588. San Diego: Academic Press.
1983
Discourse analysis. Cambridge: Cambridge University Press.
Bublitz, Wolfram
1999
Introduction: Views of Coherence. In: Wolfram Bublitz, Uta Lenk and Eija
Ventola (eds.), Coherence in Spoken and Written Discourse, 17. Amsterdam:
Benjamins.
Byron, Donna K., Sarah Brown-Schmidt and Michael K. Tanenhaus
2008
The overlapping distribution of personal and demonstrative pronouns. In:
Gundel, Jeanette K. and Nancy Hedberg, (eds.), Reference Interdisciplinary
Perspectives, 143175. Oxford: Oxford University Press.
366
Caramazza, Alfonso and Bradford Mahon

2006
The organization of conceptual knowledge in the brain: the futures past and
some future directions. Cognitive Neuropsychology 23: 1338.
Carter, David M.
1987
Interpreting Anaphors in Natural Language Texts. Chichester: Ellis Horwood.
Chafe, Wallace L.
1994
Discourse, consciousness, and time. The flow and displacement of conscious
experience in speaking and writing. Chicago: Chicago University Press.
Clark, Herbert H.
1977
Bridging. In: Philip Nicholas Johnson-Laird and Peter C. Wason (eds.), Thinking. Readings in cognitive science, 411420. Cambridge: Cambridge University Press.
Clark, Herbert H. and Susan E. Haviland
1977
Comprehension and the given-new contract. In: Roy O. Freedle (ed.), Discourse
production and comprehension, 140. Hillsdale, NJ: Lawrence Erlbaum.
1981
Webber and Ivan A. Sag (eds.), Elements of discourse understanding, 1063.
New York: Cambridge University Press.
1983
Common ground and the understanding of definite reference. Journal of Verbal
Learning and Verbal Behavior 22: 245258.
Clark, Herbert H. and C. J. Sengul
1979
In search of referents for nouns and pronouns. Memory and Cognition 7:
3541.
Consten, Manfred
2003
Towards a unified model of domain-bound reference. In: Friedrich Lenz (ed.),
Deictic Conceptualisation of Space, Time and Person, 223248. Amsterdam:
Benjamins.
Consten, Manfred
2004
Anaphorisch oder deiktisch? Zu einem integrativen Modell domnengebundener Referenz. Tbingen: Niemeyer.
Consten, Manfred and Mareile Knees
2008
Complex Anaphors in Discourse. In: Anton Benz and Peter Khnlein (eds.),
Constraints in Discourse, 181199. Amsterdam: Benjamins.
2007
The Function of Complex Anaphors in Texts. In: Monika Schwarz-Friesel, Manfred Consten and Mareile Knees (eds.), Anaphors in Texts, 81102. Amsterdam: Benjamins.
2009
Complex Anaphors. The Impact of Ontology, Cotext and Conceptual Knowledge. In: Jordan Zlatev, Marlene Johansson Falck, Carita Lundmark and Mats
Andrn (eds.), Studies in Language and Cognition, 285302. Newcastle: Cambridge Scholars Publishing.
Consten, Manfred and Annegret Loll
2009
Indirekte Anaphern ein Zirkularitts-problem zwischen Grammatik und
Pragmatik. In: Attila Peteri (ed.), An der Grenze zwischen Grammatik und
Pragmatik, 325332. Frankfurt am Main: Peter Lang.
367
Cornish, Francis
1999
Anaphora, Discourse, and Understanding. Oxford: Clarendon.
Cornish, Francis
2005
Degrees of indirectness: two types of implicit referents and their retrieval via
unaccented pronouns. In: Antnio Branco, Tony McEnery and Ruslan Mitkov
(eds.), Anaphora Processing: Linguistic, cognitive and computational modelling, 199220. Amsterdam: Benjamins.
Cornish, Francis, Alan Garnham, H. Wind Cowles, Marion Fossard and Virginie Andr
2005
Indirect anaphora in English and French: A cross-linguistic study of pronoun
resolution. Journal of Memory and Language 52 (3): 363376.
Dahl, sten and Christina Hellmann
1995
What happens when we use an anaphor? Paper. Stockholm: University, Department of Linguistics.
Damasio, Hanna, Thomas J. Grabowski, Daniel Tranel, Richard D. Hichwa and Antonio R.
Damasio
1996
A neural basis for lexical retrieval. Nature 380: 499505.
Derry, Sharon J., Morton Ann Gernsbacher and Christian D. Schunn (eds.)
2005
Interdisciplinary Collaboration An emerging cognitive science. London:
Routledge.
Donnellan, Keith S.
1966
Donnellan, Keith S.
1978
Speaker reference, descriptions and anaphora. In: Peter Cole (ed.), Pragmatics, 4768. New York: Academic Press.
Dowty, David R.
1979
Word Meaning and Montague Grammar: the Semantics of Verbs and Times in
Generative Semantics and in Montagues PTQ. Dordrecht: Reidel.
Ehlich, Konrad
1982
Anaphora and deixis: Same, similar or different? In: Robert J. Jarvella and
Wolfgang Klein (eds.), Speech, Place and Action: Studies in Deixis and Related Topics, 315339. Chichester: Wiley.
Erk, Feride and Jeanette K. Gundel
1987
The pragmatics of indirect anaphors. In Jef Verschueren and Marcella Bertucelli-Papi (eds.), The pragmatic perspective, 533545. Amsterdam: Benjamins.
Fauconnier, Gilles
1994
Mental spaces: Aspects of meaning construction in natural language. 2nd ed.
Francis, Gill
1986
Anaphoric Nouns. Birmingham: University of Birmingham.
Fraser, Bruce
1999
What are Discourse Markers? Journal of Pragmatics 31: 931952.
Fraurud, Kari
1992
Processing noun phrases in natural discourse. Stockholm: University, Department of Linguistics.
Fraurud, Kari
1990
Definiteness and the processing of NPs in natural discourse. Journal of Semantics 7(4): 395434.
368
Frege, Gottlob
1892
ber Sinn und Bedeutung. Zeitschrift fr Philosophie und philosophische Kritik C: 2550.
Freitas, Sergio and Jos G. Lopes
1996
Solving the reference to mixable entities. Paper read at the Workshop Indirect
Anaphors in Lancaster.
Garnham, Alan and Jane Oakhill (eds.)
1992
Discourse representation and text processing. A special issue of Language and
Cognitiv processes. Hove: Lawrence Erlbaum.
Garrod, Simon
1995
Distinguishing between Explicit and Implicit Focus during Text Comprehension. In: Gert Rickheit and Christopher Habel (eds.), Focus and Coherence in
Discourse Processing, 317. Berlin/New York: de Gruyter.
Garrod, Simon C. and Anthony. J. Sanford
1994
Resolving sentences in a discourse context: How discourse representation affects language understanding. In: Morton Ann Gernsbacher (ed.), Handbook of
psycholinguistics, 675698. San Diego: Academic Press.
Gernsbacher, Morton Ann
1989
Mechanisms that improve referential access. Cognition 32: 99156.
Gernsbacher, Morton Ann and Talmy Givon (eds.)
1995
Coherence in spontaneous text. Amsterdam: Benjamins.
Givon, Talmy
1992
The Grammar of Referential Coherence as Mental Processing Instructions.
Givon, Talmy
1995
Coherence in text vs. coherence in mind. In: Morton Ann Gernsbacher and
Talmy Givon (eds.), Coherence in Spontaneous Text, 59115. Amsterdam:
Benjamins.
Givon, Talmy
2005
Context as Other Minds. The Pragmatics of Sociality, Cognition and Communication. Amsterdam: Benjamins.
Gordon, Peter C., Barbara J. Grosz and Laura A. Gilliom
1993
Pronouns, names, and the centering of attention in Discourse. Cognitive
Science 17: 311347.
Grosz, Barbara
1981
Focusing and description in natural language dialogues. In: Aravind K. Joshi,
Bonnie L. Webber and Ivan A. Sag (eds.), Elements of discourse understanding, 84105. Cambridge: Cambridge University Press.
Gundel, Jeanette K., Nancy Hedberg and Ron Zacharski
1993
Cognitive Status and the Form of Referring Expressions in Discourse. Language 69(2): 274307.
1976
Hasselgrd, Hilde, Stig Johansson, Bergljot Behrens, Cathrine Fabricius-Hansen (eds.)
2002
Information Structure in a Cross-linguistic Perspective. Amsterdam: Rodopi.
Hawkins, John
1978
Definiteness and Indefiniteness: A Study in Reference and Grammaticality Prediction. London: Groom Helm.
369
Heim, Irene
1991
Artikel und Definitheit. In: Arnim von Stechow and Dieter Wunderlich (eds.),
Semantik. Ein internationales Handbuch der zeitgenssischen Forschung,
487535. Berlin/New York: de Gruyter.
Heim, Irene
1988
The semantics of definite and indefinite noun phrases. New York: Garland.
2000
Accessibility, discourse anaphora, and descriptive content. In: Paul Baker
(ed.), UCREL Technical Papers Vol. 12: Special Issue Proceedings of the Third
International Conference on Discourse Anaphora and Anaphor Resolution
(DAARC2000. 173180. Lancaster: Lancaster University).
2003
The double dynamics of definite descriptions. In: Jaroslav Peregrin, (ed.),
Meaning in the Dynamic Turn, 150168. Amsterdam: Elsevier.
Hintzman, Douglas L.
1987
Schema abstraction in a multiple-trace model. Psychological Review 93:
411428.
Hobbs, Jerry R.
1979
Coherence and coreference. Cognitive Science 3: 6790.
Jackendoff, Ray
1983
Semantics and Cognition. Cambridge, MA: MIT-Press.
Jackendoff, Ray
2002
Foundations of language. Oxford: Oxford University Press.
Johnson-Laird, Philip Nicholas
1983
Mental Models. Towards a Cognitive Science of Language, Inference, and
Consciousness. Cambridge, MA: Harvard University Press.
Kamp, Hans
1981
A theory of truth and semantic representation. In: Jeroen Groenendijk, Theo
Janssen and Martin Stokkoff, M. (eds.), Formal methods in the Study of language, 277322. Amsterdam: Mathematical Theory.
Kibble, Rodger and Richard Power
1999
Using centering theory to plan coherent texts. Proceedings of the 12th Amsterdam
Colloquium, 187192. ftp://ftp.itri.bton.ac.uk/pub/reports/ITRI-9919.ps.gz
Lambrecht, Knud
1994
Information structure and sentence form: topic, focus and the mental representation of discourse referents. Cambridge: Cambridge University Press.
Lenk, Uta
1998a. Marking Discourse Coherence. Functions of Discourse Markers in Spoken English. Tbingen: Narr.
Lenk, Uta
1998b
Discourse Markers and Global Coherence in Conversation. Journal of Pragmatics 30: 245257.
Lenz, Friedrich
1997
Diskursdeixis im Englischen. Tbingen: Niemeyer.
Lyons, John
1989
Semantic Ascent: A Neglected Aspect of Syntactic Typology. In: Doug Arnold, Claire Grover and Louisa Sadler (eds.), Essays on grammatical theory
and universal grammer, 153186. Oxford: Clarendon Press.
370
Maes, Alfons
1997
Referent ontology and Centering in Discourse. Journal of Semantics, 14:
207235.
Mahon, Bradford and Alfonso Caramazza
2009
Concepts and Categories: A Cognitive Neuropsychological Perspective. Annual Review of Psychology 60: 2751.
Marslen-Wilson, William and Lorraine Komisarjevsky Tyler
1982
The resolution of discourse-anaphors. Text 2: 263291.
McDonald, Janet L. and Brian MacWhinney
1995
The time course of pronoun resolution: Effects of implicit verb causality and
gender. Journal of Memory and Language 34: 543566.
1992
Inference during meaning. Psychological review 99: 440466.
Murphy, Gregory L.
1985
Processes of Understanding Anaphora. Journal of Memory and Language 24:
290303.
Prince, Ellen F.
1981
Toward a taxonomy of Given-New-information. In: Peter Cole (ed.), Radical
Recanati, Franois
2004
Reinhart, Tanya
1980
Conditions of coherence. Poetics Today 1: 161180.
Rickheit, Gert and Christopher Habel (eds.)
1995
Focus and Coherence in Discourse Processing. Berlin/New York: de Gruyter.
Russell, Bertrand
1905
On Denoting. Mind 30: 479493.
Sanders, Ted and Wilbert Spooren
2001
Text representation as an interface between language and its users. In: Ted
Sanders, Joost Schiperoord and Wilbert Spooren (eds.), Text Representation:
Linguistic and psycholinguistic aspects, 125. Amsterdam: Benjamins.
1994a
Resolving sentences in a discourse context. In: M. A. Gernsbacher (ed.),
Handbook of Psycholinguistics, 675698. San Diego: Academic Press.
1994b
Selective processing in text understanding. In: M. A. Gernsbacher (ed.), Handbook of Psycholinguistics, 699719. San Diego: Academic Press.
1981
Understanding Written Language: Explorations in Comprehension Beyond the
Sentence. Chichester: Wiley.
Sanford, Anthony J. and Linda M. Moxey
1995
Aspects of coherence in written language: A psychological perspective. In:
Morton A. Gernsbacher and Talmy Givon (eds.), Coherence in Spontaneous
Text, 161187. Amsterdam: Benjamins.
Schmid, Hans-Jrg
2000
English abstract nouns as conceptual shells. From corpus to cognition. Berlin:
Mouton de Gruyter.
371
Schwarz, Monika
1996
Reference as Process. In: Richard A. Geiger (ed.), Reference in a Multidisciplinary Perspective, 461470. Hildesheim: Olms.
Schwarz, Monika
2000a
Indirekte Anaphern in Texten. Studien zur domnengebundenen Referenz und
Kohrenz im Deutschen. Tbingen: Niemeyer.
Schwarz, Monika
2000b
Textuelle Progression durch Anaphern Aspekte einer prozeduralen ThemaRhema-Analyse. Linguistische Arbeitsberichte 74: 111126.
Schwarz, Monika
2001
Establishing Coherence in Text. Conceptual Continuity Within the Domain of
Text-world Models. In: Logos and Language 1523.
Schwarz, Monika
2008
Einfhrung in die Kognitive Linguistik. 3. Aufl.Tbingen: Francke.
2007a
Indirect anaphora in text: a cognitive account. In: Monika Schwarz-Friesel,
Manfred Consten and Mareile Knees (eds.), Anaphors in Text, 320. Amsterdam: Benjamins.
2007b
Sprache und Emotion. Tbingen: Francke.
Schwarz-Friesel, Monika, Manfred Consten and Mareile Knees (eds.) 2007 Anaphors in
Text. Amsterdam: Benjamins.
Searle, John R.
1969
Singer, Murray
1994
Discourse inference processes. In: Morton Ann Gernsbacher (ed.), Handbook
of Psycholinguistics, 479515. San Diego: Academic Press.
1995 [1986]Relevance: Communication and Cognition.2nd ed. Oxford: Basil Blackwell.
Spooren, Wilbert
1999
Textlinguistik. In: Ralf Prings and Ulrich Schmitz (eds.). Sprache und
Sprachwissenschaft, 187210. Narr: Tbingen.
Spooren, Wilbert and Rodie Risselada
1998
Introduction: Discourse Markers and Coherence Relations. Journal of Pragmatics 30: 131133.
Stenning, Keith D.
1978
Anaphora as an approach to pragmatics. In: Morris Halle, Joan Bresnan, and
George A. Miller (eds.), Linguistic Theory and Psychological Reality, 162
Stenning, Keith D.
1992
Representations of individuals and the processing of reference change. Quaterly Journal of Experimental Psychology. Human Experimental Psychology
45A 3: 373398.
Strawson, Peter F.
1950
On Referring. Mind 59: 320344.
Thrane, Torben
1980
Referential-Semantic Analysis. Cambridge: Cambridge University Press.
372
Van Dijk, Teun A. and Walter Kintsch

1983
Strategies of discourse comprehension. New York: Academic Press.
Walker, Marilyn A., Aravind K. Joshi and Ellen F. Prince (eds.)
1998
Centering Theory in Discourse. Oxford: Clarendon Press.
Webber, Bonnie L.
1991
Structure and ostension in the interpretation of discourse deixis. Language and
Cognitive Processes 6: 107135.
Zelinsky-Wibbelt, Cornelia (ed.)
2003
Text, context, concepts. Berlin: de Gruyter.
13.
Speech acts
Elena Collavin
1.
Introduction
Speech act theory originated from the works of the Oxford philosopher of ordinary
language John Langshaw Austin (19111960). Austin first presented the main tenets of his theory in the lectures he gave at Oxford in the years 19521954 under
the title Words and deeds, and subsequently in the William James Lectures he
delivered at Harvard University in 1955. However, Austin pointed out that those
ideas were formed as early as 1939 (1962:iii). After Austins premature death the
notes of his William James Lectures were edited, supplemented with earlier notes
and tape-recorded lectures, and published in 1962 under the title How to do
Things with Words (Austin 1962). In 1969 one of Austins pupils, the American
philosopher John R. Searle, published his own version of the theory under the title
Speech acts An Essay in the Philosophy of Language, followed by several
more works dedicated to speech acts. Searles systematization and development of
Austins ideas has been very influential, to the point that Searles interpretation of
the theory is at times taken as the definitive view of speech acts.
Two main ideas are at the core of speech act theory: the first holds that the
meaning of an utterance is distinct from the function that the utterance performs
(what we shall call the force of an utterance); the second is that all utterances
amount to the execution of an act. Both these ideas well predate Austins theory
(for a detailed overview see Sbis 2009a). However, his is the first account that incorporates both in a radically innovative philosophical explanation of linguistic
communication. In a nutshell, speech act theory advances the fundamental claim
that speech is a form of action rather than a device for describing the world. The
theory compels us to see communication not simply as the passing of information
between a speaker and a hearer, but rather as the consequential and mutual acting
of participants upon each other. The relevance of the theory extends to the analysis
of linguistic communication in its many forms, and more fundamentally to the
understanding of human communication tout court. Discussion of the theory soon
crossed the borders of language philosophy. As a consequence the literature on the
subject is vast, and some references to speech acts are a staple in artificial intelligence, discourse analysis, second language acquisition, interpreting, education,
and social psychology.
374
2.
Elena Collavin
Origins
During the nineteen thirties Logical Positivism was a mainstream philosophical

movement. Born of polemic with metaphysical theories, the approach charged
philosophy with an apparently modest task: to elucidate, clarify and expose nonsense. According to this disciplinary frame, to understand a sentence amounts to
knowing what its truth conditions are. Only statements that are potentially empirically verifiable make sense. For instance one understands the sentence
(1) Giovanni owns a red car
if one knows under which conditions (1) would be true. Within this framework the
meaning of the sentence can be described as its truth conditions. In order to calculate the truth value of a sentence one must be able to unequivocally identify its referent, in this case, for example, Giovanni. As a consequence much debate at the
time was devoted to statements with non existing referents, like
(2) The unicorn owns a red car
According to some authors, such a statement would be nonsensical since in our
world, no referent could ever be found for the unicorn, and hence no truth value
could ever be bestowed upon it; according to others, (2) is simply false. By either
account a large number of perfectly common statements are rendered uninteresting
for the theory. Only certain kinds of descriptive sentences of ordinary language can
be dealt with by the theory. This theoretical approach is chiefly concerned with the
capacity of human language to depict or represent nature, very much like natural
sciences aim at doing in formalized ways. Although language is obviously used to
do many other things aside from describing, Logical Positivism only concerned
itself with the referential, descriptive faculty of language, other uses being deemed
marginal. Ludwig Wittgenstein, whose Tractatus (1922) had been a main source of
inspiration for Logical Positivism, famously changed his mind and went on to propose a radically different theory of linguistic meaning, which bears affinity with
Austins. In a nutshell, in his late works Wittgenstein identifies linguistic meaning
with whatever participants display they believe it to be, in conformity with the type
of social activity they are engaged in. Meaning in such a view becomes an actionoriented business of strictly local, interpersonal conventions, and is de facto detached from the linguistic forms of communication supporting it. While Wittgenstein believes that linguistic forms obey rules, contrary to Austin he does not focus
his interest on what these rule are.
Austin knew Wittgensteins work. Wittgensteins Philosophical Investigations
were discussed during some of the Saturday morning weekly meetings with colleagues that Austin held over the years in various colleges (Warnock 1969:14).
Austins proposal espouses Wittgensteins action-oriented view of linguistic meaning but strives to regiment it by painstakingly trying to pin down, categorize, and
Speech acts
375
peel apart the layers of actions performed linguistically on the one hand, and on the
other to correlate form and force of utterances. First, he starts with a rebuttal of the
received view of linguistic meaning.
3.
Austin on performative utterances
The pars destruens of Austins argument starts from a kind of utterance which
looks like a statement, and grammatically, I suppose, would be classed as a statement, which is not nonsensical, and yet is not true or false (Austin 1979:235).
Here are some examples:
(3)
(4)
(5)
(6)
(7)
I promise I will come tomorrow

I apologize
I will kill you
I award the plaintiff one million dollars in compensation
I bet you ten dollars that tomorrow it will not rain in San Diego
Coining a new word and an ugly word Austin calls these utterances performatives to indicate that issuing the utterance is performing an action (Austin
1962: 6). In the appropriate circumstances with (3) the speaker takes on an obligation to come tomorrow, with (4) the speaker issues an apology, with (5) the
speaker issues a threat (that is a promise to do something dreadful to the interlocutor), with (6) the speaker attributes an entitlement to the plaintiff, and with (7) the
speaker proposes to bet ten dollars that tomorrow it will not rain in San Diego.
Noticeably, the truth value of the above utterances does not seem to have much
role to play. For instance, to the utterance
(8) A: I promise I will come tomorrow
it would not be in order to respond
(9) B: Thats true
but rather something to the tune of ok, or I do not believe you, or even you always say that which all acknowledge the fact that the speaker was making a (perhaps unreliable) promise. Explicit performatives share a similar grammatical
structure characterized by the use of the present tense indicative and a first person
singular personal pronoun as subject. Such performatives are used to simultaneously name and perform the action denoted by the verb.
The following two utterances
(10) Susan bets you $10 that tomorrow it will not rain in San Diego
(11) I awarded the plaintiff $1,000,000 in compensation
376
Elena Collavin
are descriptive rather than performative due to the use of a third person subject and
(10) and the past tense (11); accordingly, uttering (10) does not result in Susan betting $10, and uttering (11) does not award any sum of money.
The type of verb involved also plays a part, as in
(12) I own a red car
which shares the grammatical structure of an explicit performative but lacks a verbum dicendi, which is another prerequisite for a verb being used as a performative.
Unlike performatives, (12) can thus be assessed on the truth value scale. Austin
calls these latter types of utterances constatives because contrary to performatives they are used to describe rather than to do something. While the linguistic form of utterances seems to be an essential aspect of their being performatives, in addition, a number of extra linguistic requirements must be fulfilled
for a speaker to perform the intended action. These conditions are of a social and
psychological nature:
(A.I) There must exist an accepted conventional procedure having a certain conventional effect, that procedure to include the uttering of certain words by certain persons in
certain circumstances, and further,
(A. 2) the particular persons and circumstances in a given case must be appropriate for
the invocation of the particular procedure invoked.
(B.1) The procedure must be executed by all participants both correctly and
(B.2) completely.
(C.1) Where, as often, the procedure is designed for use by persons having certain
thoughts or feelings, or for the inauguration of certain consequential conduct on the part
of any participant, then a person participating in it and so invoking the procedure must
in fact have those thoughts or feelings, and the participants must intend to so conduct
themselves, and further
(C.2) must actually so conduct themselves subsequently.
(Austin 1962: 1415)
When (5) is uttered by an actor on stage it does not result in the (real) speaker
threatening the (real) addressee, and (6) generates an entitlement for the plaintiff
only if it is pronounced during the appropriate judicial setting by the judge in
charge of deciding the case. The same judge uttering the very same words while
talking with her husband at the restaurant does not create any entitlement. Austin
suggests a way of testing if a verb is performative by putting hereby before the
verb conjugated in the first person singular of the indicative present. As expected
I hereby own a red car is not an acceptable (and thus not a performative) utterance.
Performatives are successful or unsuccessful, felicitous or infelicitous, rather
than true or false. Austin distinguishes two main ways in which a performative utterance can fail. Violations to the conditions under A and B produce misfires; the
act is performed purposefully but is void for lack of the required social convention,
Speech acts
377
persons, and circumstances. Violations of the conditions under C produce abuses;

the act is professed but it is hollow because of lack of the appropriate psychological
state or consequent behavior. In order to be executed some performatives may need
uptake on the part of the interlocutor. An individual person can issue an order or
permission on his own, but cannot make a bet without the cooperation of the interlocutor. When it comes to the psychological state of an individual, Austin makes
clear that while a promise uttered with no intention to honor it is problematic, for
all social purposes the promise has been issued and remains valid, our word is our
bond (Austin 1962: 10). There are plenty of cases in which it is unclear if a faulty
or incomplete performance of a performative utterance results in an infelicity.
While administering the constitutionally prescribed Presidential Oath of Office
to President Obama, Chief Justice John Roberts botched it by inverting the word
order of the oath. As a result the Presidential Swearing In was not perfectly
executed word by word. It was unlikely that the imperfect recitation of the oath
witnessed by millions of citizens would have had legal bearing for the Presidential
Office; nevertheless, later in the day the oath was administered a second time in
closed chambers, and apparently perfectly executed (Toobin 2009).
4.
Austins theory of the total speech act in the total situation
Austin claims that there is a large number of performative verbs in English which,
when put in explicit performative form, effect an action. Here are some: to
christen, to bequeath, to promise, to acquit, to sentence, to resign, to excommunicate, to vote for, to bet. However, Austin had other things in mind than sieving the
dictionary for special classes of verbs. His rhetorical edifice compels us to first see
the difference between two distinct classes of verbs and utterances, only to show
that the distinction does not hold. After having first created the constative/performative dichotomy, he ultimately erodes it and argues that all utterances are in
fact used to perform speech acts. His argument is twofold.
Firstly, there is the matter of linguistic form. We have seen that performatives
are a special class of utterances with a particular linguistic form, but this is not always the case. A sign on the wall of a room that says:
(13) No smoking
or the utterance
(14) Smoking is not allowed
are unequivocally performatives but do not have the explicit first person present indicative performative form. Furthermore,
(15) I most definitely will come tomorrow
378
Elena Collavin
amounts to assuming a commitment, but could not be made more explicit by adding hereby.
The same action, moreover, can be performed by different linguistic means.
Consider:
(16)
(17)
(18)
(19)
(20)
(21)
I compliment you on your wonderful performance!

Wonderful performance!
You were marvelous!
I have never heard a more beautifully executed performance in my life
Glinka himself would have loved it
Bravo!
All amount to complimenting the performer. Clearly, utterances can be performative without having the form of a performative. Austin suggests that explicit
performatives are just more unequivocal than implicit performatives in signaling
the action that is being carried out. Austins argument implies that underlying a
multiplicity of linguistic forms some aspect of the action performed (what we call
the force of the utterance) in (16) to (21) is the same and can be made explicit using
the performative form. The verb used might not be a good force indicator after all;
rather, the mood of the verb, the intonation, adverbs, connecting particles, and overall the circumstances of the utterance all provide indications of what kind of action
is performed (Austin 1962: 7377). Linguistic form, that is, can be deceiving, and
many utterances that do not look like explicit performatives can be performatives.
Hence the domain of performatives grows, while that of constatives shrinks.
Secondly, in Austins original distinction, using a performative utterance is an
act, i.e. is doing something, whereas using a constative utterance is just stating
something but not doing something (Austin 1979: 247). Statements are supposed
to be either true or false; performative utterances either felicitous or infelicitous,
which is unsatisfactory. First, the very concept of truth is problematic. Consider the
utterance
(22) France is hexagonal
which illustrates the difficulties that can arise if we try to use the concept of truth in
the domain of everyday language. If with the statement we meant that France is of
a perfect hexagonal shape than the statement would be false, but in many contexts
we would agree that the description is perfectly acceptable, or true. The truth value
of many descriptive statements is not easily calculable. Furthermore, statements
are liable to be assessed not only for their correspondence or failure to correspond
with facts; they are also liable to the same infelicities that can affect performative
utterances. Let us consider what is wrong with the following sentences:
(23) The cat is on the mat but I do not believe it is
and
Speech acts
379
(24) All Johns children are bald but John hasnt got any children (Austin 1979:
248)
In these examples of Moores paradox, Austin points out that there is avowal of
insincerity that makes the statements nonsensical or rather infelicitous for violation of the conditions under C (cf. above). Frege proposed that to assert a proposition is to present it as true. (But see Alston 2007: 11). With a shift from the domain of propositions to that of utterances we can say that asserting something is
expected to be associated with the belief that what is asserted is true.
Besides, the same criteria that determine the felicity of a performative (appropriate persons and circumstances, the existence of social conventions, correctness
of the execution, subsequent conduct, and the speakers psychological state (feelings, thoughts and intentions)) all come into play as much in uttering statements as
in uttering an order or a promise.
Furthermore, performatives can also be assessed for their correspondence to
fact. Both the following utterances
(25) I killed them
and its explicit performative version
(26) I confess that I killed them
constitute an admission of responsibility, but at the same time both proffer a statement that is either true or false.
We see then that stating something is performing an act just as much as giving an order
or giving a warning; and we see, on the other hand, that, when we give an order or a
warning or a piece of advice, there is a question about how this is related to fact which is
not perhaps so very different from the kind of question that arises when we discuss how
a statement is related to fact. Well, this seems to mean that in its original form our distinction between the performative and the statement is considerably weakened, and indeed breaks down. (Austin 1979: 251)
Austin rejects the constative/performative dichotomy in favor of a general theory

of Speech acts, arguing that all utterances (as actual occurrences of linguistic behavior) are the performance of an action. Utterances do not correspond to well
formed grammatical sentences, or to conversational turns (Sacks, Schegloff, and
Jefferson 1974). Rather, the speech act can be isolated in utterances as something
different and distinct from the meaning of the uttered sentence, something that
Austin, using Freges terminology, calls the force of the utterance.
[] besides the question that has been very much studied in the past as to what a certain
utterance means, there is a further question distinct from this as to what was the force, as
we may call it, of the utterance. (Austin 1979: 251)
What we need besides the old doctrine about meanings is a new doctrine about all the
possible forces of utterances [] (Austin 1979: 251)
380
Elena Collavin
For instance:
(27) Is the door shut?
and
(28) Shut the door
have the same propositional content but differ in the action that the utterance performs, hence in their force, the first utterance being a question, the second an order.
From now on the unit of interest for the theory is the speech act, intended as a verbally performed action. If the business of a speech act is for the speaker to do something with the words he or she utters, a theory of speech acts must explain in which
respect saying something may amount to doing something. After all, doing
something is a very vague expression (Austin 1962: 91). According to Austin
saying something amounts to simultaneously performing three types of acts:
i) A Locutionary act, which is the uttering of a sentence with sense and reference.
The locutionary act can be broken down into a phonetic act (the act of uttering
certain noises), a phatic act (the act of uttering sounds that have sense and reference), and a rhetic act (the act of uttering noises that belong to a certain language vocabulary and grammar). A speech act is always the utterance of a
phone, a pheme and a rheme.
ii) An Illocutionary act, which is the action performed by virtue of the force associated with a given linguistic expression. It is the performance of an act in
saying something, as opposed to the performance of an act of saying something
(Austin 1962: 99). This level of action depends on the social conventions that
allow us verbally to carry out clearly recognizable actions. The illocutionary
force of an utterance can be isolated by asking in which sense we were using a
given utterance. When we describe metapragmatically an utterance as an order,
a request, a suggestion, a statement, a promise, a threat, we refer to its illocutionary force. For instance, we could say: By saying the soup you cooked was
very poor she was criticizing him. We can say so fairly unequivocally because we usually can recognize a criticism (and an order, a promise, and so on)
when we hear or see one. Some verbs (in English, for instance, order, request, promise) are inherently illocutionary: when put in the first person
present indicative they provide an explicit performative, whereas others cannot
be used in the same way (I convince you or I offend you do not work the
same way as an explicit performative).
iii) A Perlocutionary act, which is the production of a consequence by the utterance. Contrary to what occurs at the illocutionary level, perlocutions are not directly achieved by the conventional force of an utterance. They can be intentional or unintentional, and they might involve unexpected effects, different
from those of an illocutionary act. Also, perlocutions occur at a further level, as
Speech acts
381
the interlocutors actual reaction to the speech act. A perlocutionary effect of

the utterance the soup you cooked was very poor might be that ones interlocutor is chagrined, gets angry, pours the soup down the sink, or all three.
The distinction between illocution and perlocution is not clear cut, and it is not at
all apparent that the perlocutionary level can be fully considered as a single level of
action embodied in a given utterance (Austin 1962: 108115). In order to identify
the difference it might help to distinguish between the effects that are built in to
given expression by virtue of its illocutionary force, and those that are not. For instance a (felicitous) order just by virtue of its illocutionary force will generate an
obligation in the person who receives it, but how the individual acts upon the obligation (complying, refusing, questioning, being offended by it, and so on) belongs to the realm of perlocutions.
The locutionary and the perlocutionary levels of the speech act have generated
much debate (Searle, 1969; Searle, 1973, Warnock, 1969, Davis, 1980) but the illocutionary level is the nucleus of the Austinian theory and its later developments.
According to Austin the illocutionary act is connected with the production of effects in the following sense:
(i) The force and the content of the utterance must be understood. I cannot say to
have warned an audience unless it hears what I say and takes what I say in a
certain sense (Austin 1962: 115).
(ii) The illocutionary act needs to take effect (see the above example of how giving an order generates an obligation).
(iii) In many instances there needs to be the manifested cooperation of the audience for the illocutionary act to come into being. The necessary uptake encompasses an interlocutors understanding of the content and force of an utterance, but it can also require a specific second part (or response) to the
utterance that ratifies the first part. While one may protest or give an order on
her own, one definitely cannot bet $100 that tomorrow it will not rain in San
Diego without someone taking up her bet.
Austins theory is rich with suggestions and loose ends, a number of which remain
unresolved. Scholars continue to debate how to distinguish clearly the three levels
of the speech act, how to classify speech acts, how to assess the role of force indicators, and how to apply the values of truth and falsity to speech acts.
5.
Grices theory of non natural meaning
After Austin, the development of speech act theory was greatly influenced by Paul
Grices notion of non natural meaning (Grice 1957; for an overview of Grices theory cf. Huang, this volume). In his work, Austin did not focus on the role that the
382
Elena Collavin
speakers intentions play in the production of speech acts. While he recognized

that the appropriate inner state of the speaker may be fundamental for carrying out
felicitous acts of a certain kind, Austin was chiefly concerned with how particular
linguistic expressions, social conventions, the interlocutors uptake, and allocation
of responsibility to the speaker are involved in carrying out verbally performed actions (Sbis 2009a). Grices theory of speakers meaning revolutionized the study
of language, including speech act theory, by bringing to the fore the communicative intentions of speakers. Grice distinguishes between the meaning of certain
natural events and the intentional communicative acts of humans, or natural signs
or symptoms on the one hand, versus signals (Clark 1996: 126) on the other.
The former have a natural relationship with what they stand for; they are what
Peirce calls indexes (Peirce 1989). The latter instead rely in part at least on the
speakers intentions to signal something. For instance, the characteristic red spots
on a childs face are natural signs that the child has measles, and have what Grice
calls a natural meaning (thus, the red spots on her face mean she has measles).
However, the doctors telling gesture after having examined the child, or her declaring that the child has measles manifests the doctors intention to communicate
what she believes to be the case (thus the doctors gesture meant that the child has
measles). The latter case in Gricean terms is an instance of non natural meaning. Linguistic meaning is always of a non natural kind, even if our talking is
likely also to provide a wealth of information that belongs to the realm of indexes
(indicating our gender, age, geographic origin, social background, and so forth).
Grice further distinguishes between speakers meaning and sentence meaning.
This is based on the notion that understanding someone amounts to understanding
what they intend to communicate to us. In a modified formulation that takes into
account some later arguments (proposed by, among others, Strawson, Searle, and
Sperber and Wilson), speakers meaning is defined as follows:
In presenting s to audience A, a speaker S means for A that p if and only if: (i) S intends
in presenting s to A that A recognize that p in part by recognizing that i. (Clark 1996:
130)
where i is the speakers intention. Speakers meaning is a type of reflexive intention, i contains reference to itself. A key part of the meaning resides in the interlocutors understanding of the effect that the speaker intends to achieve. A communicative act occurs through signaling and recognizing, and a key part of the
process involves understanding the communicative intentions of the speaker. Communication is a joint action, as becomes apparent in cases where the literal meaning of a communicative act appears to be lacking, unclear, incomplete, false or absurd, but communication still succeeds. Indeed, Grice later proposed a general
theory of meaning based precisely on what he calls the Cooperative Principle,
which alongside the notion of conversational implicature provides a universal explanation of why, with the cooperation of ones interlocutor and with mutual infer-
Speech acts
383
ential calculations, it is possible to communicate with the most eccentric linguistic

and non linguistic means.
6.
Searles theory of speech acts
Searle (1969) criticizes Grices account of meaning because it relies almost exclusively on the speakers intentions. A given meaning can be produced by any sentence given that the circumstances make possible the appropriate intentions
(Searle 1969: 45). Searle tries to strike a balance between intentionality and the
role of conventions in the formation of meaning.
In the performance of an illocutionary act in the literal utterance of a sentence, the
speaker intends to produce a certain effect by means of getting the hearer to recognize
his intention to produce that effect; and furthermore, if he is using words literally, he intends this recognition to be achieved in virtue of the fact that the rules for using the expressions he utters associate the expression with the production of that effect. It is this
combination of elements which we shall need to express in our analysis of the illocutionary act. (Searle 1969: 45)
Assuming that the illocutionary force is signaled by indicators of force (that Searle
calls Illocutionary Force Indicating Devices) the challenge is to explain exactly
how the two force and force indicators are linked. For Searle the force indicators are constitutive of the force. Rather than revealing it they create it. Whereas
for Austin felicity conditions needed to be fulfilled for the speech act to be carried
out successfully, according to Searle felicity conditions are constitutive of the act
itself. He distinguishes two elements in the syntactic structure of the sentence,
which he calls the propositional indicator, and the illocutionary force indicator.
(Searle 1969: 30) This is his version of the distinction between the illocutionary act
and the propositional content of the illocutionary act.
Searle rejects Austins threefold distinction within the locutionary level of the
speech act (phonetic, phatic, rhetic) in order to locate the locutionary act at the
level of enunciation (the utterance act). Others have also criticized Austins
tripartite distinction claiming that it is hard to keep separate the locutionary act
from its constituent levels (Warnock 1969). While some scholars have described
Searles work as largely a systematization of Austins (Levinson 1983: 238), according to Sbis (2009a) Searles theory sharply diverges from Austins. For Searle
the speech act coincides with the illocutionary act, a conclusion that stems from the
fact that Searle views speech acts as physical and psychological gestures of the
speaker, whereas for Austin the action corresponds with the (necessarily interpersonal) ascription of responsibility to the agent for the production of a given illocutionary effect.
Searles illocutionary act/speech act is divided between the illocutionary force
and the propositional content, schematized by: F(p), where F describes the illocu-
384
Elena Collavin
tionary force indicators and p is the proposition. The propositional content

determines the conditions of satisfaction, truth conditions in the case of belief,
obedience in the case of a promise. Searle proceeds by analyzing the conditions
sufficient for the successful performance of illocutionary acts.
A proposition is what is asserted in the act of asserting, what is stated in the act of stating
[] an assertion is a [] commitment to the truth of a proposition (Searle 1969: 29)
The illocutionary force indicator shows how the proposition is to be taken, or to put it
another way, what illocutionary force the utterance is to have; that is, what illocutionary
act the speaker is performing in the utterance of the sentence.
(Searle 1969: 30)
Some conditions are common to most illocutionary acts, while others are specific
to some types of acts. Austin proposed a classification based on a list of performative verbs, whereas Searles is based on the felicity condition of illocutionary
acts: all have a propositional content which specifies which propositional content
the speaker has to express, preparatory conditions which are contextual but linked
to the inner states of the speaker (what he thinks, what he believes, and so on), sincerity conditions and the essential condition which specifies which kind of illocutionary act a speaker intends to carry out. In his later work, Searle focuses on three
main criteria for the classification of speech acts: the purpose of the type of act
(what he calls the illocutionary point), the direction of fit of the type of act (from
words to world, or from world to words) and the expressed psychological state of
the speaker (Searle 1975).
In either case, in Searles writing the illocutionary effect that results from successfully carrying out an illocutionary act depends on the interlocutors understanding of the force and propositional content of the act, and coincides with the interlocutor recognizing the speakers intentions. Sbis (2009a, 2009b) notices that
Searles theory is problematic precisely because it makes the illocutionary effect of
the act coincide with the recognition of the speakers intentions by the hearer.
Rather than looking at the specific illocutionary effects of an act, Searles analysis
focuses on the types of intentions of the speaker, thus constructing action as the
psycho-physical gesture of the speaker. Sbis (2009b) contends that contrary to
Searle, to Austin illocutionary acts are conventional in the sense that they have
conventional effects and underlie the interpersonal rather than the mental aspect of
uptake of the illocutionary force. On the role of intentionality Searle explains that
he sees language as the natural extension of non linguistic biological capacities,
and linguistic meaning as an extension of fundamental pre linguistic intentionality
of a biologically nature like hunger and sexual desire. Perceptions and beliefs are
also forms of intentionality because to hold beliefs is to hold something that determines conditions of satisfaction, and to hold the capacity to recognize if they are
fulfilled or not (Searle 2007: 21). Indeed, to Searle linguistic action is in essence a
mental act of the speaker. What gets communicated in speech acts are intentional
Speech acts
385
states (Searle 2007: 28). Thus, the analysis focuses on how propositional content,
conditions of satisfaction, psychological mode and direction of fit function in the
manifestation of intentions. For instance, beliefs aim to represent how things are,
thus they have a mind-to-world direction of fit, while desires have a world-to-mind
direction of fit. Searle (2002: 7) conceives of the structure of mind and the structure
of speech acts as parallel because the structure of mental states and the structure of
intentional states are similar. While he concedes that at times our verbal actions do
not have the principal scope to transmit our mental states, he underlines that even
when we carry out speech acts that have not as principal role the function of expressing feelings, we still express a mental state in the form of a sincerity condition
(Searle 2002: 7).
7.
Searle on indirect speech acts
Searle points out that a speakers utterance meaning and the sentence meaning frequently diverge (Searle 1975: 59). This is the case most obviously with irony,
metaphorical uses, and insinuation. However, even in most ordinary circumstances
we can often see no correspondence between the three major sentence types in
English (declarative, imperative and interrogative) and the forces that prototypically correspond to them (stating, ordering and requesting). Consider
(29) I want you to do it (Searle 1975: 59)
which is a request but has the declarative form of a statement. This is a very common occurrence. In English, for instance, imperatives are routinely used to invite,
as in take a seat and come in. Utterances that have the illocutionary force indicators for one kind of illocutionary act can then be used to perform another kind
of illocutionary act. This is a serious problem for the theory, which is based on the
conventional nature of the illocutionary force, and on the existence of predictable
indicators of force. A way to treat utterances whose force differs from what
their force indicators suggest is to assume that they have a literal force, which is
associated by rule, and an indirect force, which is inferred by virtue or further information available to an interlocutor. Searle offers an explanation for the case of
directives based on his own felicity conditions. Let us consider the utterance can
you pass the salt? Given a Speaker S, a hearer H, and an action A (passing the salt)
the felicity conditions for such illocutionary act are:
Preparatory condition: H is able to perform A
Sincerity condition: S wants H to do A
Propositional content condition: S predicates a future act A of H
Essential condition: Counts as an attempt by S to get H to do A
(Searle 1975: 71)
386
Elena Collavin
The request is indirect because the speaker seeks to ascertain if the interlocutor is
capable of performing the action while in fact the speaker is actually requesting the
hearer to pass the salt. The ability of the hearer to pass the salt is a preparatory condition for requesting to pass the salt. Still, this does not explain why it is the case
that only certain requests are successfully made by asking if the interlocutor can
perform the wanted action. As a preparatory condition for any request the speaker
must hold the belief that the interlocutor can carry out the request, so why is it the
case that sometimes the speaker asks? The answer may be found in the rituals of
social interaction and the obligations of politeness (Brown and Levinson 1987), a
dimension that is not integrated in the notion of illocutionary force.
8.
Bach and Harnishs theory of speech acts
Although by far the most influential, Searles approach is not the only one to emphasize the role of speakers intention for explaining illocution. With the aim of
offering a philosophically inspired explanation of linguistic communication, Bach
and Harnish (1979) propose a different but equally Gricean-inspired, speaker-intention-focused classification of speech acts. They distinguish between conventions and rules, the conventions being actions that when performed in the appropriate circumstances count as something else, while rules are socially expected
forms of behavior (Bach and Harnish 1979: 121). Based on their categorical division between conventions and rules the authors draw a distinction between communicative and conventional illocutionary acts: Communicative illocutionary
acts succeed by means of recognition of intention, whereas conventional ones
succeed by satisfying a convention (Bach and Harnish, 1979: 110). Conventional
illocutionary acts are endemic to particular institutions. For an utterance to be a
conventional illocutionary act not only must it be the utterance that the convention requires, it must be issued by the right person under the right circumstances
(Bach and Harnish 1979: 110). Examples of conventional illocutionary acts are
christening, billing, sentencing, disqualifying. Conventional illocutionary acts all
belong to the class of what they call effectives and verdictives. These types of acts
are considered not essentially communicative because they are not implemented
by recognition of intention but by convention. Communicative illocutionary acts
instead are based not on conventions but on the recognition of the intentions of
the speaker (the illocutionary effect), and their classification is based on the
illocutionary intents of the speaker. Illocutionary intents correspond to types of
attitudes expressed by the speaker. The recognition of intentions occurs when the
hearer takes the speakers utterance as reason to believe that the speaker has the
expressed attitude. Although Bach and Harnish assign a different role to conventionality in the discussion of illocutionary force, they side with Searle in limiting
the illocutionary force to the recognition of intention of the speaker, and leaving
Speech acts
387
the conventional effects of illocutionary acts within the realm of perlocutions

(Sbisa 2009b).
9.
A plan based theory of speech acts
Philip Cohen and Raymond Perrault (1979) propose to describe speech acts as
planning operators. While espousing Searles recognition of an intention based notion of speech act, they envision a formalization that would allow, for any communicative act, to isolate which intentions are involved and on which bases the
speaker expects and intends those intentions to be recognized (Cohen and Perrault
1979: 178). They propose a theory that formally models the possible intentions
underlying speech acts. Intentions are conceived as plans and a link is shown between plans and non linguistic behavior.
This approach is focused on the individual who in the pursuit of a goal, plans
and enacts a sequence of actions, the communication of desires, intentions, and
emotional states, and which might include facilitating or making sure that others
recognize these plans. Actions are defined in terms of operators that are broken
down in preconditions, effects and bodies, the latter being the means by which effects are achieved (Cohen and Perrault 1979: 178). Operators are representations;
they are evaluated on the basis of ones representation of the world (Cohen and
Perrault call the speaker problem-solver). The idea is that part of ones model of
the world consists of a symbolic description of other peoples models of the world
(something akin to I know what you know or more modestly I believe that I
know what you believe). Speech acts act as operators on the models that speaker
and hearer maintain of each other.
This is not much different from Grices model for calculating conversational
implicatures (Grice 1975), which occur when what is said is different from what is
meant, so that the accurate comprehension of the literal meaning of what is being
said would not constitute understanding. Implicatures are pragmatic inferences
that require the hearer to make a mental calculation in order to decide what the
speaker really meant to say, for if taken literally what was said would not be fitting
in the current conversation. Grice bases his model for calculating implicatures on
the general Cooperative Principle, a descriptive rule stating that participants contributions to a conversation are relevant to the task at hand. The principle acknowledges that in order to comprehend non literal meaning we need to believe that our
interlocutor is being cooperative, and as a consequence that his or her contribution
must be relevant to the task we are involved in. A successful performance of a
speech act produces some modification of the speakers model of the hearer, and of
the hearers model of the speaker (Cohen and Perrault 1979: 179). The question is
which modifications? Modeling their proposal on Chomskys notion of the ideal
competent speaker, the authors focus on a competence theory of speech acts that
388
Elena Collavin
enables to state the conditions under which a competent participant can make the
determination necessary to classify ones utterances as amounting to the performance of a given speech act.
The model has been embodied in a computer program, and detailed for two
kinds of speech acts: making requests and giving information. But even though
Cohen and Perraults theory has been influential on artificial intelligence studies
applying speech acts theory, it suffers from the same flaws that plague Searles
model: The speech act occurs within the speakers mind as the expression of his or
her intentions, leaving no room for a constitutive contribution on the part of the interlocutor, whose role is only to decipher the speakers intentions.
10.
Speech acts and the analysis of discourse
To apply speech act theory to the analysis of discourse means to try and isolate the
actions that are being performed in ongoing interaction or in a text. No matter how
speech acts are conceived, the analysis will consist in mapping discourse onto categories of speech acts. As Levinson (1983: 278) points out, this could be either
trivial, if we were to wrongly assume that the force of an utterance is indicated literally by linguistic devices, or otherwise it could involve the attempt to predict the
force of a part of discourse in real circumstances. This proves to be largely an impossible task because the function of a given utterance is so essentially dependent
on the overall circumstances in which it occurs that there may be little left to do for
any theory of speech acts. Consider the following example given by Levinson,
which is quoted from Sinclair and Coulthards ethnographic work in the classroom:
(30) Teacher: What are you laughing at?
Child: Nothing
(Sinclair and Coulthard 1975: 30)
It is the second part of the exchange as much as the scholastic context and the social roles of the participants that provide the cue that the teachers utterance is an
order. The following example is taken from a 1993 televised political debate in
Italy:
(31) Casini: I want to ask you this question. Do you want to solve the problems of
the south of Italy, or is it the case that you do not give a damn about the problems of the south of Italy?
The closed question, which seems to syntactically anticipate either yes or no
as an answer, is most likely a challenge and an accusation, which as it has been
shown on the bases of corpus studies, is usually met with a denial (Schegloff, Jefferson and Sacks 1977). Further, an assertion like The door is open in the appro-
Speech acts
389
priate circumstances can function as a declarative, a request, a criticism, a question, a complaint, and possibly other kinds of actions as well. Because ultimately it
is not possible to associate form with function in the analysis of speech acts, the
usefulness of the theory to explain how actions are performed in discourse is limited.
As Levinson points out, failing the hypothesis of the literal force, we are left
with an infinite array of language games (Wittgenstein 1953: 1011) within which
utterances acquire their meaning (and their force, if we wish to keep the two separated) by virtue of local, situated agreements among participants. Within a theory
of speech acts it must be possible to unpack an expression and reveal the underlying illocutionary force of it, and assume that the original expression and the explicit one will display the same force. However, in the analysis of interaction we
often find that it is not possible to univocally isolate one particular illocutionary
force as it has been categorized within the theory. Further, some of the very reasons
why it is interesting to study how people discursively act with one another might be
seriously impeded if we are to assume that in the analysis of the speech act other elements, such as those indexing social relationship in terms of closeness and hierarchy, are irrelevant for the force of the utterance. For instance, the utterances
(32) Can you pass the salt?
(33) Would you be so kind as to pass the salt?
(34) Salt please.
cannot really be described as fully equivalent in terms of their illocutionary force
precisely because they are not equivalent in how they index social relations between participants. Arguably, this aspect, while it does contribute to the forcefulness and the compulsory strength of a request, is not clearly accounted for in terms
of illocutionary force. This is even more apparent in languages that use grammatical markers to distinguish formal versus informal forms of addressing. In Italian
for instance
(35) Mi passi il sale? Do you [+informal] pass me the salt?
and
(36) Mi potrebbe passare il sale? Could you [+formal] pass me the salt?
have their illocutionary force at least partially embodied in the very grammatical
markings that reveal the relationship between participants: the intimate or informal
relationship triggers a more forceful and direct request in (35), while the formal relationship in (36) triggers a weaker and less imposing request. Within the theory
there is no clear way to account for distinguishing between an illocutionary act of
request made between strangers and one made between intimate companions, but
to ignore the socially loaded deictic aspects of the exchange is to leave out elements that might actually weight on the force of the action performed. Further,
390
Elena Collavin
Levinson has shown that it is hard to associate a certain set of Searlian felicity conditions to actual elements of speech exchanges: questions, for instance, as a category are too variable to be tied by a given set of felicity conditions, and even
apologies do not all have the same requirements of felicity conditions (Levinson
1983: 281; Levinson 1979).
Responding to the objections moved by Levinson and Schegloff to the applicability of speech act theory to the analysis of discourse, Rees (1991) defends the
specific contribution that the theory can provide to the analysis of verbal interaction, which she suggests be used alongside Conversation Analysis. Van Rees argues that already in Searles work there is a recognition that a single utterance
might have multiple forces:
(37) Its really quite late
can be an objection, a suggestion, or even a request (Searle 1969: 70) (and a load of
further acts), and in interaction the illocutionary force might reside in supra- and
sub-sentential units which do not coincide with a sentence at all (Rees 1991: 34).
What counts in the attribution of illocutionary force to a part of discourse then is
not a simple, literal association between utterance and force but, as she claims, the
possibility to calculate in a replicable, non arbitrary way the force of a given stretch
of talk. This compromising approach leaves unresolved the key issue of linguistic
form in the matter of the calculability of the illocutionary force. After all, it just
shifts the problem in the realm of the competence of the speaker, but still profitably
uses the notions of performative and illocutionary force in the analysis of discourse, and has won over many scholars interested in the study of verbal interaction. While the difficulty to find a rule to relate form and function in a wide array of
real occurrences of speech acts remains unresolved, key concepts from the theory
have been widely adopted within the domain of discourse analysis.
11.
Beyond speech act theory: Action and the body
The speech act approach has largely been abandoned by those dedicated to the ethnographic study of linguistic action. Critics note that in both Austins and Searles
formulation, the unit of analysis is the surgically delimited, well-formed, grammatical sentence, and that despite Austins early intuitions the developments of the
theory devote no attention to discourse level phenomena which cannot be analyzed
without taking social interaction (Goffman 1983) into account. Echoing Austin,
Agha points out that
the utterance of a performative locution is only a fragment of the performative act; it is
merely the leading incident (to use Austins phrase) in a larger order of happenings
whose shape must be specified by further semiotic accompaniments in order for the utterance to have the effect in question (Agha 2007: 58).
Speech acts
391
In Austins formulation the total speech act is to be understood in the total situation, a claim that recognizes the fundamental role of participants other than
the speaker, and of social arrangements alongside syntactic rules in carrying out
speech acts. The division echoes the dichotomy between the meaning of a sentence
on the one hand and the role of context on the other, where context supplements,
clarifies or modifies the force of the verbal action. A distinction between language
as an autonomous and primordial code and additional information provided by
extra linguistic cues is rejected by those who study cognition in action and who
advocate an ethnographic approach, which considers different kinds of phenomena instantiated in diverse media (Goodwin 2000: 1490) to understand action as
jointly achieved by participants through talk, body movement and material affordances. Goodwin argues that a primordial site for the study of cognition and human
action is an actual situation where several people talk while engaged in a task
which involves their bodies in space, as well as tools and other physical objects
around which they are working. Goodwins approach relies on the detailed transcription and analysis of video recorded data of naturally occurring multi party interactions. The data can be girls enthralled in a game of hopscotch, archeologists
intent on unearthing an artifact, or surgeons performing a laparoscopic operation.
Methodologically it is a far cry from the study of linguistic action which starts with
a set of verbs culled from a languages lexicon, and proceeds with the analysis of
hypothetical utterances pronounced in imagined contexts the confines of which can
be drawn at leisure. While studying linguistic action, this view pays attention to the
multiplicity of semiotic resources available to participants and argues that
the construction of action through talk within a situated interaction is accomplished
through the temporally unfolding juxtaposition of quite different kinds of semiotic resources [] [T]hrough this process the human body is made publicly visible as the site
for a range of structurally different kinds of display implicated in the constitution of the
actions of the moment (Goodwin 2000: 1490).
Rather than being treated as distinct domains, language and material structure in
the environment are integrated components of a common process for the social
production of meaning and action (Goodwin 2000: 1490). This perspective proposes a radically different geography of cognition (Goodwin 2000: 1490), which
is inspired by Erving Goffmans performance oriented approach to complex interactional situations (Goffman 1959, 1983) by works of the Vygotskian circle and by
ethnographic studies of workplace practices.1
Further, Goodwins approach pays close attention to the linguistic nuances of
what is said, to the synchronic details of the interaction as they have been described
within Conversation Analysis (Sacks, Schegloff and Jefferson 1974; Sacks 1995),
and to the organization of face to face interaction. The human body is the locus for
the occurrence of publicly visible displays that are implicated in the cooperative
construction of actions and meanings. In these studies action is rarely achieved by
392
Elena Collavin
one individual, and meaning is a public, social phenomenon that does not somehow reside in the mind of a speaker. In his ethnographic work on the interactional
life of a man who suffered a stroke and was left with the use of only two words,
Goodwin shows how in such an extreme communicative situation meanings are
co-constructed, and actions are the ongoing joint product of the work of participants, always subject to renegotiation and redirection. While apparently distant
from some hypothetical model of interaction between competent participants, this
aphasic mans situation is shown to be exemplary of how actions are interactionally performed, and co-constituted through the cooperation of participants.
Predictably, Goodwin rejects Searles intention centered notion of speech acts.
Searle, like most speech acts theorists, focuses almost exclusively on linguistic
properties of an utterance, everything else being relegated to perhaps necessary but
largely intractable context. A fundamental flaw that Goodwin pinpoints is the role
given to participants in the performance of a speech act. In Searles theory the
hearer exists only in the speakers imagination; the role of the hearer is to understand the force of the utterance, not to co-construct the actions that are accomplished with talk (Goodwin 2000: 1491). In Goodwins work the analysis of action
moves away from speakers intentions and returns to something closer to Austins
notion of uptake. Social action is linked reflexively to its interpretation (Goodwin 2000: 1491). It can be accomplished only when both speaker and addressee
can systematically recognize what is being done not utterance by utterance but
within the evolving scenarios that are being projected into from moment to moment. The mutual understanding of what can be the relevant next move allows
coordinated chains of which are both publicly available and jointly produced.
Speech acts cannot rely on the mental life of a single participant (the speaker);
rather, the notion relies on the sequential analysis of talk in interaction including
such intersubjective mechanism as turn taking and repair (Goodwin 2000: 1492)
where meanings are publicly negotiated. In this context action is understood as
encompassing this interactively organized process of public recognition of meaningful events reflexively linked to the ongoing production of these same events
through the use of appropriate semiotic resources within an unfolding temporal
horizon (Goodwin 2000: 1492).
In 1983 Levinson foresaw that unless a correlation could be found between
form and force in everyday language use it would be hard to maintain a level of
illocutionary force distinct from other action oriented facets of an utterance (Levinson 1983: 282) including social deixis, perlocution, and inferentially available
meanings attachable to the utterance. So far, no version of the theory has offered
such a correlation. Nonetheless, the main tenets of Austins and Searles work
stand as a groundwork that any general theory of language use needs to confront.
Speech acts
393
Notes
1. In particular Ed Hutchins notion of cognition in the wild, which he developed during
his long lasting research of collaborative cognition in the domain of piloting military and
commercial airplanes and freight ships (Hutchins 1995; Hutchins and Palen 1997).
References
Agha, Asif
2007
Language and social relations. Cambridge: Cambridge University Press.
Alston, William P.
2007
Illocutionary acts and truth. In: Dirk Greimann and Geo Siegwart (eds.), Truth
and speech acts, 1030. New York: Routledge,
Austin, John L.
1962
How to do things with words. London: Oxford University Press.
Austin, John L.
1979 [1961] Philosophical papers. Ed. by James O. Urmson and Geoffrey J. Warnock,
3rd ed. Oxford: Oxford University Press. [First published by Clarendon Press].
1979
Linguistic communication and speech acts. Cambridge, MA: MIT Press.
1987
Politeness: Some universals in language usage. Cambridge: Cambridge University Press.
Clark, Herbert H.
1996
Using language. Cambridge: Cambridge University Press.
Cohen, Philip R. and C. Raymond Perrault
1979
Elements of a plan-based theory of speech acts. Cognitive Science 6: 177212.
Davis, Steven
1980
Perlocutions. In: John R. Searle, Manfred Bierwisch and Ferenc Kiefer (eds.),
Speech act theory and pragmatics, 3755. Dordrecht: Reidel.
Frege, Gottlob
1918
Der Gedanke. Eine logische Untersuchung. Beitrge zur Philosophie des deutschen Idealismus 1(2): 5877. [Translated in Gottlob Frege, 1988. Ricerche
logiche, 4374. Milano, Guerini,].
Goffman, Erving
1959
The presentation of self in everyday life. Garden City, NY: Doubleday, Anchor
Books.
Goffman, Erving
1983
The interaction order. American Sociological Review 48: 147.
Goodwin, Charles
2000
Action and embodiment within situated human interaction. Journal of Pragmatics 32: 14891522.
Grice, H. Paul
1957
Meaning. Philosophical Review 66: 377388. [Reprinted in Peter F. Strawson
(ed.), 1971. Philosophical logic. 3948. Oxford: Oxford University Press].
394
Elena Collavin
Grice, H. Paul
1975
Hutchins, Edwin
1995
Cognition in the wild. Cambridge, MA: MIT Press.
Hutchins, Edwin and Leysia Palen
1997
Constructing meaning from space, gesture and speech. In: Lauren Resnick,
Roger Saljio, Clotilde Pontecorvo and Barbara Burge (eds.), Discourse, tools
and reasoning: Essays on situated cognition, 2340. Berlin: Springer.
Levinson, Stephen
1979
Levinson, Stephen
1983
1989
What is a sign? In: Peirce Edition Project, The essential Peirce: selected philosophical writings, Volume 2 (18931913), 410. Bloomington, Indiana: University Press.
Rees, M. Agnes van
1991
The adequacy of speech act theory for explaining conversational phenomena:
A response to some conversation analytical critics. Journal of Pragmatics 17:
3147.
Sacks, Harvey
1995
Lectures on conversation: Volumes I and II. Ed. by Gail Jefferson, with an introduction by Emanuel A. Schegloff. Oxford: Blackwell.
1974
Sbis, Marina
2009a
Speech act theory. In: Jef Verschueren and Jan-Ola stman (eds.), Key Notions
for Pragmatics, 229244. Amsterdam: Benjamins.
Sbis, Marina
2009b
Uptake and conventionality in illocution. Lodz Papers in Pragmatics 5: 3352.
1968
Sequencing in conversational openings. American Anthropologist 70: 1075
1095.
Schegloff, Emanuel. A., Gail Jefferson and Harvey Sacks
1977
The preference for self correction in the organization of repair in conversation.
Searle, John R.
1969
Searle, John R.
1973
Austin on locutionary and illocutionary acts. In: Isaiah Berlin, Lynd W. Forguson, David F. Pears, George Pitcher, John R. Searle, Peter F. Strawson and
Geoffrey J. Warnock (eds.), Essays on J.L. Austin, 141159. Oxford: Oxford
University Press.
Searle, John R.
1975
Indirect Speech acts. In: Peter Cole and Jerrold Morgan (eds.), Syntax and
Speech acts
395
Searle, John R.
1979
Expression and meaning. Cambridge: Cambridge University Press.
Searle, John R.
2002
Speech acts, minds, and social reality. In: Gnther Grewendorf and Georg
Meggle (eds.), Speech acts, minds, and social reality, 317. Berlin: Springer.
Searle, John R.
2007
Illocutionary acts and the concept of truth. In: Dirk Greimann and Geo Siegwart (eds.), Truth and speech acts, 3140. New York: Routledge.
1975
Towards an analysis of discourse: the English used by teachers and pupils.
Toobin, Jeffrey
2009
Supreme Flub, The New Yorker, Blog Post, January 20, 2009. Available online
at: http://www.newyorker.com/online/blogs/tny/2009/01/supreme-flub.html
1969
John Langshaw Austin, a biographical sketch. In: Kuang Tih Fann (ed.), Symposium on J.L. Austin, 321. London: Routledge and Kegan Paul.
1973
Some types of performative utterances. In: Isaiah Berlin, Lynd W. Forguson,
David F. Pears, George Pitcher, John R. Searle, Peter F. Strawson and Geoffrey
J. Warnock (eds.), Essays on J.L. Austin, 6989. Oxford: Oxford University
Press.
1922
Tractatus logico-philosophicus. London: Routledge and Kegan Paul.
1953
14.
Types of inference: entailment, presupposition,

and implicature1
Yan Huang
This article is dedicated to Professor Nigel Vincent, my mentor at
Cambridge, on the occasion of his sixty-fifth birthday.
Introduction
As one of the basic forms of reasoning, inference can in general be defined as a process of accepting a statement or proposition (called the conclusion) on the basis of
the (possibly provisional) acceptance of one or more other statements or propositions (called the premises). Construed thus, it includes deduction, induction, and
abduction. It also includes entailment,2 presupposition, and implicature.
The aim of this article is to assess entailment, presupposition, and implicature
the three fundamental types of inference that are recognised in semantics and pragmatics (see e.g. Huang 2007, 2009g on the scope of pragmatics). Of the three categories of inference, entailment is essentially semantic in nature, though the notion
of pragmatic entailment will be discussed. Presupposition straddles the semanticspragmatics boundary, resulting in semantic and pragmatic presupposition. Finally,
with regard to the two breeds of implicature, while conversational implicature is
pragmatic in nature, conventional implicature can be categorised either as semantic or pragmatic, depending on how the distinction between semantics and pragmatics is drawn. Concepts parallel to conversational implicature such as explicature and impliciture will also be considered. Finally, a brief comparison among
entailment, presupposition, conversational implicature, and conventional implicature in terms of four parameters: defeasibility, negation, conventionality and semantics versus pragmatics will be made.
1.
Entailment
The term entailment is derived from formal logic.3 It refers to a semantic relation
that can be defined in terms of truth in (1).
(1) Entailment
A proposition (or sentence expressing a proposition) p entails
a proposition (or sentence expressing a proposition) q if and only
if the truth of p guarantees the truth of q.
398
Yan Huang
What (1) basically says is this: if p entails q, then q can be taken as following logically and inescapably from p. If p is true, then q is also true. By contrast, if p is
false, nothing is said about the truth value of q. Defined thus, entailment represents
a truth-functional relationship in the sense that its function is to predict the truth
value of a proposition from what is known of the truth value of another. By way of
illustration, let us consider (2). (I use | to stand for entail.)
(2) a. All of the universitys professors are hard-working.
b. | Some of the universitys professors are hard-working.
In (2), if the university in question has professors, then (2a) entails (2b), but not
vice versa. The reason is that if it is true that all of the universitys professors are
hard-working, then there is no way to avoid the conclusion that it is also true that
some of the universitys professors are hard-working. Notice that entailment proceeds from our knowledge of language. It depends on the constituents of relevant
sentences rather than context. Consequently, an entailment is not defeasible, that
is, it cannot evaporate in any linguistic or non-linguistic context. This is why it
must be seen as semantic in nature (see e.g. Huang 2007: 1617).
However, on Fauconniers (1975) and Israels (2004) view, there is also pragmatic entailment. By pragmatic entailment is meant an inference that is on the one
hand default, and on the other defeasible. Two examples are given in (3) and (4).
(3) a.
b.
(4) a.
b.
John can run 100m in 9.9 seconds.

| John can run a slower 100m.
John cant run 100m in 9.9 seconds.
| John cant run a faster 100m.
According to Fauconnier and Israel, the opposing, scalar inferences of (3b) and
(4b) are pragmatic entailments. They are entailments because the inferences are
default, i.e. automatically valid; they are pragmatic because the inferences can be
defeated. But how the inferences can be cancelled seems to me to be unclear. One
type of pragmatic entailment of this sort is scalar entailment stemming from what
Levinson (2000) called a Hirschberg-scale a contextually given ad hoc scale. The
relation between the ordered elements in such a scale does not need to be that of
semantic entailment. The scale can be based on any partially ordered contrast sets
in a contextually salient way. Examples include destinations ordered in terms of
their distance, autographs ordered in terms of their authors famousness, and problems ordered in terms of their complexity (see e.g. Huang 2009d for further discussion).
Semantic entailment plays a prominent role in semantics and especially in formal semantics. It is explored to define many of the other fundamental semantic relations. For example, two sentences may be said to have a propositional equivalence, to be paraphrases of each other, or to be synonymous if and only if they have
exactly the same set of entailments, as in (5). Next, two sentences may be con-
399
sidered to be contradictories if each entails the negation of the other so that whenever one is true the other must be false, as in (6). In a similar vein, a single sentence
may be said to be a contradiction if it contains contradictory entailments, as in (7).
While contradiction is a semantic relation in which one member of the pair must be
true and the other false, contrariety is a semantic relation in which only one term
may be true, though both terms may be simultaneously false. Consequently, two
sentences S1 and S2 may be taken to have contrary propositions if and only if S1 entails the negation of S2 and the negation of S2 does not entail S1. This is exemplified
in (8). Furthermore, semantic entailment is also used to define notions like analyticity and ambiguity.
(5) (Propositional equivalence)
a. The little hut was hidden by the trees.
b. The little hut was concealed by the trees.
(6) (Contradictory)
a. No one likes dark tourism.
b. At least someone likes dark tourism.
(7) (Contradiction)
?John isnt married, but his wife is a feminist.
(8) (Contrariety)
a. This skirt is blue.
b. This skirt is red.
In formal semantics, semantic entailment is employed to advance a theory of generalised quantifiers. In this theory, which is developed out of the work of Ladusaw
(1979), the concept of directional entailingness (also called monotonicity) is invoked. Directional entailingness provides an account of patterns of semantic entailment between sets and subsets. Upward entailing or entailment (also called
monotone increasing, usually represented as MONF) involves entailment from a
subset to a set. In other words, the direction of entailment is from more specific to
less specific. By contrast, downward entailing or entailment (also called monotone
decreasing, normally represented as MONG) is concerned with entailment from a
set to a subset. Another way of saying it is that the direction of downward entailment is from less specific to more specific. Now, contrast (9) and (10).
(9) a.
b.
(10) a.
b.
Every woman is cooking.

Every woman is roasting beef.
No woman is cooking.
No woman is roasting beef.
In (9), (9a) does not entail (9b), but (9b) does entail (9a). Consequently, the quantifier every triggers an upward entailment here.4 By contrast, in (10) the reverse is
true. Therefore, the quantifier no involves a downward entailment. The distinction
between upward and downward entailment is applied particularly to the study of
400
Yan Huang
negative polarity items (NPIs). It predicts correctly that NPIs like any and ever can
be licensed in the scope of a downward entailment operator but not in that of an upward entailment one (see e.g. Saeed 2009). The contrast is shown in (11).
(11) a. ?Every woman is ever roasting beef.
b. No woman is ever roasting beef.
But there has also been evidence that downward entailment is neither necessary
nor sufficient for the licensing of NPIs (e.g. Giannakidou 1998, Horn 2009, see
also Atlas 2007).
More recently, Chierchia, Crain and their associates (e.g. Chierchia 2004,
Crain and Pietroski 2002) have used the upward/downward entailment distinction
to mount a challenge to what Horn (2006) dubbed a Golden Age Pure Pragmatics
or GAPP-style analysis of scalar implicatures. According to Chierchia, Crain, and
their associates, while a standard upper bounding Q-scalar implicature, that is, an implicature from p to at most p , does arise from positive Horn-scales,
it is quite weak and even blocked in negative Horn-scales and other downward entailing environments. On the basis of this claim, Chierchia, Crain, and their associates argued that Q-scalar implicatures must be computed compositionally. Consequently, they fall under compositional semantics, hence part of grammar. But as
pointed out by Levinson (2000) and Horn (2006, 2009), the alleged blockage of
Q-scalar implicatures is due to the fact that a Horn-scale is reversed under negation
and other downward entailing operators, as in (12a), and consequently, a different
Q-scalar implicature is derived from the inverse scale, as in (12c) (see also Huang
2009b). (I use +> to signify conversationally implicate.) If this is the case, then
Chierchia and Crains argument may not be maintained (see also Sauerland 2004,
Russell 2006, and Geurts 2009 for various (neo)-Gricean analyses of Q-scalar implicature in downward entailment contexts).
(12) a. <not some, not many, not most, not all>
b. The earthquake didnt kill many of the villagers.
c. +> The earthquake killed some of the villagers.
There are, however, serious problems at the very heart of the (semantic) entailment
based approach to meaning. One such problem, which is well-known, is concerned
with the failure of a semantic entailment in a referentially opaque context. Consider first (13).
(13) a. Barack Obama is a popular politician.
b. The first African-American President of the United States of America is a
popular politician.
At the time of writing (April, 2009), the proper name Barack Obama in (13a) and
the definite description the first African-American President of the United States of
America in (13b) have the same extension, i.e. they refer to the same person. Given
401
Leibnizs Law, which is a law of intersubstitutability salva veritate (Latin for with
the truth unchanged), it states that the substitution of expressions with the same
extension does not affect the truth-conditions of the sentences. Consequently, if
(13a) expresses a true proposition, then so does (13b). In other words, (13a) and
(13b) are synonymous because they mutually entail each other. However, in a referentially opaque context, truth cannot be preserved when the co-referential expressions are substituted for each other. This is the case in (14).
(14) a. I want to dine with Barack Obama.
b. I want to dine with the first African-American President of the United
States of America.
Here, (14a) and (14b) have different truth conditions. This is because the expression the first African-American President of the United States of America is
given an intensional reading, that is, the reading that the speaker wants to dine with
whoever happens to be the first African-American President of the United States of
American and does not care and may not have known who that person is. The same
can be said of (15).
(15) a. John believes that Barack Obama is a popular politician.
b. John believes that the first African-American President of the United
States of America is a popular politician.
If John does not know or believe that Barack Obama is the first African-American
President of the United States of America, then (15b) means that John has a belief
about someone who is the first African-American President of the United States of
America without knowing or caring about who the actual holder of the office is.
Consequently, such a belief of Johns is said to be de dicto a belief about what is
said. The same is true of Freges classical example, usually referred to as the
Morning Star Paradox (see e.g. Cann 1993 for discussion about further problems of
the (semantic) entailment based analysis). Verbs that create a referentially opaque
context are commonly referred to as intensional verbs or verbs of propositional attitude. These include believe, know, doubt, want, hope, fear, look, seem, and seek.
Other words that establish a referential opaque context include adjectives like alleged, prepositions like about, connectives like because, and modal words of various categories like must, may, probably, obviously, and permissible.
2.
Presupposition
2.1.
Defining presupposition
Presupposition is defined as a proposition or inference whose truth is taken for

granted in the utterance of a sentence.5 Its main function is to act as a precondition
402
Yan Huang
of some sort for the appropriate use of that sentence. This background assumption
will remain in force when the sentence that contains it is negated. The German
mathematician and logician Gottlob Frege is generally recognised as the first
scholar in modern times who (re)introduced the philosophical study of presupposition (see especially Frege 1892), though the notion of presupposition may go back
at least as far as the mediaeval philosopher Petrus Hispanus (see e.g. Huang 2007).
Presupposition is usually engendered by the use of particular lexical items
and/or linguistic constructions. Lexical items and linguistic constructions that give
rise to presuppositions are called presupposition triggers. While presupposition
has in general been regarded as a heterogeneous collection of quite distinct and
different phenomena (Levinson 1983: 217) since Karttunen (1973), a list of its
representative examples may be given in (16) (23) (see e.g. Levinson 1983, Atlas
2005, Huang 2007). (I use the symbol >> to stand for presuppose. The positive
and negative versions of the examples are separated by /, and the lexical presupposition triggers are italicised.)
(16) Definite descriptions
The king of France is/isnt bald.
>> There is a king of France.
(17) Factive predicates
a. Epistemic or cognitive factives.
John knows/doesnt know that smoking is a dangerous pastime.
>> Smoking is a dangerous pastime.
b. Emotive factives
John regrets/doesnt regret that he has gone on the stage.
>> John has gone on the stage.
(18) Aspectual/change of state predicates
Mary has/hasnt stopped beating her boyfriend.
>> Mary has been beating her boyfriend.
(19) Iteratives
John returned/didnt return to Berlin.
>> John was in Berlin before.
(20) Implicative predicates
John managed/didnt manage to give up binge drinking.
>> John tried to give up binge drinking.
(21) Temporal clauses
After she left school, Jane worked/didnt work as a secretary.
>> Jane left school.
(22) Cleft sentences
a. Cleft
It was/wasnt the porter who called a taxi for John.
>> Someone called a taxi for John.
403
b. Pseudo-cleft
What the porter called/didnt call was a taxi.
>> The porter called something.
(23) Counterfactual conditionals
If an ant was as big as a human being, it could/couldnt run five times faster
than an Olympic sprinter.
>> An ant is not as big as a human being.
2.2.
Properties of presupposition
Presuppositions exhibit two main properties: (i) constancy under negation, and (ii)
defeasibility. In addition, certain cases of defeasibility give rise to what is known
as the projection problem of presupposition.
2.2.1.
Constancy under negation
By constancy under negation is meant that a presupposition generated by the use of

a lexical item or a syntactic structure remains the same when the sentence containing that lexical item or syntactic structure is negated (e.g. Strawson 1952). Using
constancy under negation as diagnostic, presuppositions can be more formally defined in (24).
(24) An utterance of a sentence S presupposes a proposition p if and only if
a. if S is true, then p is true;
b. if S is false, then p is (still) true.
What (24) basically says is this: for S to presuppose p, whenever S is true, p is also
true, and whenever S is false, p is still true.
There are, however, problems with constancy under negation. On the one hand,
constancy under negation may not be necessary. For example, there is a class of
sentences which are hard, if not impossible to negate, yet they bear presuppositions, as in (25). On the other hand, constancy under negation may not be sufficient. This is illustrated by (26). Although (26) satisfies constancy under negation,
the inference is standardly analysed as a felicity condition on performing the
speech act of requesting (e.g. Levinson 1983). This is because if the window is
open, there is no need for one to perform the speech act of requesting someone else,
for example, to open it.
(25) Long live the king of France!
>> There is a king of France
(26) Do/dont open the window.
?>> The window is not open
404
2.2.2.
Yan Huang
Defeasibility
Presuppositions are deafeasible. They are cancelled if they are inconsistent with (i)
background assumptions, (ii) conversational implicatures, and (iii) certain discourse contexts. Furthermore, they can also disappear in certain intrasentential
contexts, some of which give rise to the projection problem of presupposition. Defeasibility has in general been taken as the second most important property of presupposition.
In the first place, presuppositions can disappear in the face of inconsistency
with background assumptions or real-world knowledge. Contrast (27) and (28) (I
use ~ >> to stand for does not presuppose.)
(27) John got an assistant professorship before he finished his Ph.D.
>> John finished his Ph.D.
(28) John died before he finished his Ph.D.
~ >> John finished his Ph.D.
While (27) presupposes that John finished his Ph.D. by virtue of the temporal
clause, (28) does not carry that presupposition. This is because the putative presupposition conflicts with our real-world knowledge. Consequently, the unwanted presupposition vanishes.
Secondly, presuppositions can be cancelled by inconsistent conversational implicatures. By way of illustration, consider (29).
(29) If Susan has danced on the table, her mother will be unhappy that she has
done so.
+> perhaps Susan has danced on the table, perhaps she hasnt
~ >> Susan has danced on the table
The use of the factive predicate unhappy in (29) should give rise to the potential
presupposition that Susan has danced on the table. However, there is also a Q-clausal conversational implicature due to the use of the conditional in (29), namely,
perhaps Susan has not danced on the table. In the face of the contradictory conversational implicature, the putative presupposition is defeated.
Thirdly, presuppositions are contextually cancellable, that is, they can evaporate if they run contrary to what the immediate discourse context tells us. Witness
first (30), where the putative presupposition is defeated by the inconsistent proposition that has already been established in the immediate discourse background.
(30) There is no king of France. Therefore, the king of France isnt bald.
~ >> There is a king of France
In (30), the second sentence should presuppose that there is a king of France. The
reason it does not is because such a putative presupposition is inconsistent with the
preceding proposition that has already been established in the immediate discourse
405
background. As a result, the unwanted presupposition fails to survive. Next, presuppositions can be suspended by so-called reduction arguments arguments that
proceed by eliminating each of the possibilities in a discourse. Finally, presuppositions can disappear in a discourse where evidence for their truth is being
weighed and rejected.
In the fourth place, presuppositions can be blocked in certain intrasentential
contexts, that is, they can be defeated by using other clauses in the same complex
sentence to increment the local, intrasentential context. Three sub-cases are of particular interest. The first is that a presupposition of a sentence can be overtly denied
in a co-ordinate clause without any apparent contradiction. This is exemplified in
(31).
(31) John doesnt regret snoring atrociously because in fact he never did so!
~ >> John snored atrociously
Notice that in many cases, outright denial of presupposition is not possible with
positive sentences.
(32)*John regrets snoring atrociously because in fact he never did so!
There is thus, at least in these cases, an asymmetry between negative and positive
sentences with regard to defeasibility. This asymmetry has led to an entailment
analysis for positive sentences, namely, the argument that what is allegedly presupposed in these sentences is actually what is entailed (e.g. Atlas 2005). Since semantic entailments cannot be overtly denied without producing semantic anomaly,
the anomaly displayed in (32) is entirely expected. A second point to note is that
the negation involved in overt denial of presupposition in (31) is generally taken as
a metalinguistic negation (see e.g. Horn 2006). Next, a presupposition of a sentence can be explicitly suspended in an if clause that follows. Witness (33).
(33) Mary doesnt regret being a pet sitter, if she actually ever was one.
~ >> Mary was a pet sitter
Finally, presuppositions can disappear with certain verbs of saying such as say,
mention, tell, ask and announce, and certain verbs of propositional attitude such as
believe, think, imagine, dream and want (see e.g. Huang 2007: 7273).
2.2.3.
Theoretical issues and analyses
In general, there are three main theoretical issues in accounting for presupposition.
The first and oldest one is concerned with presupposition failure. If a statement
presupposes something which does not exist, what are the consequences for that
statement? As already mentioned, a much quoted example of statements of this
kind is the positive version of (16). Given that there is no king of France, the question boils down to this: what is the truth status of the sentence? Russells (1905)
406
Yan Huang
view was that since the sentence asserts both that there is a king of France and that
he is bald, if the king of France does not exist, the sentence is false. This analysis
was widely accepted and remained practically unchallenged for almost half a century. Then in 1950, Strawson published a seminal paper (Strawson 1950) in which
he presented an alternative account. According to this analysis, Russell failed to
distinguish sentences from the use of sentences. In using the sentence, a speaker
does not assert but merely presupposes that there is a king of France. If there is no
king of France, then there is a presupposition failure. Consequently, the sentence is
neither true nor false: there is simply a truth-value gap. Strawsons account has led
to the notion of semantic presupposition.
Secondly, there is the question of the origin of presupposition, namely, where
presuppositions come from. The standard, though imperfect answer to this question is that presuppositions are engendered by certain lexical and structural
triggers. Thirdly and finally, the most important issue is concerned with the projection problem of presupposition a special case of the Fregean compositionality,
namely, to state and explain the presuppositions of complex sentences (as
wholes) in terms of the presuppositions of their component simple sentences (as
parts). The projection problem manifests itself in two opposite directions. On the
one hand, the presuppositions of a component sentence may fail to be projected
onto, and hence inherited by, the whole complex sentence. This is the case for e.g.
(31) and (33). In addition, certain complex or compound sentences formed with
if then and either or also belong to this category. On the other hand, the presuppositions of a component sentence may be preserved when that constituent sentence becomes part of a more complex sentence. This is what happens when the
sentence containing a presupposition trigger is under negation, as in (16), when it
is embedded in modal operators such as its possible that, its likely that, ought to,
may be and should, and when it is embedded in the conditional and the disjunction
formed with if then and either or.
Three formal semantic-pragmatic models are particularly influential in tackling
the projection problem. The first is what I called the filtering-satisfaction (or Karttunen-Stalnaker) analysis in Huang (2007). The central idea of this approach is that
a presupposition is not cancellable. Rather it has to be satisfied (i.e. entailed) in its
local context, which is conceived mainly as a set of propositions. The local context
is constructed in a dynamic way and developing online. This makes it possible for
an unwanted presupposition to be filtered out during the derivation of a sentence in
a bottom-up manner (see e.g. Karttunen 1973, Stalnaker 1974). Stalnakers work
has resulted in the development of the concept of pragmatic presupposition. Next,
in contrast to the filtering-satisfaction analysis, underlying Gazdars (1979) cancellation analysis is the crucial assumption that a presupposition is cancellable. On
this theory, what a presupposition trigger engenders is merely a potential presupposition. A potential presupposition will become an actual one, unless it is defeated. With respect to the projection problem, what the cancellation analysis pre-
407
dicts is that each and every presupposition of the embedded clause will become an
actual presupposition of the complex sentence, unless it is nullified by certain linguistic and non-linguistic factors. Finally, in an attempt to combine the strengths of
both the filtering-satisfaction and cancellation theories, Heim (1983) developed a
particular version of the filtering-satisfaction model, couched in her dynamic semantic theory of context change. In Huang (2007), I dubbed this model the accommodation analysis. Central to this approach is Heims belief that the meaning of an
expression, including the presupposition of a sentence, is its context change potential. The context change potential of a sentence is the intersection of any context set
in a common ground together with the proposition expressed by the sentence. By
common ground is meant a background of beliefs or assumptions which are
shared by the speaker and his audience and which are recognised by them to be so
shared (Stalnaker 1974). Accommodation, then, provides a mechanism to increment the discourse context set with new, non-controversial assumptions.
Of these analyses, Karttunens and Gazdars are representatives of the twocomponent model view of presupposition, according to which, sentences including
those containing presuppositions have two types of content: semantic content and
presuppositional content. The latter is pragmatic in character. By contrast, Stalnaker and Heim take the one-component model view, operating within the general
framework of a theory of pragmatic presupposition. On this account, presuppositions are treated as conditions on whether a sentence can be admitted into a context
(see e.g. Portner 2005).
3.
Implicature
The concept of implicature (both conversational and conventional) has its origin in
the work of H. P. Grice, though some proto-Gricean ideas can be traced back at
least to the first century B.C. rhetorician Dionysius and the fourth century rhetoricians Servius and Donatus (e.g. de Jonge 2001). The ideas were later reiterated by
the nineteenth century English philosophers John Stuart Mill and Augustus De
Morgan (e.g. Horn 2006).
3.1.
Conversational implicature
A conversational implicature is any meaning implied or expressed by, and inferred

or understood from, the utterance of a sentence which is meant without being
part of what is strictly said6 (e.g. Huang 2009a, b, f). It is derived via Grices
(1975, 1989) cooperative principle and its attendant maxims of conversation. For
example, when one utters the sentence in (34a), he or she (ceteris paribus) conversationally implicates (34b).
408
Yan Huang
(34) a. Some of the tourists are admiring the view.

b. +> Not many/most/all of the tourists are admiring the view
Conversational implicatures are characterised by a number of distinctive properties (Grice 1975, 1989, Levinson 2000, Huang 2007: 3235, 2009a, b). In the first
place, there is the property of defeasibility conversational implicatures can
simply vanish in certain linguistic or non-linguistic contexts. How? They are cancelled if they are inconsistent with (i) semantic entailments, (ii) background or ontological assumptions, (iii) contexts, or (iv) priority conversational implicatures.
As an illustrating example, consider (35) and (36). (The symbol ~ +> is used to
stand for do not conversationally implicate.)
(35) John and Mary bought a villa in Auckland.
+> John and Mary bought a villa in Auckland together
(36) The Americans and the Russians tested an atom bomb in 1962.
~ +> The Americans and the Russians tested an atom bomb in 1962 together
The utterance in (36) has the potential mirror maxim, togetherness conversational implicature, as indicated in (35). However, this potential conversational implicature runs contrary to our background or ontological assumptions. Given our
knowledge about history, it was impossible for the USA and the USSR to test an
atom bomb together in 1962. Consequently, the potential togetherness conversational implicature is defeated by inconsistent real-world knowledge. Notice that
defeasibility is a necessary but not a sufficient condition for conversational implicature (Horn 2007).
A second property exhibited by conversational implicatures is non-detachability any linguistic expression with the same semantic content tends to carry
the same conversational implicature. (A principled exception is those conversational implicatures that arise via the maxim of Manner.) This is because conversational implicatures are attached to the semantic content, rather than the
linguistic form, of what is said. Therefore, they cannot be detached from an utterance simply by replacing the relevant linguistic expressions with their synonyms.
This is illustrated in (37), which indicates that the use of any linguistic expression
that is synonymous with almost will give rise to the same conversational implicature.
(37) Gordon almost/nearly lost /came close to losing his job.
+> Gordon did not (quite) lose his job
Thirdly, we have the property of calculability conversational implicatures can
transparently be derived via the cooperative principle and its component maxims.
A fourth property is non-conventionality conversational implicatures, though dependent on the saying of what is coded, are non-coded in nature. In other words,
they rely on the saying of what is said but they are not part of what is said. A fifth
409
property is reinforceability conversational implicatures can be made explicit

without producing too much of a sense of redundancy. This is because conversational implicatures are not part of the conventional import of an utterance. For instance, the conversational implicature in (34) is made explicit in (38). But (38) is
not judged to be semantically redundant.
(38) Some, but not many/most/all, of the tourists are admiring the view.
The sixth property of conversational implicatures is universality conversational
implicatures tend to be universal, because they are motivated rather than arbitrary.
For example, if a language has all and some, the use of the semantically weaker
some will universally carry the conversational implicature not all. In Huang
(2007: 35), data are drawn from Modern Standard Arabic, Catalan, Chinese, Modern Greek, Kashmiri and Malagasy to illustrate this point. Finally, there is the property of indeterminacy some conversational implicatures may be indeterminate.
They can be taken as conveying an open-ended range of implicaures relating to
matters in hand. Suppose one says (39). He or she may generate a range of indeterminate conversational implicatures.
(39) Our new professor is a machine.
+> Our new professor is cold, or/and
+> Our new professor is efficient, or/and
+> Our new professor is a workaholic, or/and
+>
A conversational implicature can be engendered in two distinct ways. On the one
hand, it can arise from strictly observing the maxims of conversation. In Huang
(2007: 27), I called conversational implicatures thus induced conversational implicaturesO. This is the case for e.g. (34), (35) and (37) above. On the other hand, a
conversational implicature can be generated by way of a speakers ostentatiously
flouting the maxims. In Huang (2007: 29), I dubbed conversational implicatures
thus engendered conversational implicaturesF. This is the case with the generation
of the conversational implicature in (39), which deliberately exploits Grices
maxim of Quality. There is thus the first Gricean dichotomy between conversational implicatureO and conversational implicatureF. Grices achievement here was
to have provided a unified analysis of both types.
A second Gricean dichotomy, independent of the first, is between those conversational implicatures which arise without requiring any particular contextual conditions and those which do require such conditions. Grice called the first kind generalised conversational implicatures (GCIs) and the second kind particularised
conversational implicatures (PCIs). By way of illustration, consider the two conversational implicatures in Marys utterance in (40).
410
Yan Huang
(40) John: How did yesterdays guest lecture go?

Mary: Some of the faculty left before it ended.
+> (a) Not many/most/all of the faculty left before the lecture ended
+> (b) The lecture didnt go well
The conversational implicature in (40a) has a very general currency. Any utterance
of the form Some x are Y will have the default interpretation Not many/most/all
x are Y. This interpretation will go through without needing any particular context, hence (40a) is a GCI. By contrast, the conversational implicature in (40b) depends crucially on context of some kind. Marys reply points to a possible connection, namely, if some of the faculty left a lecture before it ended, the lecture may
not have gone well. Without such a specific connection, we will not have the relevant conversational implicature, thus (40b) is a PCI. The theoretical importance
of this Gricean dichotomy has recently been subject to heated debates. Hirschberg
(1991), Welker (1994) and Carston (2002), for example, doubted whether such a
distinction can be maintained. On the other hand, Levinson (2000) put forward a
rigorous defence of it.
Since its inception, Grices classical theory of conversational implicature has
revolutionised pragmatic theorising, nourishing numerous neo- and post-Gricean
variants and reformulations. Horn developed a bipartite model calling it Manichaean pragmatics. In this account, there are two genera of conversational implicatures: namely Q[uantity]- and R[elation]-implicatures (see e.g. Horn 2009). Arguing for a clear separation between pragmatic principles governing an utterances
surface form and pragmatic principles governing its informational content, Levinson (2000) proposed that the original Gricean program (the maxim of Quality
apart) be reduced to a tripartite model with three genera of conversational implicatures: what he dubbed the Q-, I[nformativeness]- and M[anner] implicatures.
Q-implicatures can then be divided into three species: what I termed (i) Q-scalar , (ii)
Q-clausal and (iii) Q-alternate implicatures in Huang (2007: 4244). Furthermore, within
Q-alternate implicatures, we have two subtypes: Q-ordered alternate and Q-unordered alternate implicatures such as <bake, boil, grill, steam, stir-fry>. Notice that Q-ordered alternate implicatures can arise either from a non-entailment scale which is given by the lexicon without requiring any specific context (e.g. <succeed, try>) or from what
Levinson (2000) called a Hirschberg-scale or what Horn (2007, 2009) dubbed a
rank order, following in part work by Lehrer (1974). As already mentioned above,
a Hirschberg-scale is essentially a nonce scale, that is, a contextually given ad hoc
scale. Such a scale can be based on any partially ordered contrast sets in a contextually salient way. Examples include <divorce, separate>, <Barack Obamas autograph, Hillary Clintons autograph>, and <full professor, associate professor, assistant professor> (see e.g. Huang 2009d for further discussion). On the other
hand, in relevance theory (Sperber and Wilson 1995), the majority of classical and
neo-Gricean GCIs is refashioned as an explicature a proposition that is an infer-
411
ential development of one of the linguistically given incomplete conceptual representations or logical forms of the sentence uttered (but see e.g. Burton-Roberts
2007 for the comment that development in the definition is not defined in relevance theory). Explicature corresponds roughly to the intuitive notion of what is
said, though the Gricean notion of what is said is abandoned in relevance theory.
Conversational implicatures in the relevance-theoretic sense, called r-implicatures
in Huang (2007: 195), are largely PCIs in the classical and neo-Gricean sense. In a
similar vein, Recanati (2004), a radical contextualist in the current divide between
contextualism and (semantic) minimalism in the philosophy of language, reduced
some cases of conversational implicature to what he called the pragmatically enriched said. Also, Bach (2004) argued that certain aspects of speaker meaning are
neither part of what is said nor of what is conversationally implicated. Consequently, he proposed a third category of communicative content, intermediate between what is said and what is conversationally implicated, and dubbed the vehicle
of such a content conversational impliciture, because it is what is implicit in what
is said. In other words, on Bachs view, some cases of conversational implicature in
the classical and neo-Gricean sense are in fact (conversational) implicitures (see
also Bach 2006, Huang 2009e and Garret and Harnish 2007).
At this point, it is useful to discuss some of the current debates on conversational implicature. In the first place, the treatment of cardinals as scalar expressions
engendering Q-scalar implicatures is highly controversial (see e.g. Carston 2002,
Bultinick 2005, Hurewitz et al. 2006, Horn 2009 for arguments against, and Levinson 2000 for arguments for the scalar analysis). Secondly, there is the issue of the
epistemic strength of Q-scalar implicatures. It is concerned with the question of
what it is a speaker Q-scalar implicates against. Two neo-Gricean pragmatic positions can be identified here: the weak epistemic one represented by e.g. Hirschberg
(1991), Sauerland (2004), Geurts (2009), and Horn (2009), and the strong epistemic one advocated by e.g. Gazdar (1979) and Levinson (2000: 7779). In the
third place, a heated debate has been going on for the last two decades or so, focusing on the nature of Gricean and neo-Gricean GCIs in general and Q-scalar implicatures in particular. One view is that GCIs in general and Q-scalar implicatures
in particular convey default meanings, sans a conscious inferential process and irrespective of a particular context (e.g. see Levinson 2000 for strong defaultism and
Horn 2009 for weak defaultism). Another, relevance-theoretic position is that they
are essentially inferred contextually (e.g. Sperber and Wilson 1995). In addition, a
more recent, third view holds that the derivation of Q-scalar implicatures relies
heavily on structural factors (cf. Section 2 above and Chierchia 2004). Furthermore, all the three views have recently been subject to studies in experimental
pragmatics. While much of the relevance-theoretically oriented experimental work
favours the contextual inference approach (e.g. Noveck and Sperber 2007, Noveck
and Reboul 2008), there is also evidence in support of the default inference theory
(e.g. Grodner et al. 2007) and the structural inference view (e.g. Panizza and Chier-
412
Yan Huang
chia 2008). Next, in recent years, there has also been an intense debate on (i)
whether or not pragmatically enriched or inferred content can enter or intrude
upon the conventional, truth-conditional content of what is said, and (ii) if the
answer to (i) is positive, then what the pragmatic intrusion under consideration is.
Concerning the first question, a dividing line can generally be drawn between pragmaticists like Levinson (2000), Recanati (2004) and relevance theorists (Sperber
and Wilson 1995, Carston 2002), who argue for pragmatic intrusion (but see e.g.
Horn 2009 for reservations), and semanticists like Cappelen and Lepore (2005)
and King and Stanley (2005), who argue against it. Next, regarding the second
question, two current positions can roughly be identified. The first is that pragmatic
intrusion is of a special kind, which differs from conversational implicature.
Within this camp, three lines of argument are of particular interest. According to
Sperber and Wilson (1995), the pragmatic content is an explicature. Secondly,
there is the position taken by Recanati (2004) that it is a pragmatically enriched
part of what is said. A third argument is due to Bach (2004), who took the view that
the pragmatic inference under consideration is his impliciture. On the other hand,
the second position is represented by Levinson (2000). Within the neo-Gricean
framework, Levinson argued that these so-called explicatures/pragmatically enriched said/implicitures are in fact the same beast as conversational implicatures.
See also my neo-Gricean and revised neo-Gricean analyses of anaphora in e.g.
Huang (1991, 2000a, b, 2004, 2007 and 2009b), which in effect argue that the presemantic, pragmatic inference involved in the interpretation of anaphora is a conversational implicature.
Finally, it is worth mentioning that there have been various attempts to integrate the classical and neo-Gricean pragmatic theories of conversational implicature with other current linguistic theories. These linguistic theories include decision theory (Merin 1999), bidirectional optimality theory (Blutner 2004) and game
theory (Benz et al. 2006).
3.2.
Conventional implicature
We turn next to the second category of implicature postulated by Grice, namely,

conventional implicature. (In fact, Freges 1892, 19181919 analysis of the Andeutung relation is a direct precursor of Grices concept of conventional implicature.
See e.g. Bach 1999, Feng 2006, Horn 2007). An Andeutung or conventional implicature is a non-truth-conditional meaning which is not derivable from general considerations of cooperation and rationality, but arises solely from the conventional
features attached to particular lexical items and/or linguistic constructions involved. A few standard examples are given in (41)-(43). (I use +>> to stand for
conventionally implicate.)
413
(41) p therefore q +>> q follows from p

Lihua is a Chinese; she, therefore, knows how to use chopsticks.
(42) p but q +>> p contrasts with q
John is poor but he is honest.
(43) Even p +>> contrary to expectation
Even a child can scoff all the small bananas.
In (41), the conventional implicature triggered by the use of therefore is that being
a Chinese provides some good reason for knowing how to use chopsticks. In (42),
there is a conventional implicature of contrast between the proposition contained
in p and that contained in q. Finally in (43), even, being epistemic in nature, conventionally implicates some sort of unexpectedness, surprise or unlikeliness (Farncescotti 1995). Other representative lexical items that are considered to engender
conventional implicatures in English include actually, also, anyway, barely, besides, however, manage to, moreover, on the other hand, only, still, so, though, too
and yet.
What, then, are the essential properties of conventional implicature? From a
traditional point of view, conventional implicatures are considered to have the following properties (Grice 1989, Levinson 1983, Huang 2007, 2009c). Firstly, conventional implicatures are not derived from Grices cooperative principle and its
component maxims, but are attached by convention to particular lexical items
and/or linguistic constructions. They are therefore an arbitrary part of meaning,
and must be learned ad hoc. Secondly, conventional implicatures are not calculable
via any natural procedure, but are rather given by convention, thus they must be
stipulated. Thirdly, conventional implicatures are not defeasible, that is, they cannot be cancelled. Fourthly, conventional implicatures are detachable, because they
depend on the particular lexical expressions and/or linguistic constructions used.
Finally, conventional implicatures do not tend to be universal.
It should be pointed out that unlike the concept of conversational implicature,
the notion of conventional implicature is not taken to be very coherent. Even Grice
himself (1989: 46) warned that the nature of conventional implicature needs to be
examined before any free use of it, for explanatory purposes, can be indulged in.
Horn (2004: 6) has gone a step further by claiming that the role played by conventional implicature within the general theory of meaning is increasingly shaky.
Since its inception, conventional implicature has been subject to numerous attempts
to reduce it to semantic entailment, conversational implicature, and presupposition
(Levinson 1983), and more recently, to part of what is said (Bach 1999), part of tacit
performatives (Rieber 1997), vehicles for performing second-order speech acts
(Bach 1999), and procedural meaning in relevance theory (Blakemore 2004).
But recently, Potts (2005) has made a brave attempt to resurrect the concept of
conventional implicature. He retain[ed] Grices brand name but alter[ed] the product (Horn 2007) by focusing on expressives like epithets, attributive adjectives
414
Yan Huang
and honorifics and supplements like non-restrictive relatives, parentheticals and

appositives rather than lexical items such as but, therefore and even. He isolated
four essential properties of conventional implicature. The first of these properties
is conventionality conventional implicatures are part of the conventional meaning of the expressions involved. The second property is commitment conventional implicatures are commitments, and thus engenders entailments. The third
property is speaker orientation the commitments are made by the speaker of an
utterance. The final property is independence conventional implicatures are logically and compositionally independent of what is said (see also von Heusinger and
Turner 2006). Taking the view that conventional implicature is semantic in nature,
Potts developed a logic of the notion by modelling it with a type-driven multi-dimensional semantic translation language (see Feng 2006, Horn 2007 for criticisms
of this analysis). Feng (2006) presented another development of Grices notion of
conventional implicature. The properties extracted by him for conventional implicature are (i) non-truth-conditionality, (ii) speaker orientation, (iii) infallibility,
(iv) occurrency, (v) dependency, and (vi) context-sensitivity. He further argued that
properties (i) (iv) are intimately associated with subjectivity. Finally, contrary to
Pottss view, Feng maintained that conventional implicature has both a semantic
and pragmatic character. This is why a conventional implicature is so named by
Grice. [I]t is so named because it involves both linguistic and contextual information. It is conventional because it is associated with the conventional linguistic meaning of a certain expression It is implicated rather than said because
its full content requires contextual information, and does not affect the truth conditions of the utterance (Feng 2006: 184). This view is also echoed in Horn (2007:
50), who said that [conventional implicature] is semantic insofar as it involves an
aspect of the conventional meaning of a given expression rather than being computable from general principles of rational behavior or communicative competence, but it is pragmatic insofar as it involves considerations of appropriateness
rather than truth of the sentence in which it appears. Whether belonging to semantics or balancing on the edge between semantics and pragmatics, Pottss, Fengs
and Horns recent works have shown that the Fregeo-Gricean concept of conventional implicature is, after all, not that incoherent.
4.
A brief comparison
In this final section, I shall make a brief comparison among semantic entailment,
presupposition, conversational implicature and conventional implicature. I shall
compare them in terms of four parameters: (i) defeasiblility, (ii) negation, (iii) conventionality, and (iv) semantics versus pragmatics.
First, defeasibility. While semantic entailment and conventional implicature
are not defeasible, most cases of presupposition and all cases of conversational im-
415
plicature are defeasible. In the study of inference, a hierarchy is set up to rank various types of inference as defeasible, nonmonotonic, default, probabilistic, and
plausibilistic, with defeasible inference at the top. An inference is defeasible when
it has the possibility of error. Looked at in this way, nonmontonic inference is a
major type of defeasible inference. It is currently reasonable but its reasonability
can be cancelled upon the admittance of new information into common ground.
Default inference is then a particular kind of nonmonotonic inference. It is characterised by what computer scientists and artificial intelligence (AI) workers call negation as failure. An example might be that from the failure of the airport departures board to list a late-night flight to London Heathrow, one may infer that such a
flight has been cancelled. In this sense, both presupposition and conversational implicature are a case of presumptive reasoning like default inference. Probabilistic
inference is stronger than plausibilistic inference (see e.g. Woods 2009). Next, we
turn to negation. While presupposition can survive negation, entailment cannot. As
for conversational and conventional implicatures, their behaviour under negation
is rather complex and needs to be further studied. Finally, let us consider the question of conventionality and the related question of where in the linguistic terrain
the four types of inference are situated. Clearly, semantic entailment is the most
conventional and falls in the province of semantics. Presupposition is less conventional than conventional implicature but more conventional than conversational
implicature. Therefore, it straddles the boundary between semantics and pragmatics, but more on the side of pragmatics. Of the three types of implicature, although GCIs are more conventional than PCIs, both types are the pursuit of pragmatics. Finally, conventional implicature, which is more conventional than both
GCIs and PCIs, can be considered either as semantic or pragmatic, depending on
how the boundary between semantics and pragmatics is placed. If semantics is
taken to be concerned with those aspects of meaning that affect truth conditions,
then the investigation of conventional implicature falls on the pragmatic side of the
divide rather than on the semantic side, since, as noted above, it does not make any
contribution to truth conditions. On the other hand, if pragmatics is conceived of as
dealing with those inferences that are non-conventional, hence cancellable, then
conventional implicature falls within the province of semantics but outside that of
pragmatics, since it cannot be defeated.
Notes
1. I wish to thank the anonymous reviewers and Wolfram Bublitz and Neal Norrick, the two
editors of this volume for their stimulating comments on an earlier version of this chapter.
The work reported on here was partially supported by a grant from the University of
Auckland Faculty Research Development Funds, which is hereby gratefully acknowledged.
416
Yan Huang
2. On Saeeds (2009) view, entailment is not an inference. However, Cummings (2005:

8384) argued convincingly that such a view is too narrow.
3. The phrase p entails q appeared first in Moore (1922). The term entailment, according
to the second edition of the Oxford English Dictionary, was first used by L. S. Stebbing in
1933, who wrote: Professor Moores analysis of the distinction between material implication and entailment makes it possible.
4. Notice that whereas every is upward entailing with respect to its right, VP argument, as in
(9), it is downward entailing in its left, NP argument (e.g. Saeed 2009). For example,
Every woman is cooking unilaterally entails Every old woman is cooking. In terms of
monotone increasing/decreasing, the former is called right monotone increasing, and the
latter is labelled left monotone decreasing.
5 The term presupposition seems to be introduced into English by Strawson (1952).
6. Note that Saul (2002) was of the view that Grices main goal is to develop a theory
of speaker-meaning. Following Saul and biting the bullet, Horn (2009) now holds that
conversational implicature is a component of speaker-meaning rather than a pragmatic
inference. By contrast, Levinson (2000), Atlas (2005) and many others are still treating
conversational implicature as a pragmatic inference. My definition is applicable to both
sides.
References
Atlas, Jay D.
2005
Logic, Meaning, and Conversation. Oxford: Oxford University Press.
Atlas, Jay D.
2007
On a pragmatic explanation of negative polarity licensing. In: Noel BurtonRoberts (ed.), 1023.
Bach, Kent
1999
The myth of conventional implicature. Linguistics and Philosophy, 22:
32766.
Bach, Kent
2004
Pragmatics and the philosophy of language. In: Laurence R. Horn and Gregory
Ward (eds.), 46387.
Bach, Kent
2006
The top ten misconceptions about implicature. In: Betty. J. Birner and Gregory
Ward (eds.), Drawing the Boundaries of Meaning, 2130. Amsterdam: Benjamins.
Benz, Anton, Gerhard Jger and Robert van Rooy
2006
Game Theory and Pragmatics. London: Palgrave Macmillan.
Blakemore, Diane
2004
Discourse markers. In: Laurence R. Horn and Gregory Ward (eds.), 221240.
Blutner, Reinhard and Henk Zeevat (eds.)
2004
Optimality Theory and Pragmatics. London: Palgrave Macmillan.
Bultinck, Bert
2005
Numerous Meaning: The Meaning of English Cardinals and the Legacy of
Paul Grice. Oxford: Elsevier.
417
Burton-Roberts, Noel
2007
Varieties of semantics and encoding: Negation, narrowing/loosening and numericals. In: Noel Burton-Roberts (ed.), 90114.
Burton-Roberts, Noel (ed.)
2007
Pragmatics. London: Palgrave Macmillan.
Cann, Ronnie
1993
Formal Semantics. Cambridge: Cambridge University Press.
2005
Carston, Robyn
2002
Chierchia, Gennaro
2004
Scalar implicatures, polarity phenomena, and the syntax/pragmatics interface.
In: Adriana Belletti (ed.), Structures and Beyond, 39103. Oxford: Oxford
University Press.
Crain, Stephen and Paul Pietroski
2002
Why language acquisition is a snap? The Linguistic Review 19: 163183.
Cummings, Louise
2005
Pragmatics. Edinburgh: Edinburgh University Press.
Cummings, Louise (ed.)
2009
The Pragmatics Encyclopaedia. London: Routledge.
de Jonge, Casper C.
2001
Natura artis magistra: ancient rhetoricians, grammarians, and philosophers on
natural word order. In: Ton van der Wouden and Hans Broekhuis (eds.), Linguistics in the Netherlands 2001, 159166. Amsterdam: Benjamins.
Farncescotti, Robert Michael
1995
Even: the conventional implicature approach reconsidered. Linguistics and
Philosophy 18: 15373.
Fauconnier, Gilles
1975
Pragmatic scales and logical structure. Linguistic Inquiry 6: 353375.
Feng, Guangwu
2006
A theory of conventional implicature and pragmatic markers in Chinese. Ph.D.
dissertation. University of Reading.
Frege, Gottlob
1892
ber Sinn und Bedeutung. Zeitschrift fr Philosophie und Philosophisch Kritik 100: 2550. [Trans. On sense and reference. In: Peter Geach and Max Black
(eds.), 1952. Translations from the Philosophical Writings of Gottlob Frege.
Oxford: Blackwell.]
Frege, Gottlob
19181919 The Thought: A Logical Inquiry. Translation in Mind 55: 289311 [(1956)].
Garret, Merrill and Robert M. Harnish
2007
Experimental pragmatics: Testing for implicature. Pragmatics and Cognition
15: 6590.
Gazdar, Gerald
1979
418
Yan Huang
Geurts, Bart
2009
Scalar implicature and local pragmatics. Mind and Language 24: 5179.
Giannakidou, Anastasia
1998
Polarity Sensitivity as (Non-)Veridical Dependency. Amsterdam: Benjamins.
Grice, H. Paul
1975
Semantics 3: Speech Acts, 4158. London: Academic Press.
Grice, H. Paul
1989
Studies in the Way of Words, Cambridge, MA: Harvard University Press.
Grodner, Daniel, Natalie Klein, Kathleen M. Carbary and Michael Tanenhaus
2007
Experimental evidence for rapid interpretation of pragmatic some. Paper presented at XPRAG, Berlin.
Heim, Irene
1983
On the projection problem for presuppositions. Proceedings of WCCFL 2:
11425.
Heusinger, Klaus von and Ken Turner
2006
(By way of an) introduction: a first dialogue on the semantics-pragmatics interface. In: Klaus von Heusinger and Ken Turner (eds.), Where Semantics
Meets Pragmatics, 118. Oxford: Elsevier.
Hirschberg, Julia
1991
A Theory of Scalar Implicature. New York: Garland.
Horn, Laurence R.
2004
Implicature. In: Laurence R. Horn and Gregory Ward (eds.), 328.
Horn, Laurence R.
2006
The border wars: a neo-Gricean perspective. In: Klaus von Heusinger and KenTurner (eds.), Where Semantics Meets Pragmatics, 2148. Oxford: Elsevier.
Horn, Laurence R.
2007
Istvan Kecskes and Laurence R. Horn (eds.), Explorations in Pragmatics: Linguistic, Cognitive and Intercultural Aspects, 3969. Berlin: Mouton de
Gruyter.
Horn, Laurence R.
2009
1: 334.
2004
Huang, Yan
1991
A neo-Gricean pragmatic theory of anaphora. Journal of Linguistics 27:
301335.
Huang, Yan
1994/2007The Syntax and Pragmatics of Anaphora. Cambridge: Cambridge University
Press.
Huang, Yan
2000a
Anaphora: A Cross-linguistic Study. Oxford: Oxford University Press.
Huang, Yan
2000b
Discourse anaphora: Four theoretical models. Journal of Pragmatics 32:
151176.
419
Huang, Yan
2003
On neo-Gricean pragmatics. International Journal of Pragmatics 13: 87
110.
Huang, Yan
2004
Anaphora and the pragmatics-syntax interface. In: Laurence R. Horn and Gregory Ward (eds.), 288314.
Huang, Yan
2007
Huang, Yan
2009a
1: 118153.
Huang, Yan
2009b
Huang, Yan
2009c
Implicature. In: Louise Cummings (ed.), 23438.
Huang, Yan
2009d
Scalar implicature. In: Louise Cummings (ed.), 441444.
Huang, Yan
2009e
Impliciture. In: Louise Cummings (ed.), 23840.
Huang, Yan
2009f
What is said. In: Louise Cummings (ed.), 520522.
Huang, Yan
2009g
Pragmatics. In: Louise Cummings (ed.), 370375.
Hurewitz, Felicia, Anna Papafragou, Lila Gleitman and Roger Gelman
2006
Asymmetries in the acquisition of numbers and quantifiers. Language Learning and Development 2: 7796.
Israel, Michael
2004
The pragmatics of polarity. In: Laurence R. Horn and Gregory Ward (eds.),
701723.
Karttunen, Lauri
1973
King, Jeffrey and Jason Stanley
2005
Semantics, pragmatics, and the role of semantic content. In: Zoltan G. Szab
(ed.), Semantics vs. Pragmatics, 111164. Oxford: Clarendon Press.
Ladusaw, William A.
1979
Polarity sensitivity as inherent scope relations. Ph.D. dissertation. University
of Texas at Austin.
Lehrer, Adrienne
1974
Semantic Fields and Lexical Structure. Amsterdam: Benjamins.
1983
2000
Presumptive Meanings: The Theory of Generalized Conversational Implicature. Cambridge, MA: The MIT Press.
420
Yan Huang
Merin, Arthur
1999
Information, relevance, and social decision making: Some principles and
results of decision-theoretic semantics. In: Lawrence S. Moss, Jonathan
Ginzburg and Maarten de Rijke (eds.), Logic, Language, and Computation,
volume 2, 179221. Stanford: CSLI.
Moore, George Edward
1922
Philosophical Studies. London: Routledge and Kegan Paul.
Noveck, Ira and Anne Reboul
2008
Experimental pragmatics: A Gricean turn in the study of language. Trends in
Cognitive Sciences 12: 42531.
Noveck, Ira and Dan Sperber
2007
The why and how of experimental pragmatics: The case of scalar inference.
In: Noel Burton-Roberts (ed.), 184212.
Panizza, Daniele and Gennaro Chierchia
2008
Two experiments on the interpretation of numerals. Paper presented at Workshop on Experimental Pragmatics/Semantics, the 30th Annual Convention of
the German Society of Linguistics.
Portner, Paul H.
2005
What is Meaning? Oxford: Blackwell.
Potts, Christopher
2005
The Logic of Conventional Implicatures. Oxford: Oxford University Press.
Recanati, Franois
2004
Rieber, Stephen
1997
Conventional implicatures as tacit performatives. Linguistics and Philosophy
20: 5172.
Russell, Benjamin
2006
Against grammatical computation of scalar implicatures. Journal of Semantics
23: 36182.
Russell, Bertrand
1905
On denoting. Mind 14: 47993.
Saeed, John I.
2009
Sauerland, Uli
2004
367391.
Saul, Jennifer
2002
What is said and psychological reality: Grices project and relevance theorists
criticisms. Linguistics and Philosophy 25: 347372.
1995
1974
Pragmatic presupposition. In: Milton Munitz and Peter Unger (eds.), Semantics and Philosophy. New York: New York University Press, 197214.
Strawson, Peter F.
1950
421
Strawson, Peter F.
1952
Welker, Katherine
1994
Plans in the common ground: Toward a generative account of conversational
implicature. Ph.D. dissertation. Ohio State University.
Woods, John
2009
Inference. In: Louise Cummings (ed.), 218220.
Part IV
The place of pragmatics in the description
of discourse
15.

Arnulf Deppermann
1.
Introduction
In this article, grammar is used as an overarching term, encompassing morphosyntax as well as the syntax of the sentential structures (phrase structure and topological structure). Grammar thus includes syntactic categories, functions, and
rules, grammatical constructions, and their realization in terms of forms and patterns of linearization. As grammatical constructions might be more or less lexically fixed and more or less restricted to, e.g., semantic classes of verbs, the boundary between lexicon and grammar is at best fuzzy, and may be better conceived of
as a continuum (Croft 2008). Pragmatics here is also used in a wide sense, referring to the use of language in context. In this way, it encompasses cognitive preconditions and functions of language, such as attention, memory, conceptualization, categorization, and inference, as well as interactional uses, such as transfer
of knowledge, turn-taking, speech acts, and rhetorical functions, displays of emotion and epistemic stance, and indexing social identities and relationships.
In order to set the scene, here are some basic questions about the relationship
between grammar and pragmatics, which will be taken up in this article:
a) The role of pragmatics for grammar
Which pragmatic factors have an impact on grammar? Candidate factors
include cognitive, interactional, situational, cultural, and bodily factors.
Which role do pragmatic factors play in defining grammatical structures?
Can or should grammatical structure be defined on a pure formal basis or
does it ultimately rest on functional categories?
How is pragmatics involved in parsing and interpreting grammatical constructions?
Which grammatical phenomena are governed by pragmatic factors both in
terms of their diachronic emergence and regarding their situated use?
How does usage shape grammar over time?
Which role do pragmatic uses and contexts play in the acquisition of
knowledge about grammar?
b) The role of grammar for pragmatics
How are grammatical structures adapted to pragmatic preferences and
functions?
How are they adapted to their interactive and situational context?
How is grammatical structure used as a flexible tool for speakers pragmatic concerns?
426
Arnulf Deppermann
Do grammatical structures as such have pragmatic meaning?

How are grammatical constructions coordinated with other orders of embodied action?
When asking questions about the relationship between grammar and pragmatics,
we are interested in a subset of issues of the more general question of how language
form relates to language function (see Newmeyer 1998).
2.
Theoretical positions: Formal vs. functional approaches to grammar
No linguist would deny that the choice and at least some aspects of the interpretation of grammatical structures are determined by pragmatic considerations.
However, it is a disputable matter how deeply grammar and pragmatics are intertwined, and consequently, if grammatical analysis needs to attend to pragmatic factors or if it should rather treat grammar as a self-contained module which is working and needs to be analyzed by its own self-sufficient principles. The latter would
be the Chomskian view, coined the autonomy of syntax-thesis (Chomsky 1965).
According to this view, pragmatic factors neither explain the basic parameters and
principles of grammar (e.g., universal organizing principles such as recursivity and
head-specifier-structures), nor do grammatical features arise from pragmatic functions and pressures: In general, it is not the case that language is readily usable or
designed for use (Chomsky and Lasnik 1993: 509). Grammar (which, in this
view, is language proper) is considered as a formal, algebraic apparatus for generating possible, well-formed linguistic structures and for deriving hypotheses about
linguistic structure from perceived speech. While this generativist stance is sometimes dubbed as a formalist approach, functionalists of various kinds (see below)
assign pragmatics an important role in explaining grammatical structure. Basically,
they view language as a means for the communication of thoughts and for social
interaction. They hold that pragmatic functions (such as information-transfer, turntaking, etc.), principles (such as iconicity, categorical organization around prototypes, etc.), and pressures (having to do with processing time, attention span, etc.)
motivate, constrain, and sometimes even fully explain the becoming, use, and interpretation of grammatical structures. There is, however, a broad variation between different functionalist approaches in terms of which grammatical phenomena and which pragmatic factors they focus on and how they conceive of the
impact pragmatics is said to have on grammar (see below and Newmeyer 1998).
2.1.
Pragmatic approaches to grammar
In what follows, we will shortly review approaches which see grammar rooted in
pragmatic motivations.
2.1.1.
427
Functional Grammar and Functional Discourse Grammar
The Dutch theory of Functional grammar (FG; Dik 1997) starts from the assumption that language is structured according to its function as a means of communication of knowledge. It assigns pragmatics the hierarchically dominant role in
determining the semantics and the grammar of clauses. In essence, however, pragmatics in this approach boils down to the information status of speaker and addressee in the ongoing discourse, which is said to determine the information structure of the clause in terms of choosing referring expressions, presuppositions, and
organizing focus vs. background. Expression rules determine how pragmatic and
semantic structures are mapped onto word order and morphosyntax. FG is a nonderivative, cognitivist theory, which requires typological, psychological, and pragmatic adequacy as standards for modelling linguistic structures in formalisms,
which provide for testable predictions (Dik 1997).
Functional Discourse Grammar (FDG) (Hengeveld and Mackenzie 2008) revises FG so as to become a discourse grammar, centering around the discourse act,
which may be larger or smaller than a sentence. Discourse acts are dominated by
speakers intentions (interpersonal level). In a top-down fashion, they govern the
representational (semantics), the structural (morphosyntax) and the phonological
level, which are each formally characterized in strict separation. In contrast to the
other approaches discussed here, (D)FG formalizes its assumptions in a predicate
calculus.
2.1.2.
Systemic-Functional Grammar
Since the 1960s, Systemic-Functional Grammar (SFG) has developed out of the
Firthian British functional linguistics. Its leading figure, M.A.K. Halliday (1985),
assumes that grammar is a system for the creation of meaning, which is structured
according to three metafunctions: the interpersonal (enacting relationships between speaker and addressee), the ideational (construing experience), and the
textual (organizing interpersonal and ideational meaning in the unfolding text or
interaction). Grammars are taken to consist of systems of options, which are available for realizing the metafunctions. Examples are the systems mood for realizing the interpersonal metafunction, transitivity (referring to types of processes
and the participants involved) for realizing the ideational metafunction, and
theme for realizing the textual metafunction. Systems are specific for ranks in the
constituency structure (clause, phrase, word, morpheme). An important place
within SFG is assigned to the study of registers, i.e., varieties of language which
are tied to certain types of situation (e.g., different degrees of formality, languages
for special purposes). SFG assumes that the context of the situation is to be categorized according to three domains, which roughly correspond to the metafunctions: tenor (social/interpersonal relationships), field (processes talked about), and
428
Arnulf Deppermann
mode (medium/channel). Some scholars add genre (relating to the context of culture) as a fourth important domain (Eggins 1994: Ch. 3).
2.1.3.
Functional Pragmatics
Functional Pragmatics is a German approach which sees language as a means designed for the transfer of knowledge and the achievement of mutual understanding
(Ehlich 1991; Redder 2008). Building on Speech Act theory and Karl Bhlers theory of linguistic fields (Bhler 1934; Daalder and Musolff this volume), Ehlich
(1991) distinguishes five linguistic procedures which are used to label five corresponding functional fields with respect to a specific addressee. Each procedure is
realized by specific linguistic means.
Field
Procedure
Linguistic means
Symbol
symbolic
lexemes of content words
Pointing
deictic
local, temporal, personal deictics; tense markers
Guiding
expeditive
interjections; intonation; imperative and vocative

mood
Operational
operational
anaphora; determiners; sub-/conjunctions;

prepositions, accent
Painting
expressive
imitation; exclamative intonation
Figure 1. Fields, procedures, and linguistic means in Functional Pragmatics

(adapted from Hoffmann 2001)
2.1.4.
Cognitive Linguistics and Construction Grammar
Most approaches of cognitive linguistics, such as Cognitive Linguistics (Lakoff

1987), Construction Grammar (Goldberg 1995; 2006), Cognitive Grammar (Langacker 1987; 2008), and Radical Construction Grammar (Croft 2001), consider
grammar to be a semiotic system. Grammatical constructions, which are seen as
the basic, irreducible building blocks of linguistic knowledge (see, e.g., Croft
2001; Goldberg 1995), are defined as form-meaning-pairings (Croft 2008). According to Langackers symbolic thesis, Grammar (or syntax) does not constitute
an autonomous formal level of representation. Instead, grammar is symbolic in nature, consisting in the conventional symbolization of semantic structure (Langacker 1987: 2). The meaning-pole of grammatical constructions includes not only
semantics in a narrow sense, but also discursive, interactional, textual, genre-related, rhetorical, and other pragmatic meanings.1 Besides, following Fillmore
(1985), Lakoff (1987), and others, Cognitive Linguistics denies that there is a clear
boundary between semantics and pragmatics, claiming that all linguistic knowl-
429
edge is in fact encyclopaedic and based on experience. It is only very recently that
cognitive linguistic approaches have turned to the analysis of text-corpora. When
dealing with pragmatic aspects of the interpretation of grammatical constructions,
they mainly focus on information structure2 and implicature (Kay 2004), while discursive phenomena are largely disregarded (but see Langacker 2001). Cognitive
linguists hold that grammar is basically organized in the same way as the lexicon,
linguistic units (Langacker 2000) or constructions (Croft 2008) being the
overarching term. All linguistic structures are subject to processes of categorization (Langacker 2000). Just like lexical items, grammatical constructions are
structured in terms of categories which exhibit a prototypical structure, resulting in
radial and sometimes fuzzy categories, where the individual members need not
share a common set of necessary and sufficient features, but may be related to one
another by family resemblances. Grammatical constructions extend over time
according to principles of metonymical and metaphorical extension, which supply
for motivated options of the change of categorical structures, which, however, cannot be predicted.
2.1.5.
Usage-based approaches
The usage-based approach (Barlow and Kemmer 2000; Bybee 2006, 2007) judges
frequency of usage to be the key factor for the becoming of grammatical constructions. Exemplar models predict that frequency of use is stored and determines the
cognitive entrenchment of linguistic units because it correlates with the probability
of storing a linguistic structure as an unanalyzed unit. Usage-based approaches resonate with connectionist models (MacWhinney and Bates 1989; Bod, Hay and
Jannedy 2003), which rely on neurobiological principles. They predict that high
frequency of reception and production of some linguistic structure strengthens its
activation pattern and facilitates automatic processing. Simultaneously, the occurrence of linguistic structures leads to an inhibition of activation and de-automatization of competing structures (e.g., prepositional genitive with of vs. morphosyntactic genitive with -s). In this view, grammaticality (and conventionality in
general) is a matter of degree and preference, depending on weighted strengths,
which are associated with supposed neural activation patterns and which are subject
to constant change. Synchronically, usage provides for prototype effects of grammatical structures, frequent structures becoming templates for categories and extensions organized around them. Diachronically, it is seen as a major cause of language change and, in particular, of grammaticalization (see 4.8). Such findings are
in line with the position of Emergent Grammar advocated by Hopper (1998, 2004;
Auer and Pfnder 2010). He rejects the assumption of any innate grammatical capacity and he denies that there was anything like the idea of a linguistic system. In
his view, language exists only as an ever changing process of linguistic usage. It is
informed by prior linguistic experiences, which provide the grounds for the routine
430
Arnulf Deppermann
use of (often schematically organized) fragments, which are creatively used as resources and adapted to the situated contingencies of local linguistic practice. In
Hoppers view, sense-making does not rest on rules for well-formed linguistic
structures, but emerges from hermeneutic interpretation, which is in need of interactional ratification.
2.1.6.
Interactional Linguistics
Interactional Linguistics (IL) is strongly based on the methodology of Conversation Analysis (see Domke and Holly this volume). It regards the sequential
organization of talk in oral interaction as the primary habitat of language.3 IL assumes reciprocal influences between linguistic and interactional structures: Linguistic structures are functionally adapted to discursive tasks, while patterns of discourse are sensitive to the linguistic options which individual languages afford and
which speakers choose. IL insists that a theory of the grammar of spoken language
must elucidate the principles of the online-production and -reception of talk-in-interaction (Auer 2009). Most important in this respect is the mechanism of projection of possible/required grammatical continuations and possible/expectable next
actions given the production of an ongoing turn so far (Auer 2005 and 2009). IL
studies put special emphasis on the interplay of grammar and prosody with respect
to action formation (see also Couper-Kuhlen this volume). In contrast to the other
approaches discussed here, IL is not a linguistic theory with a specific model of
grammar and a definite descriptive apparatus, but rather a methodological approach. IL refuses abstract theorizing and insists on the analysis of collections of
linguistic phenomena in talk-in-interaction based on audio-/video-recordings and
transcripts. During the last years, however, researchers in IL increasingly adopt
concepts from construction grammar as descriptive framework4 and positions from
usage-based approaches and emergent grammar (Auer and Pfnder 2010).
3.
The relevance of pragmatics for grammar
In what follows, the impact of pragmatic tasks, processes, and conditions on grammatical structure both in terms of its situated use and of properties of the linguistic
system will be discussed. Pragmatic key factors include temporality (online production and reception), interaction, cognition, action, the temporal and spatial
arrangement of the participants, the modes and materiality of embodied and mediated communication and frequency of usage. For grammatical phenomena relating to the pragmatics-semantics interface, like the interpretation of connectives
and scalar terms and the role of explicature, impliciture, enrichment, conventional
and generalized implicatures in relationship to propositional content and what is
said, see Saeed (this volume).
3.1.
431
Processing constraints and the grammatical packaging of information
The communication of thoughts and the performance of actions by linguistic

practice in real time is constrained by speakers and hearers capacities of cognitive processing. The temporal scope of attention and working memory delimits the
complexity of syntactic and propositional structures which can be processed as
units of planning and understanding. According to Chafe (1994: 109), there is a
one new idea constraint, meaning that neither the speaker nor the listener is able
to handle more than one new idea at a time. This constraint guides the packaging
of information in prosodic and syntactic units. Preferentially, speakers formulate
one new information per intonation unit, which has its (one) focus accent on the
item which expresses the new information. The intonation unit almost never comprises more than one clause. As a consequence, syntactic structures with multiple
recursions, which force speaker and hearer to keep track of two or even more syntactic and semantic structures simultaneously, are strongly dispreferred because of
the processing difficulties they cause.
An important syntactic repercussion of the one new idea constraint can also
be seen in the constraints of Preferred Argument Structure (Du Bois 2003), which
have been proved to hold for a large variety of typologically unrelated languages
(Du Bois, Kumpf and Ashby 2003). Du Bois (2003) claims that speakers avoid
more than one lexical core argument (which encodes the new information). For
transitive constructions, this amounts to the constraint to avoid lexical subjects
(and to express them pronominally), while direct and indirect objects can be lexicalized. The constraints of preferred argument structure explain, e.g., preferences
for cleft-constructions which can be observed in many languages, such as French
(cf. Lambrecht 1987). Instead of producing a transitive clause with both lexical
subject and object, speakers prefer a cleft-construction, which package the information in two clauses with only one lexical argument each.
(1) Transitive clause
Beaucoup de gens naiment plus cette faon de parler.
Many people do not like this way of talking anymore.
(2) il y a-cleft
Il y a beaucoup de gens qui naiment plus cette faon de parler.
There are a lot people who do not like this way of talking anymore.
Such cleft-constructions fulfil the same pragmatic function as left dislocation and
free topics: They are used to introduce a referent in a segment of its own, which, in
addition, is often produced as a separate intonation unit apart. Only after this, the
referent is topicalized and predicated upon.
(3) Left dislocation
aber ich mein ne gewisse neugier (.) die muss ja bleiben
but I mean a bit of curiosity (.) it must remain
432
Arnulf Deppermann
(4) Free topic

die polnische Wurst da (.) ist das so wie Salami?
the-FEM Polish sausage over there (.) is it-NEUTR like Salami?
Grammatical constructions as in (3) and (4) give evidence that in contrast to what
Speech Act Theory holds (cf. Searle 1969), an elementary speech act need not consist of reference plus predication. Rather, establishing a referent often is a fullblown action by itself. It is performed by clausal and phrasal constructions which
separate it from the following predication and serve as a starting point for it, which
first is grounded independently. Moreover, referential act and predicative act each
occupy an intonation unit of their own.
In a similar vein, speakers prefer right-branching constructions over leftbranching constructions because they allow for a cumulative processing of structure and meaning, which correlates with the temporal production of phrases. Leftbranching constructions, however, require language users to put elements on
hold in the working memory, before being able to assign them to their syntactic
heads, which are produced later. The need to keep track of several syntactic and
semantic representations in parallel also strongly disfavors central embedding, let
alone multiple recursion (such as multiple center-embedded relative clauses).
3.2.
Structuring information with respect to relevance and the recipients

knowledge
Text and talk is designed with respect to what speakers/writers assume to be common ground with their addressees. New information, which is either unknown or
not salient to the addressee at the moment, is introduced into the discourse by relating it to given information, which is judged to be salient, i.e., activated and retrievable with low processing effort for the hearer. In order to structure and encode
information accordingly, speakers keep track of the cognitive availability of referents (but also other denotata, such as states, actions, and events).
Given vs. new is not a simple dichotomy. Rather, there are different degrees of
accessibility of referents (e.g., Chafe 1994), which depend crucially on their identifiability and their state of prior activation. Denotata may be accessible because
they are situationally salient (via perception), established as parts of world knowledge (like the prime minister), have been introduced in prior discourse, or can be
inferred from it via bridging assumptions which, e.g., relate parts to wholes (Clark
1977).
(5) Bridging assumption (whole f part)
The couple moved on to the altar. The bride wore white.
f Bridging assumption about co-reference: The bride is part of the couple mentioned before.
433
The assumed familiarity of referents for the recipient thus can range from (brand-)
new over inferrable to (exophorically or anaphorically) evoked items (cf. Prince
1981). Accessibility of referents depends on several factors, such as its topical vs.
non-topical status, recency of last mention of the referent, or the degree of cohesion of the syntactic unit which contains the prior mention to the ongoing unit
(Ariel 1990).
Ariel (2008: 4452) proposes a referential marking scale. It orders referential
expressions according to the (assumed) degree of accessibility of the referent for
the recipient, which in turn depends on the degree of its activation in the recipients
memory that can be assumed when producing the referential expression (see also
Gundel, Hedberg and Zacharsky 1993). While new elements are introduced by indefinite NPs, accessible and given elements are referred to by definite NPs or (if
situationally salient or evoked) by pronouns. Starting with the most marked expression to be used for the least accessible referents, the referential marking scale
for person reference looks like this in more detail (Ariel 2008:44):
full name > long definite description > short definite description > last name > first name
> distal demonstrative > proximate demonstrative > stressed pronoun > unstressed pronoun > cliticized pronoun > verbal person inflections > zero
This scale shows how the two pragmatic preferences for person reference, namely
minimization and use of (recipient designed) recognitionals (Sacks and Schegloff
1979), translate into preferences for the use of grammatical resources. Their relevance can also be seen in cases of repair like (6), where more minimized forms are
used first and get replaced by (grammatically and semantically) more elaborate
forms if they fail to engender the recipients recognition.
(6) Preference for minimization and use of recognitionals (example from Sacks
and Schegloff 1979: 19)
A: I was the only other than the tch Fords? Uh Mrs. Holmes Ford? You know
uh the the the cellist?
B: Oh yes. She is the cellist.
Apart from morphosyntactic coding, word order is used as a syntactic device to
index the assumed familiarity of referents. Halliday (1985: 275) claims that there is
a natural ordering principle: the Given typically precedes the New, i.e., the
speaker organizes the message as a trajectory starting from the current state of consciousness of the recipient to an updated state (see also Chafe 1994). Therefore,
given information is expressed as theme in the prototypical case. In Germanic and
Romance languages, the theme is the first constituent of a sentence, whereas in
Japanese it is marked by the post-positioned theme marker -wa. New information
is presented in the rheme, i.e., the remainder of the sentence (Halliday 1985: Ch.3).
In addition to the assumed familiarity of referents, a second important dimension
of information structure is the organization of focus vs. background (Bring 2007;
434
Arnulf Deppermann
Krifka 2007). Focus refers to the part of an intonation unit which is framed as
being most relevant. Some of the means speakers can use to mark focus are the
same as for marking accessibility: word order and specific topological constructions like (pseudo-)clefts and passives. In addition, most important is the placement of the so-called focus accent (Uhmann 1991; Bring 2007), which marks
the focal word (but can project to higher-order nodes such as phrases or even the
whole clause). All other things being equal a Speaker will choose the Theme from
within what is Given and locate the focus, the climax of the New, somewhere
within the Rheme (Halliday 1985: 278). While there is a high correlation between
focus, new information and rheme, they need not coincide (Gundel and Fretheim
2004). The focus does not necessarily need to be put on new information, but it
can, e.g., be contrastive, i.e., marking an element which stands in contrast to a prior
turn or the recipients probable expectations. The focus can also comprise the
whole sentence, as in the case of offering requested information to open questions
like
(7) Focus on whole sentence (= requested answer)
A: Whats up?
B: The computer doesnt work.
or as in newspaper headlines
(8) Focus = whole headline (The New York Times, November 27, 2009)
Pilots Were Distracted, Transcripts Show
In elliptical responses, only the focus elements are realized.
(9) Adjacency ellipsis: Question-answer sequence
A: wieviel Lieder knnen wir spielen? was meint ihr?
How many songs can we play? What do you think?
M: Z::irka fuffzehn;
About fifteen.
In spoken language, the focus is often placed in thematic position, particularly, if it
is contrastive.
(10) Focus on theme
A: WANN ist er gekommen?
when did he come?
B: GEStern war er=s erste mal da.
yesterday he was there for the first time.
Another grammatical construction for focus marking by assigning it to the theme is
preposing a constituent (off he went; cf. Ward and Birner 2004). Focus preposing
needs to be distinguished from topicalization of a referent, which does not carry a
focus accent and contains neither focus nor new information.
435
As stated above, focus marking can be independent of accessibility status:

(11) Contrastive accent
Das war nicht ICH. Das warst DU.
It wasnt ME. It was YOU.
In fact, focus marking and marking of accessibility are not determined by identifiability or activation status of referents, but they can be used as rhetorical devices
to introduce new information by simply presupposing it as hearer-old (see also Edwards 1997:117122):
(12) Coding givenness as a rhetorical device (advertisement by Volkswagen)
Just when you thought our legendary hatchback couldnt get any better,
along comes the stunning new Golf.
In sum, signalling familiarity of referents and assigning focus vs. background results from an interaction between grammatical forms (such as definiteness, pronominalization, deictics, etc.), word order, and prosodic cues (accentuation),
which at times can be very complex.
3.3.
Projecting turn-construction and turn-taking
In spoken language, the clause is not only the preferred format for packaging ideas
(see 4.1), it is also used as the primary construction format for turns and actions
(Thompson and Couper-Kuhlen 2005). Participants use clausal structures for indexing and recognizing transition relevance places at which turn-transition might
be due (Selting 2000; Sacks, Schegloff and Jefferson 1974). A basic mechanism to
do this is temporal projection (Auer 2005 and 2009), which rests on the incremental production of linguistic structures in time. Linguistic structures produced in the
turn so far open up syntactic gestalts, i.e., probable or required continuations,
which have to be closed by producing the required missing elements. E.g., the apodosis of a conditional projects a protasis (and this is: a clause-combination) to follow, verbs project required arguments by virtue of their valence frame, determiners
project noun phrases, etc. Projection depends on typological features of the language. E.g., in German the verbal brace (Verbklammer) requires speakers to finish
clauses with an infinitival form (infinitive or participle) of the main verb (unless it
is used in present tense without a modal). Listeners can use this feature to generate
expectations about when the turn is possibly completed. SVO-languages like English permit an early projection of turn trajectories because the verb makes specific
choice of thematic roles expectable. An OV-language like Japanese, however, does
not allow for early projection, because constituents other than V do not project the
number of arguments and because Japanese allows for a much higher degree of ellipsis. Still, it marks endings more clearly in terms of syntax than SVO-languages
because of the regular positioning of the verb in sentence-final position and be-
436
Arnulf Deppermann
cause of the additional use of sentence-final particles (Thompson and CouperKuhlen 2005). Projectability of turn-endings makes itself felt not only in smooth
turn-transition, but also in cases of overlap, which tend to occur more often if elements are added after a possible clausal gestalt has been completed and thus a
possible transition relevance place has been reached (Jefferson 1973; Schegloff
1996). The collaborative production of sentences is another evidence for participants orientation to the exigencies of grammatical structure in terms of projectable trajectories of clauses (Hayashi 2003; Lerner 1991, 1996, 2004): After the
first speaker has produced the first clause of a complex sentence, second speakers
manage to deliver the second clause which was type-projected (e.g., a then-clause
after an if-clause has been completed) or they may add phrases or words at the
places where their occurrence is due.
Although clauses have more or less determinate endings, which may signal
turn-transition, they can be expanded by additional elements, so-called expansions or increments, which are syntactically related to the syntactic structure
constructed so far.5 Such turn-continuations occur regularly if a projected next
speaker does not take the turn when a transition relevance place is reached. They
address problems of understanding and affiliation with the turn so far, which might
account for an actual lack of response or an anticipated difficulty in responding.
They serve to repair the respective deficits of the turn and to renew the occasion for
turn-transition. Whether clauses project turn-transition and whether expansions are
a continuation of an ongoing turn-constructional unit or a new unit does not only
depend on syntax, but also on the prosodic contour of the clause and on the degree
of prosodic integration of the continuation. Transition relevance places are not
only determined by syntax, but they emerge from an interplay of syntactic, prosodic, semantic, and, at times, also pragmatic and genre-related considerations (as,
e.g., narrative structure).
3.4.
Social action formats
Grammar is a primary resource for performing recognizable social actions. Speech

act theory conceives of the relationship between speech acts and grammar in terms
of illocutionary force indicating devices, which index and sometimes even encode
types of speech acts. Most important indicators are sentence type (declarative,
imperative, interrogative), word-order, modal verbs and modal particles, performative formulae, punctuation, and sentence-final pitch contours (Searle and
Vanderveken 1985; Liedtke 1998). Empirical research into the relationship between action and grammar has refined this picture quite a lot: Instead of relying on
theoretically derived speech act taxonomies and trying to find linguistic correlates
for very broadly defined (universal) types of speech acts, corpus-based research
shows that linguistic action needs to be understood in terms of much more specific
emic types of actions to which speakers and recipients orient as such. At times,
437
they are tied to very specific sequential contexts, genres, and speech communities.
These actions are realized by idiomatic grammatical constructions, which have become closely associated with them by routine usage in specific pragmatic contexts
and genres, leading to the conventionalization of pragmatic meaning (Feilke 1996
and 2004). Examples of social action formats (Fox 2007) are constructions with
German nicht verstehen (do not understand; Deppermann 2010), which are used
for building reproaches:
(13) Social action format: [nicht verstehen (knnen) COMP] indexes reproach
B1: ich versteh auch andereseits NICHT wenn dieon the other hand I dont understand when the
klgerin zum arzt nach BRUCHtal MUSS JA?
plaintiff (=A1) needs to go to the doctor in Bruchtal
dass sie dann in de innenSTADT im maxim
that she then goes to Maxims in the city
KAFfee trinkt
to have a cup of coffee
(1.8)
A1: wieso is des verBOte?
why is that forbidden?
Deontic infinitives in German are another example. Deontic infinitives are not syntactically integrated, i.e., they are free infinitives, which are neither governed by
a matrix sentence nor are they part of a composite verbal phrase made up by an
auxiliary and an infinitive (cf. Fries 1983; Deppermann 2006 and 2007). With
deontic infinitives, the speaker expresses a normative stance towards some kind of
(future) action denoted by the verb. They are employed for a variety of deontic actions, which may vary in terms of the degree of obligation, such as commands, requests, recommendations, and suggestions, but also the expression of a wish (optative) or a complaint.
(14) Social action format: Deontic infinitive used for suggestion
A: also der Ablauf is grob (--)
so the procedure will be something like this
Anmoderation; (.) oder lAbern;
anchorage or talk-INF
M: vorstellen auf jeden fall
introduce-INF in any case
(15) Social action format: Deontic infinitive used for complaint
noch drei wochen arbeiten!
still three weeks of work-INF
arbeiten arbeiten immer nur arbeiten
work-INF work-INF always only work-INF
438
Arnulf Deppermann
An English example is the use of the grammatical construction [PRO COPULA

(Grading particles: really/so/quite/pretty) PRED-ADJ] for assessments, in particular, for compliments (Goodwin and Goodwin 1987):
(16) Social action format: [PRO COPULA (Grading particle) PRED-ADJ] used
for assessments
This is so nice.
Thats quite good.
Another example is the use of post-positioned causal adverbial clauses (starting
with because) as justifications dealing with actual or anticipated disaffiliation on
the recipients part (Ford 1993):
(17) Social action format: Post-positioned causal adverbial clause dealing with a
missing recipients reaction (example from Ford, Fox and Thompson 2003:
126)
A: well do you think its umm (0.2) ahm (0.2) stress? (.)
Cause a lot of back I know back pain (0.2) comes from stress.
Actions are rather contextually indexed than invariably coded by grammatical
constructions: Sequential, genre-related, and other contextual conditions, lexical
choice of the verb, and prosody determine which precise action is performed by a
deontic infinitive; causal adverbial phrases may also be used to secure understanding or to refute a statement, which an opponent had asserted before. Grammar only
very rarely stands in a context-free, one-to-one relationship to some type of action.
3.5.
Displaying stance and marking discursive coherence
Texts and interactions do not simply consist in the exchange of information and the
performance of actions. Speakers use various grammatical means to display their
epistemic and evaluative stance with regard to information and action, to convey
how they relate to recipients assumed understandings and to connect their utterance with prior and subsequent clauses and turns. Various kinds of discourse particles (see Fischer 2006) such as discourse markers (like you know, well, and in
some of its uses; Schiffrin 1987), response particles (like uhum, yes, okay; Gardner
2001; Sorjonen 2001), and modal particles (like just, germ. wohl, eben, halt, ruhig;
Diewald 2006a) are grammatical means which are specialized for displaying how
some action relates to prior (or presupposed) context in terms of coherence, epistemic stance, or affiliation with the partner. With the exception of modal particles,
discourse particles are not syntactically integrated. Mostly, they occupy turn-initial
positions in the so-called pre-front field (Auer 1996b), or, in the case of response
particles, they sometimes make up the whole turn. Pre-front field elements can also
be adverbials, address terms, and routine formulae.
439
(18) Pre-front field: Connective + response particle

also gut das haben sie verstanden
so okay you got this
(19) Pre-front field: Address term + discourse marker
herr meier ich mein wir haben uns ja gestern schon drber unterhalten
mister meier I mean we talked about this already yesterday
Pre-front field elements serve as pivotal elements which retrospectively tie the upcoming action to prior context and prospectively project the type of action to come
in various respects.
Projector constructions comprise another set of prefatory grammatical structures which serve for anticipatory framing of upcoming discursive action (Gnthner 2008; Hopper and Thompson 2008). Projector constructions can be realized in
various forms, which are in part language specific: as pseudo-clefts (wh-clefts;
Gnthner and Hopper 2010; Pekarek-Doehler 2010), it-extrapositions (Prince
1978), [N be that]-constructions (Gnthner 2008; Schmid 2001), il y a-clefts, or
the je veux dire que-construction (Pekarek-Doehler 2010).
(20) German pseudo-cleft (from Gnthner and Hopper 2010)
was so die sachn in natUrkostldn angeht,
as far as the things in health food shops are concerned
die sind ja dann meistens
they are mostly
erfahrungsgem auch relativ teuer
also quite expensive
(21) English pseudo-cleft (from Gnthner and Hopper 2010)
and so what happens is,
your mic runs into i:t,
and then it plugs into your receiver
(22) French pseudo-cleft (from Pekarek-Doehler 2010)
ce que je voulais dire
what I wanted to say
cest vrai qu il y a quand-mme tout ce phnomne,
its true that there is nevertheless this whole phenomenon
Examples (20)-(22) show that the two parts of the pseudo-clefts need not be linked
by a copula and a subordinating construction (what p is that q). Moreover, the second part often does not consist of a clearly delimited clause, but rather of some
more vaguely defined longer stretch of talk. The projector construction thus is a
very interesting case of the intersection of grammar and pragmatics (cf. Hopper
and Thompson 2008). The first parts (the wh-clauses etc.) do not project some spe-
440
Arnulf Deppermann
cific clausal structure to follow, but they rather serve as flexible fragments for holding the turn by projecting a multi-unit turn to follow. Depending on the metacommunicative framing or the announcement made by the wh-clause, it projects an
argumentation, a complex statement or a small story.
3.6.
Presupposing context
Ellipsis provides the most striking evidence for the claim that syntax is tied to the
here and now of talk (Bhler 1934) and that it is positionally sensitive (Schegloff
1996). According to traditional definitions, ellipsis applies to grammatical structures, which are acceptable in context (i.e., they are not anakoluths), but which are
phonologically reduced with respect to the exigencies of syntactic completeness
and of their semantic interpretation in situ. Ellipsis results from economical motivation (cf. Grices second maxim of quantity, see Grice 1975). It requires that
elided elements can be recovered from context (Halliday 1985: Ch.9.3). Context
can offer the missing elements in basically two ways: Ellipsis may be anaphorical
or situational.
a) Anaphorical ellipsis copies presupposed elements from prior text or talk (save
adaptations of speaker-listener deixis). Adjacent ellipsis is a most obvious case.
Typically, a second turn is restricted to the production of the (requested, problematic etc.) focus element, while the prior turn is presupposed as background:
(23) Adjacent ellipsis: Answer
IN is sexualitt fr sie total WEG?
is sexuality completely off for you?
IE als unterhAltungsthema (.) JA.
as a topic for talk (.) yes
(24) Adjacent ellipsis: Follow-up question
IE: das finde ich falsch
that is wrong in my view
IN: warum?
why
However, phonological reduction is not the full story. There are in cases where
there is no prior clausal structure which can simply be copied. Consider cases like
(25) and (26):
(25) Unattached NP increment (example from Ford, Fox and Thompson 2002)
She had it [a baby] yesterday.
Ten pounds.
441
(26) Free NP
CH: der abwasch ist FERtig (-) darf ich jetzt FERNsehen?
the dishes are cleaned (-) may I watch TV now?
PA: ERST hausaufgaben DANN fernsehen.
first homework then TV
While the subject can be copied from the preceding turn constructional unit in (26),
there is no verbal antecedent to be presupposed for the elliptical structure. Instead, it
has to be constructed on the basis of the representation of the conceptual content of
the discourse so far and by resorting to encyclopaedic knowledge and sequential expectations with regard to next actions. E.g., in (26), the request for permission makes
permission vs. refusal conditionally relevant and assigns normative power to the respondent. In addition, the construction [first NP/VP then NP/VP] is an elliptical social action format, which codes a normative order of actions. Thus contextual resources of interpretation can combine with the semantics and pragmatics which are
tied to elliptical grammatical constructions as such. Ellipsis may not only be motivated by striving for economy, but it may also have rhetorical functions. Deontic infinitives (cf. (14)-(15)) are a case in point: By neither encoding subject (i.e., an agent)
nor mood, the construction neither specifies who has to carry out the action nor the degree of obligation assigned to the agent. In some contexts, the lack of mood-coding
and the omission of arguments (agents, but also often patient, location, instrument,
etc.) is a politeness strategy (for requests and commands), while in others (like advertisements, see (27)) it allows the recipient to define for him/herself which degree
of normative obligation suits and if s/he feels addressed by the deontic infinitive at all.
(27) Deontic infinitives in advertisement
Jetzt mitspielen und gewinnen!
Join-INF the game and win-INF now!
In addition, these omissions help to profile (Langacker 2008: 6670) the action as
the most important element of the situational frame evoked by the infinitive clause.
These factors seem to account for the fact that in many contexts deontic infinitives
are preferred over imperatives and modal constructions, which require the speaker
to encode agents and the degree of normative obligation. Ellipsis can reach back
far beyond an immediate prior interactional context. Raymond (2004) shows how
stand-alone so in American conversation is used to prompt an action from a recipient as an upshot from the ongoing interactional sequence which had been due already (much) earlier in the interaction. The syntactic incompleteness of so is a case
of aposiopesis. It invites the recipients collaboration by indexing that the recipient
is supposed to know on behalf of the prior interactional sequence which kind of
action is warrantably expected from him/her. A similar case is sentence-final
but, which indexes a hanging objection which can be recovered from context
(Mulder and Thompson 2008).
442
Arnulf Deppermann
b) When situational ellipsis is used, it is the non-verbal context with respect to the
speakers origo (Bhler 1934), which supplies the missing information.
(28) Situational ellipsis: pragmatic routines (customer to waiter)
The bill, please.
(29) Situational ellipsis: pragmatic routine and local specification (announcement
at a railway station platform while train is approaching)
Zurckbleiben!
Keep away!
(28) and (29) are examples for routinized elliptical constructions which are tied to
specific social and (in the second case) spatial configurations which they require
for their interpretation. Note that there is a reflexive relationship between the construction and the context: While the situational context is necessary for their precise situated interpretation, the formulae themselves organize contexts by indexing
features which are (possibly) relevant.
3.7.
Multimodal coordination
Situational ellipsis mostly does not simply rest on some spatiotemporal context
which is taken for granted. Rather, the production of grammatical structure is coordinated in a fine-grained way with other modes of communication (Kress and
van Leeuwen 2001; Norris 2004) which make contexts selectively salient and
which cooperate in determining meaning. In oral interaction, the meanings of
spatial reference and situational ellipsis are established by a complex interplay of
linguistic structure, pointing, available visual clues, background knowledge, and
larger activity frames, which provide for putative relevance and possible semiotic importance of spatial structures (Goodwin 2003). E.g., pointing gestures
need to precede deictic terms used for spatial reference within or shortly before
the same turn in order to draw the recipients attention to the objects designated
and thus to constrain the situated meaning of the terms used. An example for the
multimodal embedding of ellipsis would be a director saying much too slow
while watching a film take on the monitor at the film set. Her visual orientation
to the screen, knowledge about some prior interaction with the camera woman,
and the routine practice of the director reviewing film takes in order to decide if
they are acceptable will be required in order to understand that the turn is taken
to mean that the camera moves much too slow and that the take needs to be redone. As in this example, coordination of gaze between speaker and hearer, visually available context, and gesture mostly provide for the thematic referential
aspects of communication, whereas linguistic communication is restricted to
rhematic predications expressing some stance or request for action. However,
gesture can also be used to elaborating a predication, as in the case of modal in-
443
dexicals like germ. so and engl. like (Streeck 2002). They need to be accompanied by an illustrative gesture in order to clarify which semantic aspect (like
size, form, direction of the referent, kind of movement) is indexed by the
modal expression.
3.8.
Grammatical variation relating to genre and medium
There are huge differences between spoken and written varieties in virtually
all areas of grammar, e.g., regarding the use of (complex) clause patterns (such
as extraposition, subjunction), tense and aspect marking, valency, or discourse
markers (Biber, Johansson and Leech 1999; Bublitz, Lenk and Ventola 1999). A
closer look, however, reveals that there is no simple dichotomy between spoken
and written language. Grammar is sensitive to fine-grained differences of register
(e.g. Biber 2006), genre and medium. During the last decades, micro grammars
related to genres of computer mediated communication (CMC) have developed.6
In genres and technical forms of communication with a high degree of interactivity like chats and SMS in particular, new grammatical conventions have emerged
for the display of emotions (multiplication of vowels and diacritics, expressive
use of capital letters, emoticons), the mimicry of nonverbal communication (root
inflectives known from comics such as germ. heul, grins), the adaptation of writing to oral interaction (eye dialect, dialectal and substandard spellings: boyz,
wuddya; use of discourse markers and non-standard grammar), and we-codes,
which are intransparent to out-groups (e.g. by using acronyms like lol, CTN, 4U;
Schlobinski 2006). In CMC in general, standards of linguistic correctness have
become liberalized, giving way to innovative playful, expressive, and economical uses.
In print and internet media, texts often combine with the visual design of pictures (Kress and van Leeuwen 2006; Stckl 2004) and other semiotic modes such as
sounds, animations, and movies (e.g., in vlogs). On websites, e.g., large parts of
texts are elliptical, consisting of free nominals which are not integrated into any
syntactic structure. However, their meaning is supplied by the visual line-up they
are embedded in and by the users pragmatic knowledge about how to use them on
behalf of there being links to other pages, forms to be filled out, etc. (Schmitz 2010).
3.9.
Pragmatic factors of grammaticalization
Language-internal and social factors have already been discussed for a long time as
determinants for linguistic change (see Aitchison 2001). Over the last three decades, inferential, discursive, and usage-based processes have increasingly become
recognized as major factors promoting linguistic change. A core area of change is
grammaticalization. The term refers to the emergence of new grammatical forms
(from old ones); more specifically, grammaticalization involves loss of lexical
444
Arnulf Deppermann
meaning (semantic bleaching), gain of functional meaning (e.g., word-class conversion), loss of phonological substance, restricted syntactic distribution or morphological incorporation, and may involve the development of a discursive pattern
into a grammatical convention. Grammaticalization has been found to be a probably unidirectional process, which runs along clines of increasing grammaticalization (Haspelmath 1999; Hopper and Traugott 2003: Ch.5). One such cline, which
can be attested for several languages, is the following diachronic path: content
word f function word (analytic) syntax (e.g., PRO) f agglutinative morphology
(clitization) f flexive morphology (e.g., subject marker) f zero.
Grammaticalization can be motivated by various strategies and preferences of
language users:
Articulatory economy leads to phenomena like reduction/attrition, clitization,
assimilation/co-articulation (Lass 1987).
Processing economy can be seen at work in heavy NP-shifts (Wasow 2002),
which correspond to the Gesetz der wachsenden Glieder (law of increasing
constituents, Behaghel 1909).
The enrichment of expressive power by innovations.
The desire to communicate unambiguously leads to, e.g, the decay of homonyms.
Iconicity (Haiman 1985) is a cognitive motivation which builds on principles
of naturalness (e.g., constructing superlatives by reduplication: chef chef;
again and again)
The desire to accrue social prestige, to gain the status of a member and to identify with a social group leads to the use of linguistic practices of prestigious ingroups (Labov 2001; LePage and Tabouret-Keller 1985). Innovations in youth
language provide paramount examples.
To be sure, linguistic change in general and grammaticalization in particular do
not only involve change of forms, but also semantic and pragmatic change. A common mechanism of change is metaphorical extension, which works by analogical
transfer of concepts from a source to a target domain. Metaphoric change can be
observed in many languages for deictics, where a metaphorical transfer of expressions for body parts and bodily orientation to spatial reference in general
comes about (Heine, Claudi and Hnnemeyer 1991; Heine 1997). A well attested
path of metaphorical change in the meaning of prepositions and connectives is spatial f temporal f causal f epistemic/speech act/discursive/textual (Sweetser
1990; Haspelmath 1997). The general dynamics of metaphorical extension consists in using anthropocentric, visually based, concrete concepts for more abstract,
discursive, and subjective entities. The diachronic change of, e.g., connectives
(like while, since, after (all)) thus shows increasing subjectification and abstraction
in terms of being used for displaying speakers attitudes, epistemic perspectives
and expectations (Sweetser 1990).
445
While metaphor operates on the paradigmatic level (supplanting one meaning

with another), metonymically based change rests on reanalysis of syntagmatic relationships, which in addition often become contracted over time (Hopper and
Traugott 2003: Ch.4). In metonymic change, a lexical item or (parts of) a construction adopt meanings from their collocational contexts. Metonymic change is more
gradual than the more abrupt and drastic metaphoric change, because what develops is rather informative strengthening or enrichment (i.e., an implicature is added;
Ariel 2008:163) and successive transformation. A much discussed example of metonymic change is the development of to be going to (e.g., Hopper and Traugott
2003: 8893; Bybee 2006): Originally, it denoted physical movement to some
place; later it became a purpose-construction, referring to the intention of an actor;
nowadays, it has become a generalized future construction, which may also be
used with inanimate, non-intentional subjects. Morphologically, there is a reanalysis of going to, which is no longer interpreted as a progressive form of a full verb,
but as a tense marker, and there is phonetic reduction (gonna), which attests to the
unit status of the construction. This reduction is absent in the progressive use (Im
going to a party vs. * Im gonna a party).
In many cases, implicatures (or invited inferences, see Traugott 2004) are at the
beginning of grammaticalization. The starting point are conversational implicatures which heavily depend on specific supportive contexts. Gradually, they become generalized implicatures, i.e., prototypical meanings (Traugott 2004). Grammaticalization has come about when the new meaning is firmly established, so that
there are (at least) two meanings, i.e., the item/structure has become ambiguous.
The next step then is reorganization and differentiation of the two meanings in
terms of the development of specialized distributions and collocational potentials,
differences in potentials for TMA-marking, etc. (Diewald 2006b).
While cognitive motivations and interpretations can be seen as important
driving forces for grammaticalization, cognitive factors as such are never directly
responsible for language change. They can leave their fingerprint on language
structure only via their discursive realization. Moreover, discourse is not only a
medium for the linguistic reflection of cognitive preferences. There are genuine
discursive motivations for the development of grammatical structures. Probably
all successful grammaticalization paths had previously led to salient discourse
patterns, themselves only necessary (but not sufficient) conditions for conventionalization (Ariel 2008:150). Except for language planning, coining of terminology, and some innovations promoted by the media and by subcultural actors,
people only rarely intend to promote language change. And even if there is some
definite intention of innovation, its success fully depends on how it is taken up
and spread by recipients, both in terms of the interpretation of the innovation and
the frequency of use. At least in the early stages of innovation, intentional strategies of people pursuing communicative goals often are responsible for language
change, which, however, as such mostly happens as an unintended by-product of
446
Arnulf Deppermann
these strategies. Keller (2004) terms such changes invisible hand-processes.

Sometimes, the long-term consequences of the factual collaboration of intentional strategies of a community of linguistic actors can even lead to a result
which is quite the opposite of what was originally intended. An example is the
dynamics of diachronic change of terms for women in German (Keller 2004):
The collective adherence to a strategy to select distinctive terms for addressing
and talking about women eventually leads to a pejorization of these terms (Weib,
Weibsbild, Frauenzimmer): Frequent use makes them lose their distinctive potential, thus speakers strive to replace them by more marked and highly valued
terms, which get increasingly used as a consequence and thus also lose their distinctive value, etc. (The same process can be observed in an accelerated pace
with innovations in youth language, e.g., regarding evaluative predicates). So, in
the long run politeness and distinction strategies have the inverse impact on the
terms used.
Other discursive preferences are conservatism (in a more macro view) and, in
terms of a micro view of dialogical sequences, priming and resonance (Pickering
and Garrod 2004; Brennan and Clark 1996). They account for the preference to follow precedents, which thus can become canonical. These strategies of repetition
are most promising in terms of securing understanding and building on intersubjectivity achieved before (Feilke 1996). They provide for a systematic interactive
motivation of the grammaticalization of options which initially may simply have
been chosen arbitrarily among a set of possible options (see Clark and WilkesGibbs 1986).
Patterns of use lead to the gradual emergence of new grammatical patterns
from non-innovative, grammatical usage. The emergence of discourse particles
(well, but, though) from connectives and other discourse markers (like yknow, verstehste do you understand, voil ) from mental verbs are examples of gradual
change. Precursor for the development of discourse markers from mental verbs is
the emergence of the construction [matrix sentence with mental verb + dependent
main clause] (Auer 1998), i.e., a semantically dependent clause follows which
does not exhibit any marker of subordination. The matrix sentence can increasingly change into a semantically bleached discourse marker, which then, in later
phases, may fully grammaticalize in terms of adopting a specific distribution (often
syntactically unembedded in turn-initial or turn-final position) and retaining only
light lexicosemantic content (if any) (Thompson 2002; Imo 2007; Karkkinen
2003). Dialogical patterns can also be a source of grammaticalization. A case in
point are clause combinations which arise from interactional sequences and transform them into grammatical formats to be used by individual speakers in single
turns (Givn 1979). An example is the emergence of conditionals from a questionanswer-response sequence of three turns:
447
(30) Grammaticalization of interactional sequences in complex sentences: Conditionals

A: Yes/No-Question (Will you help me?)
B: Agreement (Yes.)
A: Consequence (Then Ill give you two dollars.)
f A: Protasis (If youll help me),
Apodosis (then Ill give you two dollars.)
Discursive and cognitive factors are often intertwined in processes of grammaticalization. The category noun exemplifies how cognitive concerns are mediated
by discursive ones. Cognitive grammar claims that the grammatical category
noun has a schematic semantics, which consists in conceptualizing referents as
things which are defined as any product of grouping and reification (Langacker 2008:105). The analysis of the realization of nouns in discourse, however,
shows that prototypical representatives of nouns, which are count nouns and exhibit all grammatical features which are regarded as typical of nouns (determiners,
gender-, number-, and case-marking, etc.) introduce discourse topics which are
manipulable, i.e., available for subsequent co-reference (Hopper and Thompson
1984). Grammatical categoriality thus is achieved by discursive patterns, not
by mental conceptualization alone. Discursive considerations can even override
cognitive motivations. In the overwhelming majority of cases, causal adverbial
phrases (with because, weil, par-ce que etc.) follow their nucleus assertion, although adherence to an iconicity principle would predict that they precede because
causes are prior to effects (Ariel 2008: 175 f.). However, it is neither the way the
world is (temporal order) nor cognitive preference for iconic representation, but
the discursive function of dealing with the recipients actual or anticipated disaffiliation by supplying justifications, causes, etc. (see Ford 1993), which motivates
the syntactic distribution of causal adverbial phrases.
According to usage-based approaches to linguistic change, frequency of use of
linguistic forms is the driving force in grammaticalization (Bybee and Hopper
2001; Bybee 2006, 2007). Neither morphological nor semantic innovation needs to
be the starting point for change, but it is the preferential, salient discourse pattern
(Ariel 2008), which, via routinization, finally becomes entrenched as a cognitive
unit (cf. Langacker 2000). An example of this is the cline of PRO becoming clitic,
then morphosyntactic marking and finally zero. Another case in point is the increasing phonetic reduction of the French negative pre-verbal particle ne, so that
only the post-verbal pas is retained as marker of sentential negation. Entrenchment
has effects of its own, phonological reduction (leading to clitization, assimilation/
fusion, and elision; see above) being the most prominent one. Frequency has also
two other effects (Bybee 2006): conservation, and autonomy. The conservation effect is evidenced by the resistance of irregular forms to change because of their
high token frequencies. A case in point are irregular English past tense forms like
448
Arnulf Deppermann
was, came, had. They survive, while verbs with low token frequency successively
succumb to the high type-frequency of the regular, schematically derived past
tense morphology with -ed inflection (ex. wept f weeped). Autonomy (or emancipation) means the re-analysis of old and the emergence of new constructions,
which become used as unanalyzed units. They change and sometimes even cut
across syntactic categories, as could be seen in the case of gonna or the emergence
of discourse markers. Autonomy and unit status are also evidenced by increasing
speed of production and comprehension, prosodic segmentation from the discursive context and syntactic parsing as a unit. However, it is important to note that although frequency may play a major role in (the stabilization of) grammaticalization, this is not necessarily its cause. Rather, we need to ask why some pattern
becomes frequent. Motivations like economy (against redundancy of coding and
high predictability/inferrability in context) and others (see above) may cause or at
least support high frequency of use. Also, frequency does not automatically lead to
grammaticalization, because not all frequent patterns change, and the effect of frequency is mediated by factors like genre, speakers self-monitoring, linguistic
ideologies, etc.
3.10.
Acquiring grammatical constructions
Usage-based approaches show that the process of acquiring increasingly abstract

constructions rests on interactional experience and general cognitive abilities (Tomasello 2003; 2006). According to Tomasello (1999, 2003; Tomasello et al. 2005),
anthropological, social and cognitive prerequisites are the driving forces for the
childs construction of a language: the ascription of communicative (Gricean) intentions to interlocutors, the coordination of joint attention (Bruner 1983, Clark
2001), intersubjective cooperation (Tomasello 2008), and cognitive capabilities
like categorization, inference and induction of schemata, statistical comparison
and analogy, which are used for the identification of patterns of experience, action,
and linguistic practice. In contrast to Chomskys assumption about the poverty of
the stimulus, which is said to account for the need for a specialized language acquisition device, social interaction in terms of situated routines and modelling and
correction by care-givers in child-directed speech have also been shown to be of
major importance for language-learning (Cameron-Faulkner, Lieven and Tomasello 2003). Tomasello (2006: 258) questions the necessity to assume a languagespecific capacity, as Universal Grammar does, because children acquire the more
regular and rule-based constructions in the same way they acquire the more arbitrary and idiosyncratic constructions. Instead of applying abstract class-specific
rules to words according to their membership in syntactic categories (cf. Pinker
1999), knowledge about abstract syntactic structures emerges from lexical strings
by applying universal cognitive mechanisms of learning with the aid of the specific
pragmatic structures of context. The first schematic constructions to develop are
449
two-word pivot constructions, where a lexically fixed pivot is combined with various other items. The next step are item-based constructions and constructional islands (Tomasello 2003). These are constructional frames associated with specific
lexical items according to the frequency of their occurrence. They contain categorical slots for arguments and first syntactic markers (like word-order, morphosyntactic markers, prepositions). Children start to use and understand frequent
verbs in the context of various constructional frames, while less frequent verbs are
mastered only in a smaller number of constructional contexts (e.g., cut is only used
with direct object, while draw is also used with additional locative, instrumental or
beneficient prepositional phrases). So, syntactic rules do not seem to be categorically available in abstraction from lexicalization and context, but they are acquired
only incrementally by a process of increasing schematization. This process rests on
the discovery of analogies and variational patterns by comparison (see also Clark
1992 for the heuristics of conventionality and contrast). It lays the base for inferring slot-filling categories, which are increasingly abstracted and generalized to
fully schematic, abstract constructions via induction. The discovery of analogies
does not merely rest on perceptual features and cognitive operations over formal
categories, but more importantly on functionally based distributions, which cooccur with recurrent patterns of action (like object transfer or caused motion),
and interactional participation in action sequences (like referring and naming,
question-answer, narratives).
4.
Future prospects
As the paradigm of armchair linguistics gives way to empirical research,

scholars tend to conceive of pragmatics in a more comprehensive way, including
areas of research which were formerly assigned to text-, socio-, psycho- or discourse linguistics. Theoretical notions and methods originating from these areas
are increasingly integrated in pragmatic theorizing. In the last decade, a variety of
new empirical methods has emerged. They provoke changes in the view of the relationship between pragmatics and grammar and lead to new questions.
The increasing availability of large machine-readable (still overwhelmingly
written) corpora and the use of procedures of data-mining allows for the discovery of preferential patterns of co-occurrence and of contextual conditions of
use by statistical analyses (Stefanowitsch and Gries 2008). Such usage-based
studies strongly favor probabilistic and context-sensitive models of grammar
instead of universalistic rule-based approaches.
Video recordings of embodied multimodal interaction are prerequisites for
understanding the fine-grained sequential and simultaneous coordination between grammar, prosody, gesture, facial expression, and other body movements in real time. They also help to elucidate how the roots of linguistic rou-
450
Arnulf Deppermann
tines may lie in gestural routines, both ontogentically and phylogenetically

(Tomasello 2008).
Psycholinguistic experiments using parameters like latency of reaction are
used to test hypotheses about the psychological reality of alleged relationships
between (syntactically based) semantic vs. inferred meanings in terms of cognitive processing (Meibauer and Steinbach 2009).
Researchers have started to use physiological and neurobiological measures
like event-related potentials (Kutas and Federmeier 2007) and brain-imaging
(Bornkessel-Schlesewsky and Friederici 2007) as evidence for assumptions
about mental processing and cerebral localization of phenomena located on the
grammar-pragmatics interface.
The process of the acquisition of grammatical structures based on some basic
cognitive capabilities and discursive experience has started to be modelled by
learning robots, e.g., in the approach of Fluid Construction Grammar (Steels
and de Beule 2006).
Changes of the subject matter of linguistic pragmatics, which are due to sociocultural and technological developments, also call for revised models and raise
new questions. New internet and communication technologies, which are creatively adopted by users, lay the grounds for a seemingly endless rush of new
genres, multimodal configurations and modes of interactivity, in which new grammatical opportunities and routines in new communities of practice emerge. Processes of migration and globalization lead to revisions and re-formations of speech
communities, including the emergence of new linguistic varieties and inter-cultures of mixed codes which may become conventionalized.
Notes
1. See already the seminal paper by Fillmore, Kay and OConnor (1988).
2. Cf. ditransitive vs. to-dative (Goldberg 1995 and 2006) and argument omission (Goldberg 2005).
3. Ochs, Schegloff and Thompson (1996), Selting and Couper-Kuhlen (2001), Barth-Weingarten (2008).
4. Cf. e.g. Ono and Thompson (1995), Deppermann (2006, 2007), Gnthner (2006) and Imo
(2007).
5. Auer (1996a and 2007), Schegloff (1996), Ford, Fox and Thompson (2002), CouperKuhlen and Ono (2007a).
6. Crystal 2006; see also vol. 9 of the Handbook of Pragmatics handbook series: Pragmatics of Computer-mediated Communication, ed. by Susan C. Herring, Dieter Stein and
Tuija Virtanen.
451
References
Aitchison, Jean
2001
Language change: Progress or decay? 3rd ed. Cambridge: Cambridge University Press.
Ariel, Mira
1990
Accessing noun phrase antecedents. London: Routledge.
Ariel, Mira
2008
Pragmatics and grammar. Cambridge: Cambridge University Press.
Auer, Peter
1996a
On the prosody and syntax of turn-continuations. In: Couper-Kuhlen, Elizabeth and Selting, Margret (eds.), Prosody in conversation. Interactional
studies, 57100. Cambridge: Cambridge University Press.
Auer, Peter
1996b
The pre-front field in spoken German as a grammaticalization position. Pragmatics 6(3): 295322.
Auer, Peter
1998
Zwischen Parataxe und Hypotaxe: abhngige Hauptstze im Gesprochenen
und Geschriebenen Deutsch. Zeitschrift fr Germanistische Linguistik 26(3):
284307.
Auer, Peter
2005
Projection in interaction and projection in grammar. Text 25(1): 736.
Auer, Peter
2007
Why are increments such elusive objects? Pragmatics 17(4): 647658.
Auer, Peter
2009
On-line syntax: Thoughts on the temporality of spoken language. Language
Sciences 31: 113.
Auer, Peter and Stefan Pfnder (eds.)
2010
Emergent constructions. Berlin: de Gruyter.
Barlow, Michael and Suzanne Kemmer (eds.)
2000
Usage-based models of language. Stanford CA: Center for the Study of Language and Information.
Barth-Weingarten
2008
Interactional linguistics. In: Gerd Antos, Eija Ventola and Tilo Weber (eds.),
Handbook of Applied Linguistics. Vol. 2: Interpersonal Communication,
Barth-Weingarten, Dagmar and Couper-Kuhlen, Elizabeth
2002
On the development of final though: A case of grammaticalization? In: Ilse
Wischer and Gabriele Diewald (eds.), New reflections on grammaticalization,
Behaghel, Otto
1909
Beziehungen zwischen Umfang und Reihenfolge von Satzgliedern. Indogermanische Forschungen 25: 110142.
Biber, Douglas, Stig Johansson and Geoffrey Leech
1999
Longman grammar of spoken and written English. London: Longman.
Biber, Douglas
2006
University Language. Amsterdam: Benjamins.
452
Arnulf Deppermann
Bod, Rens, Jennifer Hay and Stefanie Jannedy (eds.)

2003
Probabilistic linguistics. Cambridge, MA: MIT Press.
Bornkessel-Schlesewsky, Ina and Angela D. Friederici
2007
Neuroimaging studies of sentence and discourse comprehension. In: Gareth
1996
Conceptual pacts and lexical choice in conversation. Journal of experimental
psychology: Learning, memory, and cognition 22: 14821493.
Bruner, Jerome
1983
Childs talk: Learning to use language. New York: Norton.
Bublitz, Wolfram, Uta Lenk, Uta and Eija Ventola (eds.)
1999
Coherence in spoken and written discourse. Amsterdam: Benjamins.
Bhler, Karl
1934
Sprachtheorie. Jena: Gustav Fischer.
Bring, Daniel
2007
Semantics, intonation and information structure. In: Gillian Ramchand and
Charles Reiss (eds.), The Oxford handbook of linguistic interfaces, 445474.
Bybee, Joan
2006
From usage to grammar: The minds response to repetition. In: Language
82(4): 711733.
Bybee, Joan
2007
Frequency and the organization of language. Oxford: Oxford University
Press.
Bybee, Joan and Paul J. Hopper (eds.)
2001
Frequency and the emergence of linguistic structure. Amsterdam: Benjamins.
Cameron-Faulkner, Thea, Elena Lieven and Michael Tomasello
2003
A construction based analysis of child directed speech. In: Cognitive Science
27: 843873.
Chafe, Wallace
1994
Discourse, consciousness, and time. Chicago IL: University of Chicago Press.
Chomsky, Noam
1965
Aspects of the theory of syntax. Cambridge, MA: MIT Press.
Chomsky, Noam and Howard Lasnik.
1993
The theory of principles and parameters. In: Jacobs, Joachim, von Stechow,
Arnim, Sternefeld, Wolfgang and Theo Vennemann (eds.), Syntax. Vol. 1,
Clark, Eve V.
1992
Conventionality and contrast: Pragmatic principles with lexical consequences.
In: Eva Feder Kittay and Adrienne Lehrer (eds.), Frames, fields, and contrasts:
New essays in semantic and lexical organization, 171188. Hillsdale, NJ:
Lawrence Erlbaum.
Clark, Eve
2001
Grounding an attention in the acquisition of language. In: M. Andronis,
C. Ball, H. Elston & S. Neuvel (eds.), Papers from the 37th meeting of Chicago
Linguistic Society. Vol. 1, 95116. Chicago: Chicago Linguistic Society.
453
Clark, Eve
2004
Pragmatics and language acquisition. In: Lawrence Horn and Gregory Ward
Clark, Herbert H.
1977
Bridging. In: Philip Johnson-Laird and Peter C. Wasow (eds.), Thinking,
Clark, Herbert H. and Deanna Wilkes-Gibbs
1986
Referring as a collaborative process. Cognition 22: 139.
2007a
Turn continuation in cross-linguistic perspective. Special Issue of Pragmatics
17(4).
2007b
Incrementing in conversation. A comparison of practices in English, German, and Japanese. Pragmatics 17(4): 513552.
Croft, William
2001
Radical construction grammar. Oxford: Oxford University Press.
Croft, William
2008
Construction Grammar. In: Dirk Geeraerts and Hubert Cuyckens (eds.), Handbook of Cognitive Linguistics, 463508. Oxford: Oxford University Press.
Croft, William and A. Cruse
2004
Cognitive linguistics. 225327. Cambridge: Cambridge University Press.
Crystal, David
2006
Language and the Internet. 2nd ed. Cambridge: Cambridge University.
Deppermann, Arnulf
2006
Construction Grammar Eine Grammatik fr die Interaktion? In: Arnulf Deppermann, Reinhard Fiehler and Thomas Spranz-Fogasy (eds.), Grammatik und
Interaktion, 4366. Radolfzell: Verlag fr Gesprchsforschung. Available at:
www.verlag-gespraechsforschung.de/2006/deppermann.htm.
Deppermann, Arnulf
2007
Grammatik und Semantik aus gesprchsanalytischer Sicht. Berlin: de Gruyter.
Deppermann, Arnulf
2010
Constructions vs. lexical items as sources of complex meanings. A comparative study of constructions with German verstehen. In: Peter Auer and Stefan Pfnder (eds.), Emergent constructions. Berlin: de Gruyter.
Diewald, Gabriele
2006a
Discourse particles and modal particles as grammatical elements. Approaches
to discourse particles. In: Kerstin Fischer (ed.), Approaches to discourse particles, 403425. Amsterdam: Elsevier.
Diewald, Gabriele
2006b
Context types in grammaticalization as constructions. In: Constructions HSV
1-9/2006 (Special Volume). Available at http://www.constructions-online.de/
articles/specvol1/686.
Dik, Simon C.
1997
The Theory of Functional Grammar. 2 vols. Berlin: de Gruyter.
Du Bois, John
2003
Discourse and Grammar. In: Michael Tomasello (ed.), The new psychology of
language. Vol. 2, 4787. Mahwah, NJ: Lawrence Erlbaum.
454
Arnulf Deppermann
Du Bois, John, Lorraine E. Kumpf and William J. Ashby (eds.)

2003
Preferred argument structure. Amsterdam: Benjamins.
Edwards, Derek
1997
Discourse and cognition. London: Sage.
Eggins, Suzanne
1994
An introduction into Systemic Functional Linguistics. London: Continuum.
Ehlich, Konrad
1991
Funktional-pragmatische Kommunikationsanalyse. In: Dieter Flader (ed.),
Verbale Interaktion, 127143. Stuttgart: Metzler.
1985
Frames and the semantics of understanding. Quaderni di Semantica 6(2):
222254.
Fillmore, Charles J., Paul Kay and Mary Catherine OConnor
1988
Regularity and idiomaticity in grammatical constructions: The case of let
alone. Language 64: 501538.
Feilke, Helmuth
1996
Sprache als soziale Gestalt. Frankfurt/Main: Suhrkamp.
Feilke, Helmuth
2004
Kontext Zeichen Kompetenz. In: Kathrin Steyer (ed.), Wortverbindungen
mehr oder weniger fest, 4164. Berlin: de Gruyter.
Fischer, Kerstin (ed.)
2006
Approaches to discourse particles. Amsterdam: Elsevier.
Ford, Cecilia E.
1993
Grammar in interaction. Adverbial clauses in American English conversations. Cambridge: Cambridge University Press.
2002
Constituency and the grammar of turn increments. In: Cecilia E. Ford, Barbara
A. Fox and Sandra A. Thompson (eds.), The language of turn and sequence,
2003
Social interaction and grammar. In: Michael Tomasello (ed.), The New Psychology of Language. Vol. 2, 119143. Mahwah NJ: Lawrence Erlbaum.
Fox, Barbara A.
2007
Principles shaping grammatical practices: an exploration. Discourse Studies
9(3): 299318.
Fries, Norbert
1983
Syntaktische und semantische Studien zum frei verwendeten Infinitiv und zu
verwandten Erscheinungen im Deutschen. Tbingen: Narr.
Gardner, Rod
2001
When listeners talk: Response tokens and listener stance. Amsterdam: Benjamins.
Givn, Talmy
1979
From discourse to syntax. Grammar as a processing strategy. In: Talmy Givn
(ed.), Syntax and semantics. Vol. 12: Discourse and syntax, 81112. London:
Academic.
Goldberg, Adele
1995
Constructions. A construction grammar approach to argument structure. Chicago IL: University of Chicago Press.
455
Goldberg, Adele
1998
Patterns of experience in patterns of language. In: Michael Tomasello (ed.),
The new psychology of language, 203220. Mahwah, NJ: Lawrence Erlbaum.
Goldberg, Adele
2004
Pragmatics and argument structure. In: Lawrence Horn and Gregory Ward
Goldberg, Adele
2005
Argument realization. The role of constructions, lexical semantics and discourse factors. In: Jan-Ola stman and Mirjam Fried (eds.), Construction
grammar, 1743. Amsterdam: Benjamins.
Goldberg, Adele
2006
Constructions at work: the nature of generalization in language. Oxford: Oxford University Press.
Goodwin, Charles
2003
Pointing as situated practice. In: Sotaro Kita (ed.), Pointing: Where language,
culture and cognition meet. 21741. Mahwah, NJ: Lawrence Erlbaum.
Goodwin, Charles and Marjorie Harness Goodwin
1987
Concurrent operations on talk: Notes on the interactive organization of assessments. IPRA Papers on Pragmatics 1(1): 154.
Green, Georgia M.
2004
Some interactions of pragmatics and grammar. In: Lawrence Horn and Gregory Ward (eds.), Handbook of pragmatics, 407426. Oxford: Blackwell.
Grice, H. Paul
1975
semantics. Vol. 3, 4158. New York: Academic Press.
Gundel, Jeanette and Thorstein Fretheim
2004
Topic and focus. In: Lawrence Horn and Gregory Ward (eds.), Handbook of
pragmatics, 175196. Oxford: Blackwell.
Gundel, Jeanette, Nancy Hedberg and Ron Zacharsky
1993
Givenness, implicature, and the form of referring expressions in discourse.
Gnthner, Susanne
2006
Von Konstruktionen zu kommunikativen Gattungen. Die Relevanz sedimentierter Muster fr die Ausfhrung kommunikativer Aufgaben. Deutsche
Sprache 1(2), 173190.
Gnthner, Susanne
2008
Projektorkonstruktionen im Gesprch: Pseudoclefts, die Sache ist-Konstruktionen und Extrapositionen mit es. Gesprchsforschung 9: 86114. Available
at: http://www.gespraechsforschung-ozs.de/heft2008/ga-guenthner.pdf.
Gnthner, Susanne and Paul J. Hopper
2010
Zeitlichkeit & sprachliche Strukturen: Pseudoclefts im Englischen und Deutschen. Gesprchsforschung 11: 128. Available at:
www.gespraechsforschung-ozs.de/heft2010/ga-guenthner.pdf
Haiman, John (ed.)
1985
Iconicity in syntax. Amsterdam: John Benjamins.
1985
An Introduction into Functional Grammar. London: Edward Arnold.
456
Arnulf Deppermann
Haspelmath, Martin
1997
From space to time. Temporal adverbials in the worlds languages. Mnchen:
Lincom.
Haspelmath, Martin
1999
Why is grammaticalization irreversible? In: Language 37: 10431068.
Hayashi, Makoto
2003
Joint utterance construction in Japanese conversation. Amsterdam: Benjamins.
Heine, Bernd
1997
Cognitive foundations of grammar. Oxford: Oxford University Press.
Heine, Bernd, Ulrike Claudi and Friederike Hnnemeyer
1991
Grammaticalization: A conceptual framework. Chicago: University of Chicago Press.
Hengeveld, Kees and J. Lachlan Mackenzie
2008
Functional Discourse Grammar. Oxford: Oxford University Press.
Hoffmann, Ludger
2001
Einleitung. In: Ludger Hoffmann (ed.), Funktionale Syntax, 117. Berlin: de
Gruyter.
Hopper, Paul J.
1998
Emergent grammar. In: Michael Tomasello (ed.), The new psychology of language, 155175. Mahwah, NJ: Erlbaum.
Hopper, Paul J.
2004
The openness of grammatical constructions. Papers from the 40th regional
meeting of the Chicago Linguistic Society 40: 239256.
Hopper, Paul J. and Elizabeth Closs Traugott
2003
Grammaticalization. 2nd ed. Cambridge: Cambridge University Press.
Hopper, Paul J. and Sandra A. Thompson
1984
The discourse basis for lexical categories in Universal Grammar. In: Language
60: 703752.
Hopper, Paul and Sandra A. Thompson
2008
Projectability and clause combining in interaction. In: Ritva Laury (ed.),
Crosslinguistic studies of clause combining, 99123. Amsterdam: Benjamins.
Imo, Wolfgang
2007
Konstruktionsgrammatik und Gesprochene Sprache-Forschung. Tbingen:
Niemeyer.
Jefferson, Gail
1973
A case of precision timing in ordinary conversation: Overlapped tag-positioned address terms in closing sequences. Semiotica 9: 4796.
Karkkinen, Elise
2003
Epistemic stance in English conversation. Amsterdam: Benjamins.
Kay, Paul
2004
Pragmatic aspects of grammatical constructions. In: Lawrence Horn and Gregory Ward (eds.), Handbook of pragmatics, 675700. Oxford: Blackwell.
Keller, Rudi
2004
On language change. London: Routledge.
2001
Multimodal discourse. London: Edward Arnold.
457

2006
Reading images: The grammar of visual design. London: Routledge.
Krifka, M.
2007
Basic notions of information structure. In: Caroline Fry, Gisbert Fanselow
and Manfred Krifka (eds.), Interdisciplinary studies on information structure
(ISIS) 6. Working Papers of the SFB 632, 1356. Potsdam: Universittsverlag.
Kutas, Marta and Kara D. Federmeier
2007
Event-related brain potential (ERP) studies of sentence processing. In: Gareth
Labov, William
2001
Principles of linguistic change. Vol. 2. Oxford: Blackwell.
Lakoff, George
1987
Women, fire and dangerous things. What categories reveal about the mind.
Chicago IL: University of Chicago Press.
Lambrecht, Knud
1987
On the status of SVO sentences in French discourse. In: Russell S. Tomlin
(ed.), Coherence and grounding in discourse, 217262. Amsterdam: Benjamins.
1987
Foundations of cognitive grammar. Vol. 1: Theoretical prerequisites. Stanford
CA: Stanford University Press.
2000
A dynamic usage-based model. In: Michael Barlow and Suzanne Kemmer
(eds.), Usage-based models of language, 160. Stanford, CA: Center for the
Study of Language and Information.
2001
Discourse in cognitive grammar. Cognitive Linguistics 12(2): 143188.
2008
Cognitive grammar. Oxford: Oxford University Press.
Lass, Roger
1987
Historical linguistics and language change. Cambridge: Cambridge University Press.
LePage, Robert B. and Andre Tabouret-Keller
1985
Acts of identity: Creole-based approaches to language and ethnicity. Cambridge: Cambridge University Press.
Lerner, Gene
1991
On the syntax of sentences in progress. Language in Society 30: 441458.
Lerner, Gene
1996
On the semi-permeable character of grammatical units in conversation: Conditional entry into the turn space of another speaker. In: Elinor Ochs, Emmanuel
A. Schegloff and Sandra A. Thompson (eds.), Interaction and grammar, 238
Lerner, Gene
2004
The place of linguistic resources in the organization of talk-in-interaction:
Grammar as action in prompting a speaker to elaborate. Research on language
and social interaction 37(2): 151184.
458
Arnulf Deppermann
Liedtke, Frank
1998
Grammatik der Illokution. Tbingen: Narr.
MacWhinney, Brian and Elizabeth Bates
1989
The crosslinguistic study of sentence processing. Cambridge: Cambridge University Press.
Meibauer, Jrg and Markus Steinbach (eds.)
2009
Experimental-pragmatics/semantics. Amsterdam: Benjamins.
Mulder, Jean and Sandra A. Thompson
2008
The grammaticalization of but as a final particle in English conversation. In:
Ritva Laury (ed.), Crosslinguistic studies of clause combining, 179204.
1998
Language form and language function. Cambridge MA: MIT Press.
2003
Grammar is grammar and usage is usage. Language 79: 682707.
Norris, Sigrid
2004
Analyzing multimodal interaction. London: Routledge.
Ochs, Elinor, Emmanuel A. Schegloff, and Sandra A Thompson (eds.)
1996
Ono, Tsuyoshi and Sandra A. Thompson
1995
What can conversation tell us about syntax? In: Philip W. Davis (ed.), Alternative linguistics, 213271. Amsterdam: Benjamins.
Pekarek-Doehler, Simona
2010
Clause-combining and the sequencing of actions: projector fragments in
French conversation. In: Ritva Laury and Ryoko Suzuki (eds.), Subordination
in conversation. Crosslinguistic analysis of form-function mappings, Amsterdam: Benjamins.
Pickering, Martin J. and Simon C. Garrod
2004
Towards a mechanistic psychology of dialogue. Behavioral and Brain
Sciences 27: 169226.
Pinker, Steven
1999
Words and rules. New York: Morrow.
Prince, Ellen F.
1978
A comparison of WH-clefts and IT-clefts in discourse. In: Language 54: 883906.
Prince, Ellen F.
1981
Toward a taxonomy of given/new information. In: Peter Cole (ed.), Radical
pragmatics, 223255. New York: Academic Press.
Raymond, Geoffrey
2004
The stand-alone so in ordinary conversation. In: Research on Language and
Social Interaction 37(2): 185218.
Redder, Angelika
2008
Functional pragmatics. In: Gerd Antos and Eija Ventola (eds.), Handbook of
applied linguistics. Vol 2. Interpersonal communication, 133178. Berlin: de
Gruyter.
Sacks, Harvey and Emmanuel A. Schegloff
1979
Two preferences in the organization of reference to persons in conversation
and their interaction. In: George Psathas (ed.), Everyday language, 1521.
New York: Irvington.
459
Sacks, Harvey, Emmanuel A. Schegloff and Gail Jefferson

1974
A simplest systematics for the organisation of turn-taking in conversation.
Language 50(4): 696735.
Schegloff, Emmanuel A.
1996
Ochs, Emmanuel A. Schegloff and Sandra A. Thompson (eds.), Interaction
and grammar, 52133. Cambridge: Cambridge University Press.
Schiffrin, Deborah
1987
Schlobinski, Peter (ed.)
2006
Von *hdl* bis *cul8er*. Sprache und Kommunikation in den neuen Medien.
Mannheim: Duden.
Schmid, Hans-Joerg
2001
Presupposition can be a bluff: How abstract nouns can be used as presupposition triggers. Journal of Pragmatics 33(10): 15291555.
Schmitz, Ulrich
2010
Schrift an Bild im World Wide Web. In: Arnulf Deppermann and Angelika
Linke (eds.), Sprache intermedial, 383418. Berlin: de Gruyter.
Searle, John R.
1969
Searle, John R. and Daniel Vanderveken
1985
Foundations of illocutionary logic. Cambridge: Cambridge University Press.
Selting, Margret
2000
The construction of units in conversational talk. Language in Society 29:
477517.
2001
Studies in interactional linguistics. Amsterdam: Benjamins.
Sorjonen, Marja-Leena
2001
Responding in conversation: A study of response particles in Finnish. Amsterdam: Benjamins.
Steels, Luc and Joachim de Beule
2006
A (very) brief introduction to Fluid Construction Grammar. Proceedings of the
3rd International Workshop on Scalable Natural Language (ScaNaLu06):
7380. Available at: http://www3.isrl.illinois.edu/~junwang4/langev/localcopy/pdf/Steels06Intro_FCG.pdf.
Stefanowitsch, Anatol and Stefan Th. Gries
2008
Corpora and grammar. In: Anke Ldeling and Merja Kyt (eds.), Corpus linguistics. Vol. 2, 933951. Berlin, New York: de Gruyter.
Stckl, Hartmut
2004
Die Sprache im Bild Das Bild in der Sprache. Berlin: de Gruyter.
Streeck, Jrgen
2002
Grammars, words, and embodied meanings. On the evolution and uses of so
and like. Journal of Communication 52(3): 581596.
Sweetser, Eve E.
1990
From etymology to pragmatics. Cambridge: Cambridge University Press.
Thompson, Sandra A.
2002
Object complements and conversation. Towards a realistic account. Studies
in Language 26(1): 124164.
460
Arnulf Deppermann
Thompson, Sandra A. and Elizabeth Couper-Kuhlen

2005
The clause as a locus of grammar and interaction. Discourse Studies 7(45):
481505.
Tomasello, Michael
1999
The cultural origins of human cognition. New York: Harvard University Press.
Tomasello, Michael
2003
Constructing a language: A usage-based theory of language acquisition,
94281. New York: Harvard University Press.
Tomasello, Michael
2006
Acquiring linguistic constructions. In: Deanna Kuhn and Robert S. Siegler
(eds.), Handbook of child psychology. Vol. 2, 255291. New York: Wiley.
Tomasello, Michael
2008
Origins of human communication. Cambridge MA: MIT Press.
Tomasello, Michael, Malinda Carpenter, Josep Call, Tanya Behne and Henrike Moll
2005
Understanding and sharing intentions: The origins of cultural cognition. Behavioral and Brain Sciences 28: 675691.
2004
Historical pragmatics. In: Lawrence Horn and Gregory Ward (eds.), Handbook
of pragmatics, 538561. Oxford: Blackwell.
Uhmann, Susanne
1991
Fokusphonologie. Eine Analyse deutscher Intonationskonturen im Rahmen der
nicht-linearen Phonologie. Tbingen: Niemeyer.
Ward, Gregory and Betty Birner
2004
Information structure and non-canonical syntax. In: Lawrence Horn and Gregory Ward (eds.), Handbook of pragmatics, 153174. Oxford: Blackwell.
Wasow, Tom
2002
Postverbal behavior. Stanford: CSLI.
16.

John Saeed
1.
Introduction
In this article we focus on some relatively recent ideas and discussion about the
relationship between pragmatics and semantics. As a result we do not seek to characterize either enterprise comprehensively, or symmetrically, even if such were
possible in this limited space. Hence our discussion omits consideration of major
concerns of both: for pragmatics, for example, of the social and cultural dimensions of language use. In particular speech-act theorists are underrepresented, despite the importance to pragmatics of their concern with social institutions, conventions and language activities.1
Characterizing the relationship between pragmatics and semantics is made difficult by the differences in views of these fields in different linguistic theories.
Moreover, other disciplines are keen to explore the same territory. Philosophers
Lepore and Smith, for example, in prefacing a handbook of the philosophy of language, state aims that many linguists would recognize:
[] the philosophy of language now focuses on its primary concern: the nature of natural language and the extraordinary capacity of human beings to use it to express their
thoughts about the world and other subject matters. The way language works, how specific linguistic devices function to achieve their effects, how we come to know these
properties of expressions, and how we exploit them in our talk: all this is pursued by
contemporary philosophers of language. (2006: viii)
Looking the other way across the disciplinary fence, linguists may profess scepticism about the philosophical method, alleging a dependence on introspection, a
concentration on a few European languages and scant attention to other linguistic
levels such as phonology and syntax. However, the views of Frege, Russell, the
early Wittgenstein, Quine, Davidson and other philosophers have been of fundamental importance in many linguistic theories of semantics, in particular in pursuit
of the idea that languages are, or can be profitably viewed as, formal systems. As
will be clear from other contributions to this handbook, the ideas of the later
Wittgenstein, Grice, Austin, Strawson and Searle have been fundamental in the development of pragmatics.
More generally, the influence of philosophy on linguistic semantics can be seen
in the search for a role for propositions in the representation of sentence meaning.2
The idea that semantics is concerned with representations is widespread, even in
functional approaches, along with the associated idea that pragmatics is then concerned with interpersonal action.3 Problems with determining how semantic repre-
462
John Saeed
sentations relate to propositions have been crucial to the developing views of what
role pragmatics has in describing meaning. In the second half of the last century it
became clear to many scholars that if propositions are bearers of determinate truth
or falsity then the literal meanings of sentences cannot be equivalent to propositions. Typically when uttered, sentences dont provide enough information: they
leave a gap that has to be bridged by the listener. Increasingly the listeners interpretive activity has been characterized as pragmatic.
2.
Context and propositional meaning
Various types of gap have been identified. Perhaps the clearest case is deixis, such
as the italicized deictic expressions (or indexicals): Youre here now. Such expressions of course crucially require contextual information to be accessed, and
thus their linguistic meanings do not seem to be part of the content of the utterances
that contain them. They form part of what Gumperz (1992) calls contextualization
cues, which invoke the relevant contextual assumptions in the participants.
Some writers, following Kaplan (1989a, 1989b) have identified a split in the
meaning of such expressions between a context independent element, termed
character by Kaplan, and the context dependent meaning, which he termed content. The character determines how the context will determine the content: in
other words how the intended reference is identified and thus contributes to determining what proposition is expressed. A further distinction is then made between
two types of indexical: the first are pure indexicals like I and yesterday, for which
the contextual determination is thought to be straightforward or automatic, in the
terms of Perry (2001) and which are independent of speakers intentions. The
others (termed true demonstratives by Kaplan 1989a), like he and that (used deictically rather than anaphorically), require inferential work for the addressees to be
able to identify the referent. This latter set is called discretionary by Perry (2001).
This difference might be significant for setting the pragmatics-semantics boundary
since the automatic reference identification could be assumed to be a semantic process, akin to literal meaning, while the discretionary determination of reference
could be held to be pragmatic, involving inferences about speaker intentions (see
King and Stanley 2005 for discussion).
One problem is that the automatic set seems to shrink on examination. Traditionally English here and now have been considered pure indexicals. However
though we might want to say that here includes the position of the speaker and now
the time of utterance, in fact the determination of the zones these refer to seems to
depend on context-dependent inferences, sensitive to the speakers intent. Moreover, linguistic studies of richer deictic systems than we find in English have revealed referential grids that involve both literal and inferential elements of meaning, as in Yupik (Eskimo-Aleut; Alaska):
463
Yupik demonstratives (Anderson and Keenan 1985: 295)

Extended
Restricted
Obscured
mana
una
this (near SP)
tamana
tauna
that (near ADR)
imna
the aforementioned one
ukna
the one approaching the speaker
augna
ingna
amna
the one going away from the speaker
agna
ikna
akemna
the one across there
qaugna
kiugna
qamna
the one inland, inside, upriver
qagna
keggna
qakemna
the one outside
una
kana
camna
the one below, toward river
unegna
ugna
cakemna
the one downriver, by the exit
paugna
pingna
pamna
the one up there, away from river
pagna
pikna
pakemna
the one up above
This system incorporates a semantic classification of the objects to which the demonstratives refer: extended forms are for either large expanses of land or water,
or objects that are lengthy or moving; restricted applies to objects that are stationary, or moving within a confined area, and fairly small in extent, relatively near,
and visible; and obscured describes objects that are farther away and not clearly
in sight. At the same time the demonstratives crucially rely on the position of the
speaker and so are deictic (see Anderson and Keenan 1985 for details). Hanks
(1992) argues that in many languages deictic expressions, whose core function is to
identify the referent through different relations to elements of the speech act, are
typically fused with other information about the category of referents, the speakers attitude or stance and the intended speech act.
Another type of gap between sentences and propositions is caused by the need
for contextual disambiguation. Ambiguity, which may be structural as in (1), involve quantifier scope as in (2), or be lexical as in (3), results in a sentence fitting
two or more propositions and thus requires the audience to determine which fit is
more appropriate for the context:
1. He tracked the man with binoculars.
2. All reviewers watched one movie.
3. She held the note.
The ambiguity in (1) depends on whether the prepositional phrase with binoculars
is interpreted as modifying the verb track or the noun phrase the man. Both interpretations are possible but describe different situations. In (2) the ambiguity derives from interpretations of the relative scope of the quantifiers all and one: if all
is given wide scope over one, then there may be as many films as reviewers; if one
has scope over all then they all want to see the one film. While there have been attempts to determine such scope ambiguities by semantic or logical rules, other
writers (for example, Saba and Corriveau 2001: 276) suggest that quantifier scope
464
John Saeed
interpretation must involve a pragmatic level. One type of evidence is the influence
of real world knowledge. Thus while the sentence in (4) has the predicted scope
ambiguity (One and the same doctor read all the books / Each book was read by a
possibly different doctor), this typically doesnt arise in (5) because of what we
know about the world:
4. A doctor read every book.
5. A doctor lives in every village.
Lexical ambiguity as in (3), where the noun note can of course be a musical sound
or a written message and consequently different senses of held are invoked, arises
where the same phonological shape is shared by two or more distinct semantic
units (or lexemes). A distinction is often made between homonymy, like pump device for moving liquid or gas and pump shoe where the semantic units are clearly
distinct, and polysemy, where different but related senses are involved, like fault in
Whose fault was the banking crisis? and The San Andreas Fault doesnt go
through San Francisco. The decisions about relatedness are, however, not without
difficulty (Lyons 1977: 55069). The question of how the multiple senses are accessed and selected in context has been an important research issue in the psychology of language (see Simpson 1984, 1994 for discussion). It has also proved to be a
practical problem for natural language processing, for example in machine translation, and the investigation of the relevant contextual information for lexical disambiguation has been an important research strand in computational linguists, as
described in Agirre and Edmonds (2006).
Remaining with quantifiers it has been widely noted (e.g. Recanati 1989,
1993b, Bach 1994, Bring 1997) that in normal use the domain of a quantifier is
typically underdetermined and needs to be contextually enriched. So for the second
sentence in (6a) below, Recanati (2000: 678) suggests the representation in (6b),
where the leftmost square bracket identifies the situation or context to which the utterance relates:
6. a. Berkeley is a nice place. There are bookstores and coffee shops everywhere.
b. [Berkeley] << There are bookstores and coffee shops everywhere>>
Although in this case the domain is mentioned in the previous sentence, in commonly encountered examples like (7) below, the relevant domain is left for the
audience to supply from context.
7. Are you going to the party? Everyone will be there.
If quantifier restriction is freely supplied by contextual information then many
writers have concluded that the semantic content of the utterance underspecifies
the intended proposition.4
In a similar vein, several writers, including Sadock (1984), Horn (1992), and
Carston (1998), have argued about the extent of pragmatic enrichment needed to
465
explain uses of cardinal numbers, which may in different contexts for a cardinal
number x have the intended interpretations, exactly x as in (8) below, at least x in
(9), or at most x in (10):
8. He had six children but one died so now he has five.
9. She needs to reach thirty metres to make the javelin final.
10. Undergraduates may borrow six books from the library.
Another type of underspecification is shown by gradable adjectives (Kennedy
2007), for example, like tall, expensive, and fast, which require contextual fill in
to be interpreted. This is clear in implicit comparison uses like:
11. Isabel is tall.
12. Restaurants in Dublin are expensive.
Participants have to access contextually determined standards to interpret such
examples: Isabel might be tall for a seven year old child, for example. Context update models (Kyburg and Morreau 2000, Barker 2002) provide a presuppositional
account of this process. A related phenomenon is loose talk, argued by Lasersohn
(1999) to be distinct from vagueness. This is where speakers employ a pragmatically licensed degree of approximation in their use of words, even where the terms
are quite sharply defined. Thus a speaker might in casual conversations say (13)
below:
13. Ive a three am. flight tomorrow.
even if the flight is actually at 3.05 am. If the context is right, the inaccuracy is
treated as pragmatically irrelevant.
As pragmatic research has developed over the past decades more examples of
such under-determinacy have been identified. Typical examples include possessive
constructions as in (14); understood time references in (15); and the strengthening
of and in conjoined subjects as in (16):
14. The conversation turned to Johns book.
15. Ive had lunch.
16. John and Mary lifted the piano.
In (14) the relationship described by the possessive pronouns is contextually determined: the book he wrote, the one he owns, the one he was assigned to review, etc.
If (15) is a reply to an invitation, it communicates that the speaker has eaten lunch
within a contextually appropriate period of time (e.g. that day). In (16) there may
be a stereotypical or default interpretation that John and Mary lifted the piano together (Levinson 2000).
The increasing role of contextual inference in the determination of the truthdeterminate propositional meaning expressed by utterances has been called pragmatic intrusion (Levinson 2000). The growing emphasis on the context sensitivity
466
John Saeed
of language raises fundamental challenges to philosophical approaches which

base theories of meaning on theories of truth. The varying extent to which philosophers permit the intrusion of contextual information into semantics can be
viewed as grades of contextualism (Recanati 2004, 2005) as discussed in Preyer
and Peter (eds. 2005, 2007). Some philosophical defences against contextualism
have been proposed, for example by Borg (2004) and Cappelen and Lepore
(2005). These authors defend formal semantic approaches by proposing versions
of semantic minimalism, where context dependence is viewed as limited to cases
triggered by a limited set of linguistic forms. In Cappelen and Lepore (2005) this
set of forms is reminiscent of Kaplans (1989a, 1989b) indexicals, mentioned earlier. Their account pushes contextual variation into what they call speech act
pluralism. In this view, a sentence without one of the context-sensitive elements
always expresses the same proposition, regardless of context. However when uttered in context such a sentence may be expanded to express infinitely many propositions, only some of which are relevant to the context, and it is these expanded
propositions that motivate truth-value judgments. In other words semantic content
and speech act content can be different and, although the former is viewed as a
proposition, it is the latter that we have truth condition intuitions about. The cost
of this, from a linguistics perspective, is a weakening of the role of intuitions
about semantic content, which might risk the empirical underpinning of the study
of meaning.
Other philosophers of course are not troubled by a separation of semantic representations from truth conditions and propositions. The philosopher John Searle
adopts a contextualist position:
We know without hesitation what are the truth conditions of The fly is on the ceiling,
but not of The cat is on the ceiling, and this difference is not a matter of meaning, but a
matter of how our factual background information enables us to apply the meanings of
sentences. In general, one can say that in most cases a sentence only determines a set of
truth conditions relative to a set of assumptions that are not realized in the semantic content of the sentence. (Searle 1979: 86)
Searle argues that background assumptions have to be incorporated before a determinate proposition can be communicated. In particular, in this view, an addressee
cannot respond to the intended speech act, say an order, until this is done.
Noam Chomsky, whose view of a semantic representation concentrates on an
individuals mind-internal mapping between grammar and conceptual structure,
shows even greater scepticism about a role for truth conditional denotation in semantic analysis: the argument for a reference-based semantics (apart from an internalist syntactic version) seems to me weak. It is possible that natural language has
only syntax and pragmatics (1995: 26).
Chomskys view is that the meanings of sentences do not specify truth conditions even when relativized to context. This is reminiscent of Strawsons (1950)
view that although sentences can be used to assert things that are true or false, sen-
467
tences themselves are neither true nor false. Pietroski (2005: 256) characterizes
Chomskys position as follows:
Rather the meaning of S is a compositionally determined intrinsic property of S that
constrains and guides without determining how S can be used to make true or false assertions in various conversational situations making truth-evaluable assertions is one
of the things we can do with sentences, in contexts, though uses of this kind are highly
variable; while people refer to things, words dont; and sentence use may not be a theoretically tractable phenomenon.
Chomsky (1995) effectively proposes a break in the link assumed by many philosophers between the truth-conditional semantics of artificial languages and accounts of meaning for natural languages. In doing this he takes a use-based theory
deep into traditional semantic territory and makes sense of the dramatic suggestion
in the quotation above.
3.
Neo-Gricean inferential pragmatics
Within contemporary linguistics, debate about the relation between pragmatics and
semantics has followed from Paul Grices (1975, 1978, 1989) influential distinction between what is said in an utterance and what a speaker means by an utterance.5 This is often characterised (see for example Carston 2004) as a distinction
between what is said as propositional, or truth-conditional, content, and what is
intentionally implied by the speakers utterance. As noted by several writers (e.g.
Szab 2006: 386), Grice never quite seems to give us a tight definition of this notion of what is said, identifying it as closely related to the conventional meaning
of the words (the sentence) he has uttered (1989: 25). This leaves it somewhat
open as to how much resolution of context-sensitive expressions, such as indexicals, is necessary to reach what is said. The distinction in any case depends on his
notion of conversational implicature as implications that are not part of the conventional meaning of the utterance, are context-dependent and, crucially, intended
by the speaker, as in a typical (invented) example (17):
17. A: Would you like another drink?
B: Id better go home; Ive an early start in the morning.
Here Bs reply may be interpreted to implicate a refusal or simply no.
Grices work suggests a distinction between semantics and pragmatics: semantics is concerned with the proposition expressed while pragmatics is concerned with the implicatures of the utterance. His work opens up the prospect of
an inferential pragmatics that explores the gap between the linguistic content of
the speakers utterance, when viewed analytically, on the one hand, and what listeners seem to understand as speakers meaning. Pragmatic processing relies on
468
John Saeed
the predictability of inferential behaviour, seen by Grice as depending on co-operative principles underlying communication (Grice 1975, 1978). His cooperative
principle and maxims below predict that the listener assumes, and the speaker
relies on this, that the speaker is speaking truthfully, relevantly, and appropriately,
that is performing subtasks of assessing what the hearer knows, selecting the right
degree of clarity, etc. Clearly this is not a claim that speakers always behave in
this way; rather it is a claim that inferential strategies seem to rely on such assumptions.
Grices Co-operative Principle (see Grice 1989:26)
Make your contribution such as is required, at the stage at which it occurs, by
the accepted purpose or direction of the talk exchange in which you are engaged.
Grices maxims (adapted from Grice 1975: 46)
i. The Maxim of Quality
Try to make your contribution one that is true, i.e.
a) do not say what you believe is false
b) do not say that for which you lack adequate evidence.
ii. The Maxim of Quantity
Make your contribution as informative as is required for the current purposes
of the exchange (i.e. not more or less informative).
iii The Maxim of Relevance
Make your contributions relevant.
iv. The Maxim of Manner
Be perspicuous, and specifically:
a) avoid ambiguity
b) avoid obscurity
c) be brief
d) be orderly.
His approach moreover promises an account of pragmatic processing. Grice proposed that conversational implicatures must be capable of being worked out, i.e. a
reasoning process must be identifiable, even if it is not deduction. For every implicature one can predict a process of calculation involving what is said (literal
meaning), the cooperative principle and context. A characteristic Gricean formulation is:
Working out schema for conversational implicatures
(a) The speaker (S) has said that p.
(b) There is no reason to think that S is not observing the maxims.
(c) S could not be doing this unless he thought that q.
(d) S knows (and knows that the hearer (H) knows that he knows) that H can
see that he thinks that the supposition that he thinks q is required.
469
(e) S has done nothing to stop H from thinking that q.

(f) S intends H to think, or is at least willing to allow H to think, that q.
(g) And so, S has implicated that q. (adapted from Grice 1975: 50)
Grice used the term non-natural meaning (meaningnn) for interactional communication between mutually aware participants to distinguish it from simpler forms of
signification. He provided an initial characterization of this as below:
Figure 1: Grices Meaningnn (Levinson 2000: 13)
As suggested above, the first split in Figure 1 above might be taken to coincide
with the Gricean division between semantics and pragmatics, though Grice never
puts it in these terms, since what is said is presumably the truth-conditional content. However the notion of conventional implicature causes problems here since it
is conventionally marked or coded but inferentially determined. A typical example
is English but, which as shown below consistently carries a meaning of unexpectedness or contrast:
18. a. Hes a politician but hes honest.
b. Joan is tall but Libby is short.
c. But you know I dont drink coffee!
Despite this degree of conventionality, the meaning of but is excluded from what
is said because it does not contribute to the truth conditions of the proposition expressed. It also requires contextual inferences to determine the precise interpretation. Subsequent work in this general framework has described such behaviour
without this notion of conventional implicature and, as we shall see below, more
generally has suggested different borders between semantics and pragmatics.
The pragmatic inferences called conversational implicatures are claimed to
have specific features. They are described, for example, as cancellable, non-detachable and non-conventional. They are called cancellable because, for example,
we can add in (19) below another proposition to cancel the implicature without
causing anomaly:
470
John Saeed
19. a. John has three children, in fact ten.

b. John has three children, maybe more.
This contrasts with other types of inference like presupposition in (20), and entailment in (21), where anomaly results:
20. ?The King of France is bald and/but theres no King of France.
21. ?The president was assassinated yesterday but hes not dead.
Non-detachability is the claim that they are attached to a semantic level rather than
to surface linguistic form. Thus by changing the specific linguistic form, say via
synonyms, there is no change in the implicatures. Levinson (1983: 116117)
claims that if we know a context where (22a) implicates (22b) then (23a-d) will
also produce (22b):6
22. (a) Johns a genius.
(b) John is an idiot.
23. (a)
(b)
(c)
(d)
Johns a mental prodigy.

Johns an exceptionally clever human being.
Johns an enormous intellect.
Johns a big brain.
Finally non-conventionality captures the fact that the inference is not encoded in
the linguistic form of the utterance, as is clear from examples like (17) earlier.
To complete the Gricean taxonomy: particularized conversational implicatures
(PCIs) depend completely on specific context of utterance: they are once-off inferences. Generalized conversational implicatures (GCIs) are context dependent but
range over contexts, like a default. They only fail if blocked by inconsistency with
context. Carston (2004), following Levinson (2000), provides the following
example to illustrate the difference:
24. A: Did the childrens summer camp go well?
B: Some of them got stomach flu.
GCI: Not all the children got stomach flu.
PCI: The summer camp didnt go as well as hoped.
This Gricean conception of pragmatics has inspired an extensive literature on the
nature of both the maxims and the implicature themselves (cf. Huang this volume).
Horn (1984, 1989, 2004) for example argues for a reworking of the Gricean
maxims into two basic pragmatic principles, briefly outlined below:
Horns Q- and R-principles (Horn 1984, 1989)
(a) The Q-principle
Make your contribution sufficient;
Say as much as you can (given the R-principle)
471
(b) The R-principle

Make your contribution necessary;
Say no more than you must (given the Q-principle)
Horns Q-principle conflates Grices Quantity maxim and parts 1 and 2 of Grices
Manner maxims and can basically be paraphrased as say the most that you are
licensed to say. Alternatively, taking the hearers point of view in mind, it could be
paraphrased as maximize the informational content of what you say. On the other
hand the R-principle is a speaker-orientated economy principle. The two speaker
versus hearer economy principles are held to be in tension. The best known application of the Q-principle is the Q- or Horn scales of strength, schematized below:
Q-scale/Horn scale (Huang 2007: 38)
For <S, W> to form a Q- or Horn scale,
(i) A(S) entails A(W) for some arbitrary sentence frame A;
(ii) S and W are equally lexicalized, of the same word class, and from the same
register; and
(iii) S and W are about the same semantic relation, or from the same semantic
field.
Thus Horn scales are typically scales of alternates, ordered strong to weak, for
example:
25. (a) <all, some>
(b) <identical, similar>
The Q-maxim operating on scales ensures that the use of a weaker alternate implies
the negative of a stronger:
26. a. Not all the listeners enjoyed the show implicates some (i.e. not none)
did
b. Their results are similar implicates Their results are not identical
A similar but distinct revision is attempted by Levinson (1987a, 1987b, 1991,
2000), who, seeking to distinguish semantic content from surface form in relation
to economy, proposes three principles: the Q-, I-, and M-principles. The Q-principle, which corresponds to Grices Maxim of Quantity 1, is given in simplified
form below:
Levinsons Q-principle (simplified; Levinson 2000, Huang 2007)
Speaker: Do not say less than is required (bearing in mind the I-principle)
Addressee: What is not said is not the case.
This principle, like Horns Q-principle, gives rise to scales. In addition to the
simple scales as in (25) above, Levinson identifies more complex cases of alternate
sets. We can briefly mention two. The first are clausal, where the alternates are dis-
472
John Saeed
tinguished by the stronger entailing its subparts and the weaker not entailing its
components, as in (27):
27. Example of Levinson Q-clausal: conditionals
a. <since p then q>, <if p then q>
b. Since hes here, he can play entails He is here; and He can play
c. If hes here, he can play does not entail He is here, nor He can play
In this example, since p then q entails p, and entails q, as shown in (27b). However, if p then q does not entail p, and does not entail q, as shown in (27c). Because of this relationship, by uttering If hes here, he can play the speaker, by excluding the stronger, implicates: Maybe hes here; maybe not; maybe he can play;
maybe not i.e. uncertainty.
The second type is simple alternates, where the alternates are a non-entailing
contrast set. These may be ordered in strength (<stronger, weaker>) as in (28) or an
unordered taxonomy, like the colour terms in (29):
28. Q-alternate: <succeed, try>
29. <red, orange, yellow, green, blue, silver >
Hence from (29) a speaker uttering His car is blue (weakly) implicates His car
isnt silver, etc.
Levinsons I- and M-Principles concern economy. His I-principle, which corresponds to Grices Maxim of Quantity 2, is in simplified form:
Levinsons I-principle (simplified; Levinson 2000, Huang 2007)
Speaker: Do not say more than is required (bearing in mind the Q-principle)
Addressee: What is generally said is stereotypically and specifically exemplified.
This can be paraphrased as the use of a semantically general expression implicates
a semantically specific interpretation. This principle is used to explain how speakers employ or assume defaults and stereotypes, like those in (30):
30. A cheese board = for cutting cheese not made of cheese
A kitchen knife = used in kitchens not for cutting kitchens
It is also used for the examples of strengthening implications mentioned earlier, for
example the occasion when conjunctions like English and can imply temporal and
causal sequence as in (31); when if is interpreted as if and only if as in (32); and
the strengthening of negation as when (33a) implicates (33b):
31. Its always the same at parties: either you get drunk and no one will talk to
you or no one will talk to you and you get drunk (Blakemore 1992: 80)
32. If you pass your examinations, Ill buy you a car.
33. a. Susana doesnt like the Irish weather.
b. Susana dislikes the Irish weather.
473
Levinsons M-Principle relates to the form of utterances, and can be characterized

as:
Levinsons M-principle (simplified; Levinson 2000, Huang 2007)
Speaker: Do not use a marked expression without reason.
Addressee: What is said in a marked way is not unmarked.
This principle explains speakers use of unusual (marked) expressions to avoid defaults and stereotypes. It seeks to explain the pragmatic effects of negative asymmetry, where (34a) below will be interpreted differently from (34b); or where periphrasis carries distinctions of meaning as in (35):
34. a. I dont dislike Henry.
b. I like Henry.
35. a. Bob opened the door.
b. Bob caused the door to open.
Once again these principles are viewed as operating in tension; the interaction between them is a complicated and controversial issue: see Levinson (2000) and
Huang (2007) for discussion.
The work of neo-Gricean writers like Horn and Levinson has shed light on a
wide range of linguistic expressions, seeking to determine the role of semantic encoding and pragmatic inference in their interpretation. An important claim of Levinson (2000) is that generalized conversational implicatures represent a form of
conventional language use, characterized as default inferences, which can be
over-ruled in context. In this view, these linguistically-triggered default inferences contribute to propositional content. This then changes the Gricean relation
between semantics and pragmatics. Since pragmatic inferences contribute to
propositional content, then clearly if there is purely semantic representation it is
pre- or sub-propositional. Or, to put it another way: both semantic interpretation
and pragmatic inference combine in the formation of propositional truth-bearing
content.
4.
Relevance theory
This pragmatic intrusion, to use Levinsons term, is even more marked in relevance
theory (Sperber and Wilson 1995, Carston 2002), which is a more radical development from Gricean inferential pragmatics. Relevance theory is an approach to
communication based on the view that people are predisposed to pay attention to
phenomena in their environment when doing so is likely to bring about improvements in their belief system. In other words, we tend to pay attention to stimuli
which we expect will turn out to be relevant to us. Some stimuli, which include
474
John Saeed
pointing gestures and utterances, are designed to create an expectation that paying
attention to them will yield significant cognitive and other rewards. In other words,
they are designed to create expectations of their own relevance. This type of stimuli is called in this theory ostensive stimuli and the kind of communication which
involves the production and interpretation of ostensive stimuli is called ostensiveinferential communication.
In this view a phenomenon is relevant to an individual (a) to the extent that the
cognitive effects achieved when it is processed in context are large, and (b) to the
extent that the processing effort required for achieving the effects is small (Sperber
and Wilson 1995: 153).
This definition of relevance provides the basis for two principles of human cognition and communication, known as the cognitive and the communicative principles of relevance:
The Cognitive Principle of Relevance
Human cognition tends to be geared to the maximization of relevance. (Sperber
and Wilson 1995: 260)
The Communicative Principle of Relevance
Every act of ostensive communication communicates a presumption of its own
optimal relevance. (Sperber and Wilson 1995: 260)
The presumption can be described as follows:
Presumption of Optimal Relevance
(a) The ostensive stimulus is relevant enough for it to be worth the addressees
while to process it.
(b) The ostensive stimulus is the most relevant one compatible with the communicators abilities and preferences. (Sperber and Wilson 1995: 270)
The influence of Grice is clear here, with its assumption that the cost, so to speak,
of inference can only be motivated by a shared presumption of certain strategies on
both the speakers and hearers behalves.
Consequently, in this account, there are certain assumptions about economy, as
described below by Carston (2002):
Relevance-theoretic Comprehension Strategy
(a) construct interpretations in order of accessibility (i.e. follow a path of least
effort);
(b) stop when your expectation of relevance is satisfied. (Carston 2002: 380)
Of course expectations of relevance may vary across different situations of utterance and in practice adult hearers are expected to have developed strategies to account for speakers fallibility or deceptiveness; and other type of communicative
475
noise. Giving a schematic view of the sub-tasks in the comprehension process,

Wilson and Sperber summarize these as below:
Sub-tasks in the comprehension process (Wilson and Sperber 2004: 615).
a. Constructing an appropriate hypothesis about explicit content (in relevancetheoretic terms, explicatures) via decoding, disambiguation, reference resolution, and other pragmatic enrichment processes.
b. Constructing an appropriate hypothesis about the intended contextual assumptions (in relevance-theoretic terms, implicated premises).
c. Constructing an appropriate hypothesis about the intended contextual implications (in relevance-theoretic terms, implicated conclusions)
From a relevance theory point of view, the processes of recovery of explicatures
and implicatures are both equally inferential and interact freely. In distinction
to Levinsons (2000) position, relevance theorists argue that the pragmatic processes leading to explicatures (i.e. explicit propositional content) are not restricted to special linguistic constructions, nor are they implicatures. Like all
pragmatic processes though, they are motivated by the same presumption of optimal relevance. As we can see, the tasks in constructing a proposition involve context-dependent inference. In this way relevance theory reflects the influence of
context in the most basic of what has sometimes been seen as coded semantic
meaning.
Blakemore (2002) suggests that ostensive stimuli may contribute to the comprehension process in three main ways. First, they may focus the addressees attention in a particular direction (e.g. pointing gestures and acts of displaying in general). For example, a speaker might ask where the bus stop is and the addressee
might point to its location. In this case, the communicator, by displaying an object
to the addressee, suggests that it is relevant as evidence of his/her informative intention.
Secondly, verbal ostensive stimuli typically encode conceptual representations
which contribute to the truth-conditional content of the utterance directly. For
example, instead of pointing the addressee in this scene might say Its across
the road. In this case, the words used present some evidence of what the speaker
intends to communicate by virtue of the concepts that they encode. Thirdly, lexical items may encode procedures (i.e. constraints on or indications of the way
pragmatic inference should proceed). Blakemore (2002) discusses the English
examples in (3638):
36. Tom can open the safe. He knows the combination.
37. Tom can open the safe. So he knows the combination.
38. Tom can open the safe. After all he knows the combination.
476
John Saeed
Example (36) expresses two propositions. While the explicit content is unambiguous, the implicatures that can be retrieved are not. How the implicit content can be
constrained is exemplified in (37) and (38). In these examples, the discourse markers
so and after all guide the computational route that needs to be taken by the hearer. In
(37) the second statement is a conclusion drawn from the first. If someone can open
the safe then he knows the combination. So encodes a constraint on the way the utterance it introduces should be interpreted in context. In So p, p is relevant as a conclusion (which follows from some salient premises). In (38) the second proposition
strengthens the existing assumption that Tom can open the safe. A pronoun like he
in these examples encodes some procedural information which helps the hearer to
form a conceptual representation of the intended referent. So, there are two types of
procedural constraints: constraints on the derivation of implicatures (e.g. so, after
all, however) and constraints on pragmatic inferences involved in deriving the
propositional content of the utterance (e.g. pronouns, tense markers).
We can give a short example from Carston (2006) of how this approach applies
to a hypothetical conversational exchange in (39):
39. A conversational exchange (Carston 2006: 39)
Bob: How is Jane feeling after her first year at university?
Sue: She didnt get enough units and cant continue.
40. A possible interpretation of Sues reply (Carston 2006: 39)
a.
[JANE 1 DIDN T PASS ENOUGH UNIVERSITY COURSE CREDITS TO QUALIFY
FOR ADMISSION TO SECOND YEAR STUDY ]p & AS A RESULT OF P JANE 1
CANNOT CONTINUE WITH UNIVERSITY STUDY.
JANE IS NOT FEELING HAPPY.
b.
The interpretation in (40) contains two different types of propositional results. The
first, (40a) is an expanded or explicit version of the linguistically encoded material.
So although, as Carston notes, the word she doesnt conventionally contain the
meaning Jane or the word units contain an element of meaning university, it seems
that the linguistic items can be a trigger to identify the fuller version, by a process
of working out or inference. The second interpretation is somewhat different. It
provides an answer to Bobs question, but has to be seen as providing it indirectly
since the proposition cant be seen as directly connected to the linguistic content of
Sues answer. Though both parts of the interpretation in (40) seem to depend on
context they seem to do it in different ways. We can easily imagine a scenario
where (40a) is the same, but where Jane is quite happy not continuing her studies.
Perhaps she was forced to attend by her parents and this gives her the excuse to
leave. Thus the indirect answer would be the opposite. Thus the pragmatic processes in (40a) involve expanding the linguistic content while those in (40b) involve identifying intended implications of the expanded linguistic content. Both
however are seen as pragmatic processes.
477
This relevance theory view of the underspecification of encoded meaning is

clearly shown in their work on lexical meaning. The contextual variability of word
meaning is viewed as access to lexical concepts governed by relevance-based contextual inferences:
Quite generally the occurrence of a word in an utterance provides a piece of evidence, a
pointer to a concept involved in the speakers meaning. It may so happen that the intended concept is the very one encoded by the word, which is therefore used in its
strictly literal sense. However, we would argue that this is no more than a possibility, not
a preferred or default interpretation. Any interpretation, whether literal or not, results
from mutual adjustment of the explicit and implicit content of the utterance. (Sperber
and Wilson 1997:15)
The contextual variability of word meaning is seen as a product of a number of inferential processes, including narrowing, as in Wilson and Carstons (2006:409)
example:
41. All politicians drink.
Here the meaning of drink has been narrowed from the meaning drink liquid to
mean drink alcohol and possibly further, to drink alcohol in sufficient quantities
to be worth commenting on. In relevance theory terms, drinking liquid is clearly
necessary for staying alive and in most contexts the fact that a person drinks liquid
has no obvious relevance. Thus in many contexts the expected interpretation of
(42) below:
42. Gordon doesnt drink.
is that Gordon doesnt drink alcohol. Other processes include broadening or approximation (Wilson and Sperber 2004) where the sense of word is extended, as
the word square in:
43. His father has a square face too.
In this example it is unlikely that the speaker intends the usual dictionary meaning
of square. In this view, lexical narrowing and broadening are types of pragmatic
enrichment.
Relevance theorists treat such processes as general in lexical interpretation.
Wilson and Carston (2007) quote Murphys (1997) experimental work on evidence
for distinct word senses by respondents supplied antonyms that contribute to explicatures. For the English word fresh when used with different nouns these include those in Table 1:
478
John Saeed
Table 1.
Antonyms of fresh (Murphy 1997:2379)
fresh
ANTONYMS
shirt
dirty
vegetables
rotten
fish
frozen
sheets
recently slept-in
water
dirty/salt
bread
stale
air
polluted
outlook
tired
assistant
experienced
idea
old
They argue that this is a typical example where a general concept, here FRESH , gets
tailored to individual contexts.
Wilson and Carston (2007) argue that a whole range of what have been considered individual lexical processes, narrowing, broadening, hyperbole (as in 44
below), metaphor (45), category extension (46), etc. can be given a unified account
as inferential processes constrained by the encyclopaedic entries for concepts and
the expectations of relevance. In particular they suggest that this involves a process
of ad hoc concept construction, a once-off contextual interpretation, conventionally shown as starred concepts, i.e. FRESH * for the last example. One important aspect of this approach is the claim that processes like broadening, hyperbole and
metaphor are outcomes of essentially the same pragmatic processes, by different
balancing of encyclopaedic information, contextual assumptions and derived
contextual implications. Thus, their example (47) below could be intended and interpreted as literal, approximation, hyperbole or metaphor, depending on the contextual features (Wilson and Carston 2007: 2489):
44.
45.
46.
47.
5.
Nobody walks in Los Angeles.

Manchester United clawed their way back into the game.
New Orleans is the new Hollywood.
The water is boiling.
Cognitive semantics
Aspects of this relevance theory approach to lexical pragmatics suggest parallels to

that of cognitive semantics (Lakoff 1987, Langacker 1987, 1991, Talmy 2000 inter
479
alia). In cognitive semantics we also see a rejection of the view that literal encoded
meaning has a privileged position, and of the view that hyperbole, metonymy,
metaphor, etc. are special rhetorical devices or tropes. We also see a search for
common general principles of sense determination and extension in context; and
an emphasis on the role of encyclopaedic knowledge in accessing concepts. However, cognitive semanticists use these features of lexical meaning as an argument
against identifying a clearly-defined semantics/pragmatics distinction at all. Langacker (1987:154), for examples, states: The distinction between semantics and
pragmatics (or between linguistic and extralinguistic knowledge) is largely artefactual, and the only viable conception of linguistic semantics is one that avoids
such false dichotomies and is consequently encyclopedic in nature.
Langacker argues that it is impossible to distinguish between lexical and encyclopedic meaning, and, in common with other cognitive linguists, views the
principles involved in organizing concepts and lexically accessing them as part of
more general cognitive processes, rather than specialized linguistic operations.
However the similarities with the relevance theory position are clear. Cognitive
semanticists argue that the conventional meaning associated with a particular linguistic unit acts as a trigger for a process of meaning construction which involves
the determination of an appropriate interpretation for the context of the utterance.
Fauconnier and Turner (2002), for example, discuss interpretation of the English
word safe. They provide some examples of its range of meanings in the context of a
child playing on a beach:
48. a. The child is safe.
b. The beach is safe.
c. The shovel is safe.
As they note, in this context, the interpretation of (48a) is that the child will not
come to any harm. On the other hand, (48b) is not likely in this context to mean that
the beach will not come to harm, but rather that the beach is an environment in
which the risk of the child coming to harm is small. Finally, (48c) is not likely to
mean that the shovel will not come to harm, but that it will not cause harm to the
child. Their point is that there is no single fixed property that safe assigns to
the words child, beach and shovel. In order to understand what the speaker means,
the addressee employs encyclopaedic knowledge of children, beaches and shovels,
and knowledge relating to what it means to be safe. We can see a parallel with Wilson and Cartons (2007) picture of participants constructing a meaning by contextual adjustment based on encylopaedic knowledge.
Although there are individual differences among writers within this approach,
in general cognitive semanticists place no emphasis on a distinction between semantics and pragmatics, seeking instead to identify a range of cognitively motivated representations and computations that straddle the traditional border. An
important constituent of this approach is of course the rejection of truth conditions,
480
John Saeed
or correspondences to states of affairs, as a way of defining meaning. Lakoff

(1988:121) adopts a notion of experiential cognition, contrasting it with its objectivist counterpart:
Where objectivist cognition views human thought as fundamentally disembodied, experientialist cognition sees human thought as essentially involving the
kind of structured experience that comes from having human bodies, especially
from innate human sensory-motor capacities.
Where objectivist cognition sees meaning in terms of a correspondence theory, as the association of symbols with external objects, experientialist cognition sees meaning as essentially involving an imaginative projection, using
mechanisms of schematization, categorization, metaphor and metonymy to
move away from what we experience in a structured way with our bodies to abstract cognitive models.
Where objectivist cognition sees thought processes as the manipulation of abstract symbols by a great many highly-structured algorithms, experientialist
cognition posits a small number of general cognitive processes whose application to abstract highly-structured cognitive models constitutes reason.
Jackendoff (1990) makes a similar point, though couching his view in more
Chomskyan terms:
Hence a truth-conditional semantics in the Tarskian or Davidsonian sense requires a theory of E-language, of language as an abstract artefact extrinsic to speakers within a
theory of I-semantics, a statement in the Tarskian vein like Sentence S in Language L is
true if and only if condition C is met is taken as shorthand for something like Speakers
of Language L treat Sentence S as true if and only if their construal of the world meets
condition C, and it is subject to similar caveats about attentional and processing limitations. This is the basis of the approach of Conceptual Semantics, in which a level of
mental representation called conceptual structure is seen as the form in which speakers
encode their construal of the world. (Jackendoff 1990: 12)
This stance is reflected in the cognitive semantic approach to metonymy and metaphor. Though they are distinct processes for cognitive semanticists they both rely
on structured fields of real world knowledge, identified as conceptual or cognitive
models. In this view metonymy is a referential strategy that is based on links within
the same cognitive model while metaphor is a mapping across cognitive models.
Metaphor is essentially a way of understanding one conceptual model by relating it
to another, usually more concrete and familiar. There have been a number of proposals for such cognitive models including scripts (Schank and Abelson 1977),
frames (Fillmore 1982, 1985), and idealized cognitive models (Lakoff 1987).
The important fact for this literature is that these cognitive models are bodies of
real world knowledge, often culturally specific. Metaphor is then simply a special
type of the more general cognitive process of analogical mapping. Fauconnier
(1997: 19) uses the example of the coining of the term computer virus for a specific
481
type of harmful program, which he suggests is based on a conceptual model of biological viruses which is generalized or schematized away from the biological details:
Biological virus schema (Fauconnier 1997: 19)
x is present, but unwanted; it comes in, or is put in, from the outside; it does not
naturally belong;
x is able to replicate; new tokens of x appear that have the same undesirable
properties as the original x;
x disrupts the standard function of the system;
the system should be protected against x; this might be achieved if the system
were such that x could not come into it, or if other elements were added to the
system that would counteract the effects of x, or eject x, or destroy x
This schema is transferred to the general aspects of the computer situation; it provides a way of characterizing the new domain. The schema is itself based on lower
level conceptual schemas called image schemas, especially characterizing path and
force dynamics: entry, resistance etc. This analogical mapping between a health
schema and a computer domain licenses a whole system of lexical innovations so
that files can be said to be infected and can be placed in quarantine.
The conclusions cognitive linguists draw from such analyses include the following: (a) that the supposedly linguistic process is actually part of a more general
cognitive process; and (b) that such processes are based on real-world experience
and the resulting conceptual models.
These two points may be used as arguments against what are characterized as
assumptions of formal approaches to language: that language is different from
other cognitive processes and that linguistic knowledge is separable from general
knowledge.
We can take as an example of the resulting blurring of traditional semantic and
pragmatic processes Fauconniers (1994, 1997) work on reference. Fauconnier has
in his theory of mental spaces sought to integrate traditional topics such as referential opacity, modality, presupposition projection, counterfactuals, time deixis,
and indirect reference, together with metaphorical uses of language. Fauconnier
suggests that speakers and hearers construct mental models of situations and it is
through these mental models and operations upon them that reference is possible.
These mental spaces in effect can be seen as a cognitive version of possible world
semantics (Lewis 1973), with the difference that more linguistic behavior is
brought under the analysis. For Fauconnier the creation of discourse consists of the
creation and manipulation of series of mental spaces in which individuals and situations are identified, and between which connections are made, as negotiated by
the participants. He identifies linguistic markers, which he calls spacebuilders, as
invitations to the hearers to construct mental spaces and the relevant referential
links:
482
John Saeed
49. spacebuilders (Fauconnier 1994):

adverbials of location and time like in your new novel, in Peters painting,
when she was a child, after we find the crash site, etc.; modals like possibly
and really; connectives like if then; etc.
As well as these overt markers, spacebuilders, the use of mental spaces is also evident in the normal use of names, which can be shifted. As in the following example
where the same name can be used for a real, historical person; a character in a film;
and the actor playing the character:
50. (a) Truman Capote wrote In Cold Blood.
(b) In the film Capote, Truman Capote develops a close relationship with
Perry Smith, one of the killers.
(c) I didnt like the film: Capote was too handsome. (i.e. the actor Philip
Hoffman)
This shifting of referential expressions like names occurs naturally with deictic
pronouns of course:
51. If I were you, Id sell those shares now.
Such connections can be built up by a number of devices, the creation of fictional
characters, by representation in images, etc. Thus a speaker might say, looking at a
television screen:
52. Ronaldo looks a bit out of focus.
Here, in Fauconniers terminology, the name of the Ronaldo is the trigger and the
target is the image. Real Ronaldo (a) and screen Ronaldo (b) are linked by the
pragmatic function IMAGE , so a description of real Ronaldo (his name, da) can be
used to identify his television image (b). In other words, we assume that a speaker
can use a persons name to refer to an image of that person. This is clearly related to
metonymy: here an electronic pattern created within a machine is given the name
of a person whose image it relays. This relates to the traditional view that metonymy derives from a notion of contiguity or connectedness. The approach extends
naturally of course to more traditional cases:
53. a. The blue suit at the end of the bar is about to throw up.
b. The trains are on strike.
Fauconnier uses a very general statement to describe all of this referential behaviour:
The Identification (ID) Principle (Fauconnier 1994:3)
If two objects (in the most general sense), a and b, are linked by a pragmatic
function F (b = F(a)), a description of a, da, may be used to identify its counterpart b.
483
The important points here are that this is a pragmatic process and that any link is
possible: the versions of connectedness can be very idiosyncratic and culturally
specific, though some of course, like the use with images in paintings, photographs, film and video, quickly become conventionalised across cultures sharing
the technology. We might note once again that one major implication of this is that
it overtly recognises that reference depends on conceptual systems, including cultural knowledge, contextual knowledge etc. The cognitive processes freely access
all kinds of information.
Fauconnier has developed this notion of mental spaces further into conceptual
integration theory (Fauconnier and Turner 2002). In this theory knowledge from
multiple mental spaces can be blended to form a new space, forming an emergent
meaning structure that is distinct from the information provided by each input
space. This model seeks to represent speakers abilities to create blended metaphors and scenarios, as in the following newspaper example:
54. On Wednesday, he faces a fast-paced tango with Gordon Brown during the
pre-Budget report debate.7
Here we see the blend integrating cultural knowledge about British politics and
ballroom dancing.
Finally, an important element in the cognitive linguistics rejection of a clear
distinction between pragmatic and semantics is the notion of construal (Langacker 1987), which identifies the dynamic processes by which the speaker conceptualizes a situation for communicative purposes. This involves, for example, a
speaker conceptualizing entities at various levels of specificity, selecting a perspective on a scene, relating it to background assumptions and expectations and
profiling some entities as more prominent than others. In this view the speaker
imposes a structure on a scene, which is conditioned by and in turn conditions the
linguistic form, by selecting grammatical structures, lexical items, etc. This process is also in part governed by the communicative context, including the speakers communicative goals. This approach therefore dissolves any hard and fast
distinction between representation and communicative acts since construal works
at all levels of the grammar and lexicon and involves access to encylopaedic and
contextual knowledge.
6.
Conclusion
From our discussion it is clear that we can conclude with no neat characterization
of the distinction between pragmatic and semantics. There is a wide range of positions in the literature. We can however suggest some trends. More types of information and processes involved in the interpretation of utterances have come to
be viewed as extra-linguistic, and therefore seen by scholars as pragmatic. For the
484
John Saeed
many scholars who relate sentence meaning to truth conditions or propositions,

this can be seen as an encroachment, or intrusion to use Levinsons term, of pragmatics into semantics. From this perspective, the task is to establish what information about meaning is encoded in the linguistic material uttered, that is, by
the grammatical rules of the language. Or to put it another way: what is the nature
of the underspecified semantic representations? Recanati (2005) has characterized
this shift amongst philosophers of language as a movement from literalism to contextualism, with many intervening positions.
Within linguistics some pragmaticists, such as neo-Griceans and relevance theorists, without jettisoning the importance of truth conditions to meaning, have increasingly seen the notion as orthogonal to the distinction between pragmatics and
semantics. Relevance theorists in particular view the difference in terms of a computational distinction between linguistic decoding and inference. For others, like
Chomsky (1995), who reject the relation to states of the world as integral to the
semantic description of natural languages, the semantic representations are not of
course underspecified. However here the question of how linguistic rules and representations map to conceptual structures is part of semantics while a correspondingly large role is given to pragmatics in the derivation of contextually appropriate
interpretations. Finally, cognitive semanticists, as we have seen, see any clear distinction between pragmatics and semantics as an artificial construct deriving from
a mistaken distinction between expressing information and managing communication.
Notes
1. But cf. Collavin this volume.
2. Here we leave aside those writers, for example Montague (1974), who propose a formal
logical approach to meaning which treats natural languages like artificial logical languages and identifies truth relations between linguistic forms and extensions, with no intervening psychological entities.
3. See for example Stalnakers definition: Pragmatics is the study of linguistic acts and the
contexts in which they are performed (1999: 43).
4. See Stanley and Szab (2000) and Stanley (2002) for discussion of domain restriction.
5. For the following see also Huang this volume.
6. It is not certain though that this is a feature of Manner implicatures, where effects are produced by selecting longer or shorter equivalent descriptions.
7. Vincent Cable: The dance-crazy politician aiming to match Gordon Brown step for step,
Marie Woolf, Independent Newspaper, Monday, 8 December 2003.
485
References
Agirre, Enoko and Philip Edmonds (eds.)
2006
Word Sense Disambiguation: Algorithms and Applications. Dordrecht: Springer.
Almog, Joseph, John Perry, and Howard Wettstein (eds.)
1989
Themes from Kaplan. Oxford: Oxford University Press.
Anderson, Stephen R. and Edward L. Keenan
1985
Deixis. In: Timothy Shopen (ed.), Language Typology and Syntactic Description. Vol. III: Grammatical Categories and the Lexicon, 259308. Cambridge:
Atlas, Jay David
1979
How linguistics matters to philosophy: presupposition, truth, and meaning. In:
David Dinneen and Choon-Kyo Oh (eds.) Syntax and Semantics 11: Presupposition, 265281. New York: Academic Press.
Atlas, Jay David
2005
Logic, Meaning and Conversation: Semantical Underdeterminacy, Implicature, and their Interface. Oxford: Oxford University Press.
Austin, John Langshaw.
1975
How to Do Things with Words. 2nd edition. Oxford: Clarendon Press. (First
published 1962).
Bach, Kent
1994
Conversational impliciture. Mind and Language 9: 12462.
Barker, Chris
2002
The dynamics of vagueness. Linguistics and Philosophy 25.1: 136.
Blakemore, Diane
1992
Understanding Utterances: an introduction to pragmatics. Oxford: Blackwell.
Blakemore, Diane
2002
Linguistic Meaning and Relevance: The Semantics and Pragmatics of Discourse Markers. Cambridge: Cambridge University Press.
Borg, Emma
2004
Bring, Daniel
1997
The Meaning of Topic and Focus. London: Routledge.
2005
Carnap, Rudolf
1942
Introduction to Semantics. Cambridge, MA: Harvard University Press.
Carston, Robyn
1998
Informativeness, relevance and scalar implicature. In: Robyn Carston and
Sachico Uchida (eds.), Relevance Theory: Applications and Implications,
Carston, Robyn
1999
The semantics/pragmatics distinction. In: Ken Turner (ed.), The semantics/
pragmatics interface from different points of view, 85125. Oxford: Elsevier.
Carston, Robyn
2002
486
John Saeed
Carston, Robyn
2004
Truth-conditional content and conversational implicature. In: Claudia Bianchi
(ed.), The Semantics/Pragmatics Distinction, 65100. Stanford, CA: CSLI.
Carston, Robyn
2006
Linguistic communication and the semantics/pragmatics distinction. UCL
Working Papers in Linguistics 18: 3769.
Chomsky, Noam
1988
Language and Problems of Knowledge. The Managua Lectures. Cambridge,
MA: MIT Press.
Chomsky, Noam
1995
Language and nature. Mind 104.413: 161.
Davidson, Donald
1984
Inquiries into Truth and Interpretation. Oxford: Clarendon.
Davidson, Donald
1990
The structure and content of truth. Journal of Philosophy 87: 279328.
1992
Rethinking Context. Cambridge: Cambridge University Press.
Fauconnier, Gilles
1988
Quantification, roles and domains. In: Umberto Eco, Marco Santambrogio and
Patrizia Violi (eds.) Meaning and Mental Representations, 6180. Bloomington, IN: Indiana University Press.
Fauconnier, Gilles
1994
Mental Spaces: aspects of meaning construction in natural language. 2nd edition. Cambridge: Cambridge University Press.
Fauconnier, Gilles
1997
Mappings in Thought and Language. Cambridge: Cambridge University Press.
Fauconnier, Gilles and Mark Turner
2002
The Way We Think: Conceptual Blending and the Minds Hidden Complexities.
New York: Basic Books.
1982
Frame semantics. In: The Linguistic Society of Korea (ed.), Linguistics in the
Morning Calm, 111138. Seoul: Hanshin.
1985
Frames and the semantics of understanding. Quaderni di Semantica 6:
222254.
Frege, Gottlob
1980
Translations from the Philosophical Writings of Gottlob Frege. Edited by Peter
Geach and Max Black. Oxford: Blackwell.
Grice, Herbert Paul
1975
Logic and conversation. In: Peter Cole and Jerrold L. Morgan (eds.), Syntax
and Semantics. Vol. 3: Speech Acts, 4358. New York: Academic Press.
Grice, Herbert Paul
1978
Further notes on logic and conversation. In: Peter Cole (eds.) Syntax and
Semantics. Vol. 9: Pragmatics, 113128. New York: Academic Press.
Grice, Herbert Paul
1989
Studies in the Way of Words. Cambridge, MA: MIT Press.
487
Gumperz, John
1992
Contextualization and understanding. In: Duranti and Goodwin (eds.), 229
252.
Hanks, William F.
1992
The indexical ground of deictic reference. In: Duranti and Goodwin (eds.),
4376.
Horn, Laurence R.
1984
Toward a new taxonomy for pragmatic inference: Q-based and R-based implicatures. In: Deborah Schiffrin (ed.), Meaning, Form and Use in Context: Linguistics Applications, 1142. Washington, DC: Georgetown University Press.
Horn, Laurence R.
1989
A natural history of negation. Chicago: University of Chicago Press.
Horn, Laurence R.
1992
The said and the unsaid. In: Chris Barker and David Dowty (eds.), SALT II:
Proceedings of the Second Conference on Semantics and Linguistic Theory.
163192. (Ohio State Working Papers in Linguistics 40.) Columbus, OH: Ohio
State University Linguistics Department.
Horn, Laurence R.
2004
Implicature. In: Laurence R. Horn and Gregory Ward (eds.), The Handbook of
Huang, Yan
2007
Jackendoff, Ray
1990
Semantic Structures. Cambridge, MA: MIT Press.
Kaplan, David.
1989a
Demonstratives. In: Joseph Almog, John Perry, and Howard Wettstein (eds.),
Kaplan, David
1989b
Afterthoughts. In: Joseph Almog, John Perry, and Howard Wettstein (eds.),
Kennedy, Chris
2007
Vagueness and grammar: the semantics of relative and absolute gradable adjectives. Linguistics and Philosophy 30.1: 145.
King, Jeffrey C. and Jason Stanley
2005
Semantics, pragmatics and the role of semantic content. In: Szab (ed.),
111164.
Kyburg, Alice and Michael Morreau
2000
Fitting words: vague language in context. Linguistics and Philosophy 23:
577597.
Lakoff, George
1987
Women, Fire and Dangerous Things. Chicago: University of Chicago Press.
Lakoff, George
1988
Cognitive semantics. In: Umberto Eco, Marco Santambrogio and Patrizia Violi
(eds.), Meaning and Mental Representations, 119154. Bloomington, IN: Indiana University Press.
1987
Foundations of Cognitive Grammar. Vol. 1. Stanford, CA: Stanford University
Press.
488
John Saeed
1991
Foundations of Cognitive Grammar. Vol. 2: Descriptive Applications. Stanford, CA: Stanford University Press.
1999
Grammar and Conceptualization. Berlin: de Gruyter.
Lasersohn, Peter N.
1999
Pragmatic halos. Language 75.3: 522551.
Lepore, Ernest and Barry C. Smith (eds.)
2006
The Oxford Handbook of Philosophy of Language. Oxford: Oxford University
Press.
1983
1987a
Minimization and conversational inference. In: Jef Verschueren and Marcella
Bertuccellii Papi (eds.), The Pragmatic Perspective, 61129. Amsterdam:
Benjamins.
1987b
Pragmatics and the grammar of anaphora. Journal of Linguistics 23: 379434.
1991
Pragmatic reduction of the binding conditions revisited. Journal of Linguistics
27: 10761.
2000
Presumptive Meanings: the Theory of Generalized Conversational Implicature. Cambridge, MA: MIT Press.
Lewis, David K.
1973
Counterfactuals. Oxford: Blackwell.
Lyons, John
1977
Semantics. 2 volumes. Cambridge: Cambridge University Press.
Montague, Richard
1974
Formal Philosophy: selected papers of Richard Montague. Edited and with an
introduction by Richmond H. Thomason. New Haven: Yale University Press.
Murphy, Gregory
1997
Polysemy and the creation of novel word meanings. In: Thomas B. Ward,
Steven M. Smith and Jyotsna Vaid (eds.), Creative Thought: An Investigation
of Conceptual Structures and Processes, 235265. Washington, DC: American
Psychological Society.
Ortony, Andrew
1979
Metaphor and Thought. Cambridge: Cambridge University Press.
Peleg, Orna, Rachel Giora and Ofer Fein
2004
Contextual strength: the whens and hows of context effects. In: Ira A. Noveck
and Dan Sperber, (eds.), Experimental pragmatics, 172186. Basingstoke:
Palgrave.
Perry, John
2001
Reference and Reflexivity. Stanford, CA: CSLI Publications.
Pietroski, Paul M.
2005
Meaning before truth. In: Preyer and Peter (eds.), 255302.
489

2005
Contextualism in Philosophy: Knowledge, Meaning, and Truth. Oxford: Clarendon Press.
2007
Context-Sensitivity and Semantic Minimalism. Oxford: Oxford University
Press.
Recanati, Franois
1989
The pragmatics of what is said. Mind and Language 4: 295329.
Recanati, Franois
1993a
Recanati, Franois
1993b
Domains of discourse. Linguistics and Philosophy 19: 445475.
Recanati, Franois
2000
Oratio Obliqua, Oratio Recta: an Essay on Metarepresentation. Cambridge,
MA: MIT Press.
Recanati, Franois
2004
Literal meaning. Cambridge: Cambridge University Press.
Recanati, Franois
2005
Literalism and contextualism: some varieties. In: Preyer and Peter (eds.),
171196.
Saba, Walid S. and Jean-Pierre Corriveau
2001
Plausible reasoning and the resolution of quantifier scope ambiguities. Studia
Logica 67.2: 27189.
Sadock, Jerrold M.
1984
Whither radical pragmatics? In: Deborah Schiffrin (ed.), Meaning, Form, and
Use in Context: Linguistic Applications. Georgetown University Roundtable,
139149. Washington, DC: Georgetown University Press.
Saeed, John Ibrahim
2009
Semantics. 3rd ed. Oxford and Malden, MA: Wiley-Blackwell.
Schank, Roger C. and Robert P. Abelson
1977
Scripts, Plans, Goals and Understanding. Hillsdale, N.J.: Lawrence Erlbaum.
Searle, John
1979
Metaphor. In: Ortony (ed.), 83111.
Simpson, Greg B.
1984
Lexical ambiguity and its role in models of word recognition. Psychological
Bulletin 96: 316340
Simpson, Greg B.
1994
Context and the processing of ambiguous words. In: Morton Ann Gernsbacher
(ed.), Handbook of psycholinguistics, 359374. San Diego: Academic Press.
1995
Relevance. 2nd ed. Oxford: Blackwell.
1997
The mapping between the mental and the public lexicon. UCL Working Papers
in Linguistics 9: 120. (Republished in Peter Carruthers and Jill Boucher
(eds.), 1998. Language and Thought, 184200. Cambridge: Cambridge University Press).
490
John Saeed
Stalnaker, Robert
1999
Stanley, Jason
2002
Nominal restriction In: Gerhard Preyer and Georg Peter (eds.), Logical Form
and Language, 365388. Oxford: Oxford University Press.
Stanley, Jason and Zoltn Gendler Szab
2000
On quantifier domain restriction. Mind and Language 15.2/3: 219261.
Strawson, Peter Fredrick.
1950
On referring. Mind 59. 235: 320344.
Szab, Zoltn Gendler
2001
Adjectives in context. In: Istvan Kenesi and Robert M. Harnish (eds.), Perspectives on Semantics, Pragmatics and Discourse, 199146. Amsterdam:
Benjamins.
Szab, Zoltn Gendler (ed.)
2005
Semantics versus Pragmatics. Oxford: Clarendon Press.
Szab, Zoltn Gendler
2006
The Distinction between semantics and pragmatics. In: Lepore and Smith
(eds.), 361389.
Talmy, Leonard
2000
Toward a Cognitive Semantics. 2 vols. Cambridge, MA: MIT Press.
Wilson, Deirdre and Robyn Carston
2007
A unitary approach to lexical pragmatics: relevance, inference and ad hoc concepts. In: Noel Burton-Roberts (ed.), Pragmatics, 23059. Basingstoke: Palgrave Macmillan.
Wilson, Deirdre and Dan Sperber
2004
Relevance theory. In: Lawrence Horn and Gregory Ward (eds.), The Handbook
of Pragmatics, 607632. Oxford: Blackwell.
17.

This chapter is an attempt to take stock of advances in the field of prosody, in particular those which have had a direct impact on pragmatics and the study of language use in social contexts. It begins by looking backward, to where the field of
prosody began and at what has been achieved so far with respect to the pragmatic
functions of prosody, and ends by looking forward from the present standpoint to
the challenges that lie ahead. By prosody I refer primarily to those auditory but
non-segmental dimensions of speech which have a variable relation to the words
used (Crystal 1969, Couper-Kuhlen 1986).1
1.
Looking backward
Where and how did the field of linguistic prosody start? What have been the mileposts along the way to its present position? The history of prosody as a linguistic
enterprise dates from approximately the middle of the twentieth century, with
scholars such as J.R. Firth, David Abercrombie, David Crystal, Michael Halliday
in the UK and Kenneth Pike, George Trager, Dwight Bolinger in the US playing a
leading role in establishing its credentials as a discipline within the language
sciences.
Early students of linguistic prosody, particularly those working in the University College of London tradition, focused primarily on intonation and conceptualized its basic unit, the tone unit, as corresponding roughly to the sentence. Descriptive models of the tone unit were developed primarily through introspection and
the examination of read-aloud laboratory speech. The tone unit, it was thought,
could be mapped onto sentences according to syntactic category (cf. e.g. Armstrong and Ward 1926, OConnor and Arnold 1961, Halliday 1967). Five features
characterized this early view of intonation:
(i) metaphor of superposition
In the American tradition, pitch and stress were called suprasegmentals (Trager
and Smith 1957). These were conceptualized as being superimposed on the segmentally based units of language (syllables, morphemes, words) when the latter
were spoken or read aloud. In the British tradition, the stressed syllables were
treated as a frame on which the melodic line was suspended.
492
(ii) the structuralistic perspective

Pitch and stress, working in tandem, were thought to create the basic unit of spoken
language, the tone unit, which itself was divided into smaller structural parts: e.g.,
pre-head, head (or body), nucleus and tail (cf. e.g. Crystal 1969, OConnor and Arnold 1961). Speech was viewed as exhaustively segmentable into such discrete
structural units.
(iii) the context-free approach
Tunes were mapped onto sentences with different syntactic shapes in a contextfree manner. These sentences were consequently isolates, and when pitch and
stress were superimposed on them, the resulting melodies were also isolates. Intonation was acknowledged to be relative, but little thought was given to the question of relative to what? There was only one voice producing little snippets of
melody with no relation to prior tunes or to other voices. In short, there was no
provision for harmony.
(iv) syntactically based functions
The type of melody used was thought to be determined first and foremost by syntactic sentence type (Armstrong and Ward 1926). For instance, declaratives and
wh-interrogatives were said to take final falls in the neutral or default case, yes-no
interrogatives to take final rises. Departures from these neutral contours were said
to convey attitudinal meaning, declaratives with final rises to express doubt, yes-no
interrogatives with final falls to express seriousness or urgency (OConnor and Arnold 1961).
(v) phonological approach to meaning
Intonation was believed to create contrasts in closed systems of meaning, as could
be shown through minimal pairs (Halliday 1967, Schubiger 1958). For instance,
sentences such as She speaks English naturally, George has plans to write or Isnt
he sure of himself were said to be disambiguated by placement of unit boundaries,
position of tonic stress and type of main accent, respectively.
In the early days then, prosody was equivalent to intonation, and intonation (=
pitch and stress) was thought to be superimposed on the occasion of speaking on
particular syntactic sentence types, which themselves were conceptualized as
single utterances, considered in isolation from other utterances. Today we have
come to think quite differently about prosody.2 In part thanks to developments in
the field of sociology, including Berger and Luckmanns social construction of
reality (1967), Sacks, Schegloff and Jeffersons ethnomethodological conversation
analysis (1974), John Gumperz contextualization theory (1982), and the work of
many others whose names go unmentioned here, we have now come to think of
language as a tool for communication and interaction, and consequently of prosody
in terms of social action.
493
The social actions which prosody is implicated in are as varied as are the tasks
which participants must address in communicating with one another. For the conduct of mundane conversation, that primordial site of communication, the following are among those involved:
Turn construction
Action formation
Floor management
Multi-unit turn construction
Turn continuation
Sequence organization
Repair
Stance and alignment
In the following I attempt to summarize what is known at present about the contribution of prosody as a resource for each of these conversational tasks.
1.1.
Prosody as a resource for turn construction
Intonation is no longer conceptualized as something which is optionally overlaid

or superimposed on a piece of decontextualized language. Instead intonation and
prosody more generally is considered to be an integral part of language (Linell
2005: 63). Language products, we have now come to see, do not exist without prosody, i.e. without some kind of pitch, loudness and timing, even if these are realized
silently in reading and writing. Since the natural habitat of language is interaction and it is in turns-at-talk that bursts of language are found (Schegloff 1996),
we now think of intonation and prosody primordially as affording resources for the
construction of turns-at-talk. Given that there are no such things as turns-at-talk
without prosodic formatting, prosody is seen as one of the building blocks of turns.
Prosodic resources are used alongside lexical and syntactic resources as practices
for the construction of turns-at-talk.
1.2.
Prosody as a resource for action formation
Today it is no longer customary to think in terms of simple and direct equations between sentence type (e.g. interrogative vs. declarative) and final pitch movement
(rise vs. fall). Instead, grammar and prosody are thought of as being deployed together to build turns which implement particular social actions in interaction. Prosody then always has some kind of supporting role to play in action formation. A
number of studies have highlighted this function.3 For instance, in dialect survey
interviews specific prosodic features on word repetitions have been shown to cue
the difference between an understanding check of a vocabulary item and a mulling over to determine its familiarity (Kelly and Local 1989). In riddle guessing
494
sequences on call-in radio, another set of prosodic features accompanying the repetition of a guess by the compre cues the difference between an understanding
check of that guess and a negative, disparaging assessment of it (Couper-Kuhlen
1996). In child-adult picture labeling, the prosody with which care-takers repeat
the childs version of a label conveys the difference between an invitation to repair
that label and a confirmation that it is correct (Tarplee 1996). Interestingly, studies
like these testify to the fact that not only final pitch movement but also less wellknown prosodic dimensions such as phonation type (e.g. breathy vs. non-breathy
voice), pitch register (absolute vs. relative) and timing (e.g. delayed vs. welltimed) operate in tandem in specific sequential environments for action formation.
One demonstration of the power of prosodic and phonetic formatting for implementing particular actions will be found in recent work on assessments [or
evaluation] (actions which evaluate some personally experienced object or situation). Assessing turns which are produced in second position, i.e. in response to a
prior assessment, and with prosodic and phonetic downgrading, i.e. with narrower
pitch span, lower pitch accents and reduced volume vis--vis the prior assessment,
will be interpretable as projecting disagreement, rather than agreement, with the
first assessment, even though their assessing term may be the same or lexically upgraded (Ogden 2006).4 As Ogden points out, this kind of phenomenon is revealing
with respect to what prosodic skills a competent speaker of English is expected to
have: viz., they must be able to match their own prosodic and phonetic production
of a turn to that of another speaker and possibly manipulate it for the creation of interactional meaning.
Importantly, the relation between prosody and action formation is not contextfree but dependent on (a) co-occurring lexico-syntactic forms and (b) location in a
particular sequential position. For example, (a) topic proffers5 in English conversation will have final falling pitch if they are done with a wh-interrogative but final
rising pitch if they are yes-no interrogatives (Couper-Kuhlen, forthcoming, b). (b)
Turns which are repeated following other-initiation of repair are likely to be done
with prosodic and phonetic upgrading (i.e. expanded pitch range, longer word durations, greater loudness and articulatory resettings) if they were originally well
placed in the sequence but without this kind of upgrading if they were sequentially
misplaced (Curl 2004). Thus, just as there are no simple equations between sentence type and final pitch, so there are no simple relations between prosodic shape
and action type without a concomitant consideration of lexico-syntax and sequential position.
1.3.
Prosody as a resource for floor management
Acknowledging that language, and thus prosody as well, inhabits turns at talk, we
have also come to recognize its role in signaling transition readiness. Here the
studies which have been carried out so far indicate that prosody, in conjunction
495
with lexico-syntax, has an important role to play in turn-taking. Our judgment of

whether someone has finished speaking or not depends to no little extent on
whether a stretch of speech sounds prosodically complete or not. But the contribution of prosody to turn transition is sensitive to considerations of syntactic and
pragmatic completeness as well (Ford, Fox and Thompson 1996, Ford and Thompson 1996, Selting 1996a, 2000, 2005). That is, a possibly complete prosodic contour will not signal transition readiness unless it is accompanied by possibly complete syntax and unless the activity underway has also reached a point of pragmatic
completion.
The role of prosody in turn-taking is now known to involve more than simply
rising or falling intonation (Szczepek Reed 2004).6 We have learned, for instance,
that different kinds of pausing, e.g. whether a speaker trails off into silence or
produces silence by holding a glottal stop over time, can distinguish turn-yielding
from turn-holding (Local and Kelly 1986). Although we by no means fully understand how turn transition works, we do now know that English dialects can differ
significantly in terms of how syllable-based pitch, loudness and duration are configured at the ends of floor-yielding turns (Local, Kelly and Wells 1986, Wells and
Pepp 1996, Wells and Macfarlane 1998). And we know that speakers of some languages rely much more heavily on prosodic features other than pitch in managing
the floor in interaction. In Finnish, for instance, it appears to be non-modal voice
quality (including creaky phonation) which does much of the work (Ogden 2004,
Ogden and Routarinne 2005). Other languages, by contrast, appear to rely only minimally on prosody for managing the floor (Tanaka 2000, 2004).
Prosody is also known to have a crucial role to play in cueing how overlapping
talk is understood. Competitive incomings7, for instance, i.e. incomings which lay
an immediate claim to speakership well before a current speakers point of possible
completion, will be marked as such with high pitch and increased loudness (French
and Local 1983). The current floor-holder will typically respond either with a prosodic return of competition (increased loudness, decreased pace) or with a yield
to competition (decreased volume). By contrast, entries at non-completion points
which do not lay claim to the floor, e.g. quips and asides, are lower in pitch and volume than prior talk and the speakers norm. Here the current floor-holder does not
treat the incoming speech as turn-competitive but drops out immediately, only to
resume (with a small delay) once it is finished (French and Local 1983).
In sum, current research indicates that, in one way or another and to varying degrees, prosody is implicated in all aspects of turn-taking, including whether and
how speakership transfers from one participant to another in interaction. In this
sense it is central to the concerns of both interactants and analysts.
496
1.4.
Prosody as a resource for multi-unit turn construction
Today it is also customary to look beyond the single intonation unit or phrase to
consider how adjacent intonation phrases relate to one another. This means focussing on the prosodic seams between adjacent units in single-speaker turns-attalk. To do this, the beginnings of new intonation phrases must be related to the
endings of preceding ones, patterns of rhythmic timing across units identified and
attention paid to long-range articulatory and phonatory settings at unit boundaries.8
Registering this level of detail allows us to say that some neighboring intonation
phrases cohere more closely than others. It is the presence of prosodic cohesion between adjacent intonation phrases in a speakers turn which is related to various
practices for multi-unit turn construction, or building one turn out of several turn
units. For instance, at points of possible turn completion there is the practice of
abrupt joins: here what we find at the seam between two adjacent units is a local
speeding-up, a sudden diminuendo and closer than expected temporal proximity
between the two units. Interestingly, these features display that the actions involved are being treated as disjunct with one another (Local and Walker 2004).
Similarly, there is the practice of parenthetic inserts: here it is shifts to low density
of accentuation, faster tempo and lower overall pitch and amplitude compared to
prior talk which signal that the current bit of talk is being presented as parenthetic
(Uhmann 1992, Local 1992, Wichmann 2001, Mazeland 2007). Further, there is
the practice of reported speech, where speakers will be heard to use clear prosodic
breaks in pitch level, rhythm and loudness when switching from their own voice to
that of another (Couper-Kuhlen 1999, Anward 2002, Holt and Clift 2007). Such
prosodic shifts function like quotation marks in speech, although their placement
with respect to the quotation itself is more flexible (Klewitz and Couper-Kuhlen
1999).
Yet not only dis-junctive, also con-junctive relations between units within a
speakers turn are conveyed by clusters of prosodic features linking neighboring
intonation phrases. There is, for instance, the practice of rush-throughs. This involves speeding up the pace of talk towards the end of one turn-constructional unit,
withholding outbreaths and other marks of finality, and moving quickly into the
next turn-constructional unit, so as to present the two units in question as welded
together quasi-seamlessly (Schegloff 1987). Or the practice of projecting, at the
end of one unit, that another unit will follow. For instance, in conjunction with the
particle no when it is used to begin a turn following a topic proffer, mid-to-high
pitch range and increasing intensity project that the speaker will go on to produce
another unit of related talk (Ford, Fox and Hellermann 2004). Prosodic conjunction can also be created through the practice of emphatic scansions, where the
isochronous and isometric9 distribution of syllables across units in a turn-at-talk
produces a scanned multi-unit turn with special rhetorical effects (Auer, CouperKuhlen and Mller1999). Further, there is the practice of conversational listing,
497
where in conjunction with lexico-syntactic parallelism, the use of recurrent plateau contours and other forms of prosodic stylization creates an enumerating
multi-unit turn (Selting 2004, 2007). All of these practices involve particular prosodic features which, together with specific lexico-syntactic forms, link adjacent
units and signal that disjunctive or conjunctive relationships hold between them.
It is thus by comparing and contrasting prosodic formatting at the boundaries
between adjacent units within a single speakers turn that we discover distinct
combinations of pitch configuration, loudness and timing and specific patterns of
phonatory and articulatory setting which together, and along with lexico-syntax,
contribute to the construction of specific kinds of complex, multi-unit turns-at-talk.
1.5.
Prosody as a resource for turn continuation
Today we now also compare and contrast prosodic formatting across non-adjacent
same-speaker turn units and turns. Obviously, there are challenges here due to the
temporal discontinuities involved, but techniques have been developed to deal
with them. By comparing the prosody at the provisional end of a speakers turn
with that at the beginning of a subsequent continuation of that turn by the same
speaker, increments10 can be described as prosodically fitted to their hosts if they
are not significantly higher in pitch or louder in volume than the turn they are continuing and do not involve noticeable changes in articulatory setting or rate of articulation (Walker 2004). More generally, any type of same-speaker continuation
following a possible turn completion can be said to be either prosodically integrated if it is tacked on to the prior intonation contour, or prosodically exposed if it
has a significant pitch jump together with a marked change of tempo, loudness or
rhythm (Auer 1996). Further, when a speakers turn is interrupted or discontinued
and returned to again later, there is now a prosodically and phonetically grounded
way to talk about continuing that turn vs. re-starting it: resumption of a curtailed
turn is done by matching the pitch and loudness levels at drop-out, whereas recycled material is presented as a new contribution if its pitch and loudness levels
are re-set (Local 1992).
In addition to prosodic formatting across same-speaker turns, and just as importantly, prosodic formatting across turns by different speakers can now be compared and contrasted. This is no small feat, given the fact that the speakers involved
may have significantly different pitch spans and loudness ranges as well as different habitual speech rates and rhythms. Yet normalization procedures have been developed for comparing pitch and loudness, accent timing, tempo and speech rate in
cross-speaker talk (cf. e.g. Couper-Kuhlen 1996, Local 2005, Ogden 2006). As a
result, meaningful statements can now be made about the prosody of a speakers
unfinished (or finished) turn and the prosody of a collaborative completion (or continuation)11 of that turn by another speaker (Local 2005, Szczepek Reed 2006).
Collaborative completions, for instance, have been found to be typically no faster
498
or louder than the turn they are completing and to typically match the overall pitch
of the prior talk. They are thus displayed, and accountably so, not only as lexicosyntactic but also as prosodic continuations of what a prior speaker has begun but
not (yet) brought to completion (Local 2005).
In sum, when strict adjacency between stretches of talk is lacking, it is often
prosody which contributes to marking cohesive relations between them. Current
research is revealing just how sensitive participants are to the prosodic details of
continuing a turn-at-talk, be it ones own or that of another. Along with lexico-syntactic considerations, it is these details which help convey whether the talk is being
presented as part of a turn already begun or as something new.
1.6.
Prosody as a resource for sequence organization
We are now also beginning to appreciate the role of prosodic and phonetic resources in signaling how a new turn-at-talk relates to prior turns. For instance, in
radio phone-in calls following an exchange of greetings and possibly how-areyous, callers can signal that a turn in first-topic slot is the reason-for-the-call by
using extra high pitch on its first accented syllable (Couper-Kuhlen 2001).12 If it is
merely an aside, it will lack these features. Once a sequence has reached a point of
possible completion, a cluster of prosodic cues will conspire to convey whether the
next turn is beginning a new sequence or continuing the ongoing one: turns which
are new beginnings are typically louder and higher in pitch from the outset than
prior same-speaker talk and are often accompanied by rhythmic breaks (Goldberg
1978, Couper-Kuhlen 1993, 2004). Turns which continue the prior line of topical
or sequential development lack these features. New sequences which are begun
after a prior one has been closed down are also often marked as phonetically discontinuous by the use of clicks and loud inbreaths (Wright 2007).
Prosodic and phonetic features also contribute to signaling the possible end of a
sequence. Unit-final conjunctionals such as but, and and so have been considered ambivalent in terms of whether the speaker is projecting further talk or not
(Jefferson 1983), but a consideration of their prosodic and phonetic delivery resolves much of the ambivalence (Local and Kelly 1986). Stand-alone variants of
so, for instance, will be heard as potentially sequence-closing if they are relatively
low in pitch, have quiet volume and lack final glottal closure, but as projecting
more on-topic talk if they are relatively high in pitch, loud and have final glottal
closure (Local and Walker 2005).
As Schegloff (1990) has pointed out, adjacent turns-at-talk can either clump
together, i.e. form coherent courses of action, or not. Recent research shows that
such clumping is often underlined by specific prosodic and phonetic devices.
1.7.
499
Prosody as a resource for repair
In addition to signaling structural dimensions of talk, prosody, along with lexicosyntactic resources, is now known to be crucial with respect to what Goffman has
called the ritual requirements of talk. In self-repair, for instance, phonetic and glottal cut-offs will mark a potential trouble source in a speakers turn and initiate its
repair (Jasperson 2002); rhythmic breaks and extreme spikes in pitch and loudness
will accompany the upcoming self-repair itself (Cutler and Ladd 1983). As for
other-initiation of repair, interactants use prosodic resources when they produce a
noticeable silence instead of the conditionally relevant response and thereby implicitly convey that the prior turn may need some reworking. Noticeable silences
are of course notoriously problematic for analysts: how much elapsed time between turns counts as a delay? Yet for participants they are less so. Recent studies
suggest that participants orient to inter-turn silence in relation to the pacing of prior
talk (Couper-Kuhlen 1993, Auer, Couper-Kuhlen and Mller 1999). Prosodic and
phonetic resources are also deployed in other-initiated repair when a turn has been
targeted as problematic via a generalized repair initiator such as what? or huh? and
is repeated with prosodic and phonetic upgrading or downgrading (Curl 2004,
2005).
In short, from what we know so far, it is fair to say that if we did not pay attention to prosodic and phonetic practices in talk we would be unable to account for
much of repair-related behavior, including when repair is initiated, what is targeted
as a trouble source, whether a piece of talk is carrying out a repair or not and how
the responsibility for remedial action is being assigned.
1.8.
Prosody as a resource for displaying and negotiating stance and alignment
Finally, prosody is currently coming to be appreciated as a resource for the display

and negotiation of stance and alignment. The most persuasive research findings
here have situated the use of specific prosodic and phonetic practices within particular action sequences (Goodwin and Goodwin 2000). For example, in news deliveries: the valence with which a piece of news is delivered can be ascribed not
only lexically but also prosodically. Good news will often be signaled with high
pitch level, wide pitch range, animated contour and faster speech rate, while bad
news will have low pitch level, narrow pitch range, stretched vowels, slower pace
and often breathy or creaky voice quality (Freese and Maynard 1998). And if a
piece of news is receipted with a freestanding news-receipt token such as oh,
this turn will be heard as displaying surprise if the oh is lengthened and delivered
with an extra-high and pointed (rising-falling) contour13 but as displaying sympathy or disappointment if it has low pitch, compressed pitch range, soft volume
and weak and/or breathy voice (Reber 2008, Couper-Kuhlen 2009). Further, in
childrens game disputes, when cries of hey!, no! and out! are produced with sig-
500
nificant lengthening, extra high pitch and pronounced falling-rising contours, this
will mark not only disagreement but an affective dimension of strong opposition
(M. Goodwin, C. Goodwin and Yaeger-Droh 2002). As these and similar studies
show, it is by examining prosodic and phonetic resources for the implementation of
specific actions situated in particular sequences and activities that their distinctive
contribution to the display and interpretation of emotive involvement can be appreciated.
But surely one of the most fascinating tasks for which interactants deploy prosody is the negotiation of stance and alignment. In telephone openings, for instance,
callers attend to subtle indications of mood and enthusiasm as displayed by the
pitch level at which their co-participants deliver and respond to greetings and how
are yous. When pitch levels are adjusted to mutually accommodate, this leads to
the joint negotiation of the tenor at which the conversation is to be conducted
(Schegloff 1998). Similar online prosodic negotiation will be found with respect to
rhythm and pacing in conversation openings and closings on the telephone
(Couper-Kuhlen 1993, Auer, Couper-Kuhlen and Mller 1999). More generally
speaking, in virtually any kind of second position, speakers adjust the timing of
their incoming according to the pacing of the prior turn. The way they do this cues
meaningful inferences such as Interaction proceeding apace or I have a problem
with responding as expected (Auer, Couper-Kuhlen and Mller 1999). Recipients
also orient in other respects to their interlocutors prosody and thereby convey
alignment with the prior turn or speaker (Szczepek Reed 2006).14 This is the dimension of prosodic harmony which was missing in early treatments of intonation,
and it is perhaps one of the most exciting aspects of interactive prosodic practice
currently being investigated today.
1.9.
Summary
Looking back now at the work in prosody and interaction which has emerged over
the recent years, it becomes apparent that the nature of the enterprise has subtly
changed. For one, attention is no longer paid only to final, or nuclear pitch movement. Consideration is also given to onset level, pitch range and pitch register; to
the timing of accented and unaccented syllables; to amplitude shifts and levels of
loudness, to pausing and audible breathing; to phonation types and articulatory settings. Moreover, since the approach is an inductive one, further ways in which prosodic, phonetic and paralinguistic dimensions of speech are relevant to interaction
are likely to be discovered in the future.15 Some significant advances have already
been made. We have new ways of describing prosodic and phonetic relationships
with categories such as prosodic matching and prosodic orientation, phonetic
upgrading and downgrading, hypo- and hyper-articulation, etc. And we have
new tools for analyzing prosodic and phonetic patterns, e.g. the speech analysis program P RAAT , which is not only robust enough to handle non-laboratory data and
501
relatively large chunks of speech at a time but is also freely available and constantly
being updated.16 But we are still debating what kind of notation system is needed for
the representation of prosodic and phonetic features in talk-in-interaction.17
2.
Looking forward
What does the future hold in store? What challenges still remain for research into
prosody and interaction? Two recent developments are bringing about some reorientation among students of interactional prosody. They concern (i) the scope of
the field and (ii) its status as a resource for social action in interaction.
2.1.
Scope of prosody
What should we call prosodic? With the increasing awareness that many phonatory, articulatory and voice-quality parameters form clusters along with traditional prosodic parameters for structuring conversation and conveying interactional meaning, the original narrow understanding of prosody as restricted to pitch,
loudness and timing is gradually giving way. Just like the more traditional prosodic
dimensions, phonetic and paralinguistic features also convey what belongs together and what does not (the disjunctive-conjunctive dimension of talk) as well as
what matches and/or upgrades a prior turn-unit and what downgrades it (the affiliative-disaffiliative dimension of talk). For this reason, it is becoming increasingly
artificial to restrict ones attention to a mere subset of these phenomena. Local has
described the situation this way:
() in structuring their talk-in-interaction speakers systematically draw on bundles of
phonetic features which cut across the traditional classification of phonetic parameters
into prosodic on one hand versus articulatory (or segmental) on the other. This suggests that phonetic parameters are best treated as falling into functional bundles or
clusters, irrespective of their prosodic or segmental characteristics, on the basis of
how speakers deploy them to achieve particular interactional goals. (Local 2004:396)
So prosodists today are faced with a terminological dilemma: what to call the prosodic, alias paralinguistic, alias phonetic features which speakers routinely use for
the management of interaction in all its structural/organizational and its ritual/
interactional facets? Should everything be called phonetic? This runs the risk of
making the object of inquiry indistinguishable from that of linguistic phoneticians
whose concern is the phonetic-phonemic interface. Should everything be called
prosodic? This runs the risk of allowing articulatory dimensions to be forgotten
along the way. One is reminded of the debate over sexism in language: are women
made invisible through the use of linguistic forms such as everyone he? Just as
with sexist language, so too discriminatory scientific labels should be avoided. The
solution may thus lie in joint mention, with reference being made, where relevant,
502
to prosodic and phonetic dimensions of language use in interaction. On this

understanding, the present chapter might be more aptly titled: Prosody and phonetics as social action.
2.2.
Status of prosody
How central to social interaction is prosody? With recent advances in video technology there has been a concomitant increase in the attention paid to so-called
multi-modality, i.e. to those embodied and visually based semiotic systems implicated in the conduct of interaction. Prosody is thus no longer an only child; it
has acquired a number of siblings and must now take its place alongside other nonverbal practices for conducting interaction, including gaze, facial expression, gesture, body posture and kinesics. Ironically, among all the various types of embodiment, prosody for decades treated as lying at the edge of language (Bolinger
1972) now appears the most strongly linguistic. Despite the fact that many linguists continue to treat it as at best paralinguistic, in the larger picture of interactional conduct prosody is much more closely tied to language than is, say, gaze
or facial expression. This is in part because prosody is intrinsic to speech and
speech-like vocalization, and in part because prosodic resources are primarily
auditory and are thus organized linearly in ways in which visible resources are not.
So prosody, and research on prosody in interaction, must find a new place in the
multimodal world of interaction analysis. And unless language is to be prioritized,
that place may not be front stage.
Compared to the more dominant, visual semiotic systems deployed in interaction, prosodic and phonetic resources may appear somewhat limited, especially if
they are treated as auditory only. Actually there is often a visible component to the
prosodic and phonetic formatting of talk: so-called smile voice, for instance, is
quite naturally embodied in spread lips and is therefore clearly visible. Surges in
loudness and pitch are typically accompanied by noticeable changes in jaw and larynx position. The temporal onset and offset of speech is visible, and many articulations, not to mention articulatory settings (e.g. labialization), have clearly visible
dimensions. In short, it would behoove prosodists and phoneticians to pay more attention to the visible cues associated with prosodic and phonetic practices. In this
respect prosody and phonetics also have a visual contribution to make. But surely
their most important contribution to multimodal interaction is time or temporality.
In C. Goodwins words: time is an intrinsic component of the units used to
build events and action within situated human interaction. The location of relevant
units is a key task facing both analysts and participants (C. Goodwin 2002:
47). Even in the larger, multimodal picture, temporality remains the backbone of
interaction: it is in real time that interactive events emerge and position themselves
sequentially and simultaneously. And prosodists and phoneticians can help track
this temporal unfolding.
2.3.
503
The way forward
Despite the dominance of multimodality in the world of interaction, prosody and

phonetics in social action remain a central concern both for participants and for
analysts. Without prosodic and phonetic resources, interactants would lack a crucial means for implementing actions indirectly; they would have no way of determining when a particular action is due or overdue; they would be ill-equipped to
mark disjunctive relationships between adjacent pieces of talk or conjunctive relationships between non-adjacent ones; the beginnings of their courses of action
would be lacking in projectability and therefore indistinct from endings; their repairs would risk being taken for continuations; their stances towards actions and towards co-participants would remain opaque, etc. It has often been claimed that
meaning in interaction is dependent on composition (what sorts of parts) and on
position (located where) (e.g. Schegloff 2007: 81). Yet just as crucial is the question of manner (produced how): how high or low, how soft or loud, how fast or
slow is a piece of talk? Is it well-timed or delayed, glottalized or not, said with
spread lips (smile voice) or not? It is answers to such questions that help illuminate the omnipresent issue in interaction: Why that now? (Schegloff and Sacks
1973: 299). The way forward in the field of pragmatics and social interaction will
thus require continued attention to minute prosodic and phonetic detail. Concomitantly, it will require a flexible and, where needed, prosodically and phonetically
explicit transcription system. But it will also demand of its practitioners the willingness and ability to integrate prosodic and phonetic systems with other semiotic
modalities in order to fully appreciate the rich multimodal ecology of sign systems (Goodwin 2002) which underlies human social interaction.
Notes
1. As will become apparent in the course of this chapter, however, the actual delimitation of
the field of prosody is currently a subject of debate.
2. By we I mean those engaged in the empirical and in vivo investigation of prosody as a
pragmatic resource.
3. For reasons of space the following list is necessarily selective (the examples mentioned
all involve word repetition, for instance). Here and elsewhere in this chapter, barring indication to the contrary, mentions are restricted to work on English.
4. In Ogdens most telling case, a second speaker responds to the assessing turn isnt that
good at long last with thats thats (.) very good news but in a way which is prosodically
and phonetically downgraded. In subsequent talk the latter turn is treated as intimating
mild disagreement with the co-participants first assessment (2006: 1771).
5. These are interrogative turn-constructional units which propose a new topic of talk, either
after the opening section of the call or later in the body of the conversation after a prior
topic has been brought to possible closure (Couper-Kuhlen 2004, Schegloff 2007).
6. If only for this reason the negative results of de Ruiter, Mitterer and Enfields laboratory
experiments on final intonation in Dutch turn-taking (2006) are not surprising.
504
7. Competitive incoming refers to a would-be speaker trying to take over the floor immediately, although the current speaker is nowhere near completing their turn.
8. It is in part because articulatory and laryngeal settings are often co-implicated in this
kind of cohesive linking that some scholars have argued for speaking of phonetic instead of prosodic resources (Local and Walker 2005). See also section 2.1.
9. Isochronous refers to a type of rhythmization in which rhythmic intervals have
approximately the same temporal duration; isometric refers to another type of rhythmization in which rhythmic intervals have the same number of syllables (Auer, CouperKuhlen and Mller 1999).
10. An increment is a stretch of same-speaker talk which (a) is produced after a point of
possible completion in that speakers turn, and (b) is syntactically symbiotic with the
prior talk (cf. e.g. Schegloff 1996, Couper-Kuhlen and Ono 2007).
11. Collaborative completion is a term used to refer to a participant finishing a turn-constructional unit which their interlocutor has left (provisionally) incomplete. Collaborative continuation refers to a participant adding on to a turn-constructional unit which
their interlocutor has just (provisionally) completed.
12. This typically happens after the production of uh or uhm (Schegloff 2006).
13. This is especially so when the news being delivered has been designed to elicit surprise
(Wilkinson and Kitzinger 2006). Cf. also Local (1996), and Selting (1996b) for German.
14. Szecepek Reed (2006), for example, documents prosodic orientation involving intonation contour, pitch step-up, register, loudness, speech rate and voice quality.
15. One good example of a recent discovery comes from work on crying in interaction
(Hepburn 2004, Hepburn and Potter 2007), a follow-up to Jeffersons (1985) study of
laughter.
16. Developed by Paul Boersma and David Weenink, Institute of Phonetic Sciences, University of Amsterdam; available at www.praat.org.
17. Among the options are the autosegmental, laboratory-based ToBI system (Beckman and
Ayers 1994), the discourse-based Discourse Transcription system (Du Bois et al 1992)
and the CA-based Gesprchsanalytisches Transkriptionssystem: GAT (Selting et al
1998, 2009).
References
Anward, Jan
2002
Other voices, other sources. In: Per Linell and Karin Aronsson (eds.), Jagen
och rsterna: Goffman, Viveka och samtalet / Selves and Voices: Goffman,
Viveka, and Dialogue, 128147. Linkping: Tema K, Linkpings Universitet.
Armstrong, Lilias E. and Ida C. Ward
1926
A Handbook of English Intonation. Cambridge: W. Heffer and Sons.
Auer, Peter
1996
On the prosody and syntax of turn-continuations. In: Elisabeth Couper-Kuhlen
and Margret Selting (eds.), Prosody in Conversation, 57100. Cambridge:
Auer, Peter, Elizabeth Couper-Kuhlen and Frank Mller
1999
Language in Time: The rhythm and tempo of spoken interaction. New York:
505
Beckman, Mary E. and Gayle. M. Ayers

1994
Guidelines for ToBI labeling. Version 2.0. Columbus, OH, Ohio State University, Department of Linguistics.
Berger, Peter and Thomas Luckmann
1967
The Social Construction of Reality. Harmondsworth: Penguin.
Bolinger, Dwight
1972
Intonation. Selected readings. Penguin: Harmondsworth, Middlesex.
1986
An Introduction to English Prosody. Tbingen: Max Niemeyer and London:
Edward Arnold.
1993
English Speech Rhythm: Form and function in everyday verbal interaction.
1996
The prosody of repetition: On quoting and mimicry. In: Elizabeth CouperKuhlen and Margret Selting (eds.), Prosody in Conversation, 366405. Cambridge: Cambridge University Press.
1999
Coherent voicing: On prosody in conversational reported speech. In: Wolfram
Bublitz and Uta Lenk (eds.), Coherence in Spoken and Written Discourse:
How to create it and how to describe it, 1132. Amsterdam: Benjamins.
2001
Interactional prosody: High onsets in reason-for-the-call turns. Language in
Society 30: 2953.
2004
Prosody and sequence organization: The case of new beginnings. In: Elizabeth
Couper-Kuhlen and Cecilia E. Ford (eds.), Sound Patterns in Interaction,
2009
A sequential approach to affect: The case of disappointment. In: Markku
Haakana, Minna Laakso and Jan Lindstrm (eds.), Talk in Interaction comparative dimension. Helsinki, Finnish Literature Society.
Forthc. Some truths and untruths about final intonation in conversational questions. In:
Jan P. de Ruiter (ed.), Questions, 94123. Cambridge: Cambridge University
Press.
2007
Incrementing in conversation. A comparison of practices in English, German and Japanese. Pragmatics 17.4: 513552.
Crystal, David
1969
Prosodic Systems and Intonation in English. Cambridge, Cambridge University Press.
Curl, Traci S.
2004
Repetition repairs: The relationship of phonetic structure and sequence organization. In: Elizabeth Couper-Kuhlen and Cecilia E. Ford (eds.), Sound
506
Curl, Traci S.
2005
Practices in other-initiated repair resolution: the phonetic differentiation of
repetitions. Discourse Processes 39.1: 144.
Cutler, Anne and D. Robert Ladd (eds.)
1983
Prosody: Models and measurements. Berlin: Springer.
Du Bois, John W., Stephan Schuetze-Coburn et al.
1992
Discourse Transcription. Santa Barbara Papers in Linguistics 4: 1225.
Ford, Cecilia E. and Sandra A.Thompson
1996
Interactional units in conversation. Syntactic; intonational and pragmatic resources for the management of turns. In: Elinor Ochs, Emanuel A. Schegloff,
and Sandra A. Thompson (eds.), Interaction and Grammar, 134184. Cambridge: Cambridge University Press.
Ford, Cecilia E., Barbara A. Fox and Sandra A.Thompson
1996
Practices in the construction of turns: The TCU revisited. Pragmatics 6:
427454.
Ford, Cecilia E, Barbara A. Fox and John Hellermann
2004
Getting past no: Sequence, action and sound production in the projection of
no-initiated turns. In: Elizabeth Couper-Kuhlen and Cecilia E. Ford (eds.),
Sound Patterns in Interaction, 233272. Amsterdam: Benjamins.
Freese, Jeremy and Douglas W. Maynard
1998
Prosodic features of bad news and good news in conversation. Language in
Society 27:195219.
French, Peter and John K. Local
1983
Turn-competitive incomings. Journal of Pragmatics 7: 1738.
Goldberg, JoAnn
1978
Amplitude shift: A mechanism for the affiliation of utterances in conversational interaction. In: Jim Schenkein (ed.), Studies in the Organization of Conversational Interaction, 199218. New York, New York Academic Press.
Goodwin, Charles
2002
Time in action. Current Anthropology 43: 1935.
Goodwin, Marjorie and Goodwin, Charles
2000
Emotion within situated activity. In: Alessandro Duranti (ed.), Linguistic Anthropology: A Reader, 239257. Oxford: Blackwell.
Goodwin, Marjorie H., Charles Goodwin and Malcah Yaeger-Droh
2002
Multi-modality in girls game disputes. Journal of Pragmatics 34: 16211649.
Gumperz, John
1982
1967
Intonation and Grammar in British English. The Hague: Mouton.
Hepburn, Alexa
2004
Crying: Notes on description, transcription and interaction. Research on Language and Social Interaction 37: 251290.
Hepburn, Alexa and Jonathan Potter
2007
Crying receipts: Time, empathy, and institutional practice. Research on Language and Social Interaction 40.1: 89116.
Holt, Elizabeth and Rebecca Clift (eds.)
2007
Reporting Talk: Reported speech in interaction. Cambridge: Cambridge University Press.
507
Jasperson, Robert
2002
Some linguistic aspects of closure cut-off. In: Cecilia E. Ford, Barbara A. Fox
and Sandra A. Thompson (eds.), The Language of Turn and Sequence,
Jefferson, Gail
1983
On a failed hypothesis: Conjunctionals as overlap-vulnerable. Tilberg Papers
in Language and Literature 28: 133.
Jefferson, Gail
1985
An exercise in the transcription and analysis of laughter. In: Teun van Dijk
(ed.), Handbook of Discourse Analysis. Vol. 3, 2534. London, Academic.
Kelly, John and John Local
1989
Doing Phonology. Manchester: Manchester University Press.
Klewitz, Gabriele and Elizabeth Couper-Kuhlen
1999
Quote-unquote. The role of prosody in the contextualization of reported speech
sequences. Pragmatics. 9.4: 459485.
Linell, Per
2005
The Written Language Bias in Linguistics. Its nature, origins and transformations. London: Routledge.
Local, John
1992
Continuing and restarting. In: Peter Auer and Aldo di Luzio (eds.), The Contextualization of Language, 273296. Amsterdam, Benjamins.
Local, John
1996
Conversational phonetics: Some aspects of news receipts in everyday talk. In:
Elizabeth Couper-Kuhlen and Margret Selting (eds.), Prosody in Conversation: Interactional studies, 177230. Cambridge: Cambridge University
Press.
Local, John
2004
Getting back to prior talk: and-uh(m) as a back-connecting device in British
and American English. In: Elizabeth Couper-Kuhlen and Cecilia E. Ford
(eds.), Sound Patterns in Interaction, 377400. Amsterdam: Benjamins.
Local, John
2005
On the interactional and phonetic design of collaborative completions. In:
W. Hardcastle and J. Beck (eds.), A Figure of Speech: a Festschrift for John
Laver, 263282. Mahwah, NJ: Lawrence Erlbaum.
Local, John and John Kelly
1986
Projection and silences: Notes on phonetic and conversational structure.
Human Studies 9: 185204.
Local, John, John Kelly and William H.G. Wells
1986
Towards a phonology of conversation: Turn-taking in Tyneside English. Journal of Linguistics 22: 411437.
2004
Abrupt-joins as a resource for the production of multi-unit, multi-action turns.
2005
Methodological imperatives for investigating the phonetic organization and
phonological structures of spontaneous speech. Phonetica 62: 120130.
508
Mazeland, Harrie
2007
Parenthetical sequences. Journal of Pragmatics 39: 18161869.
OConnor, J.D. and Arnold, Gordon F.
1961
Intonation of Colloquial English. London: Longmans.
Ogden, Richard
2004
Non-modal voice quality and turn-taking in Finnish. In: Elizabeth CouperKuhlen and Cecilia E. Ford (eds.), Sound Patterns in Interaction, 2962.
Ogden, Richard
2006
Phonetics and social action in agreements and disagreements. Journal of Pragmatics 38: 17521775.
Ogden, Richard and Sara Routarinne
2005
The communicative function of final rises in Finnish intonation. Phonetica 62:
160175.
Reber, Elisabeth
2008
Affectivity in Talk-in-interaction: Sound objects in English. PhD diss., University of Potsdam.
de Ruiter, Jan Peter, Holger Mitterer and Nick J. Enfield
2006
Predicting the end of a speakers turn; a cognitive cornerstone of conversation.
Language 82.3: 515535.
1974
1987
Analyzing Single Episodes of Interaction: An Exercise in Conversation Analysis. Social Psychology Quarterly 50.2: 101114.
1990
On the organization of sequences as a source of coherence in talk-in-interaction. In: B. Dorval (ed.), Conversational Organization and its Development,
5177. Norwood, NJ, Ablex.
1996
Ochs, Emanuel A. Schegloff and Sandra A. Thompson (eds.), Interaction and
Grammar, 52133. Cambridge: Cambridge University Press.
1998
Reflections on studying prosody in talk-in-interaction. Language and Speech
41: 235263.
2006
On uh and uhm and some of the things they are used to do. brandial, Potsdam, September 1113, 2006.
2007
Sequence Organization in Interaction. A primer in conversation analysis I.
1973
Schubiger, Maria
1958
English Intonation: Its form and function. Tbingen, Niemeyer.
509
Selting, Margret
1996a
On the interplay of syntax and prosody in the constitution of turn-constructional units and turns in conversation. Pragmatics 6: 371388.
Selting, Margret
1996b
Prosody as an activity-type distinctive cue in conversation: the case of socalled astonished questions in repair. In: Elizabeth Couper-Kuhlen and Margret Selting (eds.), Prosody in Conversation, 231270. Cambridge: Cambridge
University Press.
Selting, Margret
2000
The construction of units in conversational talk. In: Language in Society 29:
477517.
Selting, Margret
2004
The upward staircase intonation contour in the Berlin vernacular: An
example of the analysis of regionalized intonation as an interactional resource.
In: Elizabeth Couper-Kuhlen and Cecilia E. Ford (eds.), Sound Patterns in Interaction, 201232. Amsterdam: Benjamins.
Selting, Margret
2005
Syntax and prosody as methods for the construction and identification of turnconstructional units in conversation. In: Auli Hakulinen and Margret Selting
(eds.), Syntax and Lexis in Conversation, 1744. Amsterdam: Benjamins.
Selting, Margret
2007
Lists as embedded structures and the prosody of list construction as an interactional resource. Journal of Pragmatics 39: 483526.
Selting, Margret, Peter Auer, Birgit Barden, Jrg Bergmann, Elizabeth Couper-Kuhlen, Susanne Gnthner, Christoph Meier, Uta Quasthoff, Peter Schlobinski, and Susanne Uhmann
1998
Gesprchsanalytisches Transkriptionssystem (GAT). Linguistische Berichte
173: 91122.
Selting, Margret, Peter Auer, Dagmar Barth-Weingarten, Jrg Bergmann, Pia Bergmann,
Karin Birkner, Elizabeth Couper-Kuhlen, Arnulf Deppermann, Peter Gilles,
Susanne Gnthner, Martin Hartung, Friederike Kern, Christine Mertzlufft,
Christian Meyer, Miriam Morek, Frank Oberzaucher, Jrg Peters, Uta Quasthoff, Reinhold Schmitt, Wilfried Schtte, Anja Stukenbrock, and Susanne Uhmann
2009
Gesprchsanalytisches Transkriptionssystem 2 (GAT 2). Gesprchsforschung
Online (http://www.gespraechsforschung-ozs.de/).
2004
Turn-final intonation in English. In: Elizabeth Couper-Kuhlen and Cecilia E.
Ford (eds.), Sound Patterns in Interaction, 97118. Amsterdam, Benjamins.
2006
Prosodic Orientation in English Conversation. Basingstoke: Palgrave Macmillan.
Tanaka, Hiroko
2000
Turn-projection in Japanese talk-in-interaction. Research on Language and
Social Interaction 33.1: 138.
Tanaka, Hiroko
2004
Prosody for marking transition-relevance places in Japanese conversation:
The case of turns unmarked by utterance-final objects. In: Elizabeth Couper-
510
Kuhlen and Cecilia E. Ford (eds.), Sound Patterns in Interaction. 6396.

Amsterdam, Benjamins.
Tarplee, Clare
1996
Working on young childrens utterances: Prosodic aspects of repetition during
picture labelling. In: Elizabeth Couper-Kuhlen and Margret Selting (eds.),
Prosody in Conversation, 406435. Cambridge: Cambridge University Press.
Trager, George. L. and Henry L. J. Smith
1957
An outline of English structure. Studies in Linguistics: Occasional Papers 3:
4052.
Uhmann, Susanne
1992
Contextualizing relevance: On some forms and functions of speech rate
changes in everyday conversation. In: Peter Auer and Aldo di Luzio (eds.), The
Contextualization of Language, 297336. Amsterdam: Benjamins.
Walker, Gareth
2004
On some interactional and phonetic properties of increments to turns in talk in
interaction. In: Elizabeth Couper-Kuhlen and Cecilia E. Ford (eds.), Sound
Wells, Bill and Sue Pepp
1996
Ending up in Ulster. Prosody and turn-taking in English dialects. In: Elizabeth
Couper-Kuhlen and Margret Selting (eds.), Prosody in Conversation, 101
Wells, Bill and Sarah Macfarlane
1998
Prosody as an interactional resource: Turn-projection and overlap. Language
and Speech 41: 26594.
Wichmann, Anne
2001
Spoken parentheticals. In: Karin Aijmer (ed.), A Wealth of English, 177193.
Gteborg: Gothenburg University Press.
Wilkinson, Sue and Celia Kitzinger
2006
Surprise as an interactional achievement: Reaction tokens in conversation.
Social Psychology Quarterly 69.2: 150182.
Wright, Melissa
2007
Clicks as markers of new sequences in English conversation. In: Proceedings
of the 16 th International Congress of Phonetic Sciences, Saarbrcken, 1069
1072.
18.

Jacob L. Mey
1.
Introduction
Literary pragmatics is about how language is used in the production and consumption of literary texts. Since much of literature is about telling stories (including the
stories that we value as lyrical expressions of a poets feelings), the notion of narrativity looms large here. To guide the author and reader in their collaborative work
of creating and re-creating the fictional universe of the story, the concept of voice
is of paramount importance. Voices represent characters, as they are created by the
author and perceived, re-created, by the reader; it is the characters that in the end
determine the storys importance rather than the simple plot, or the action, as it is
nowadays called. Important to notice here is that even when the storyteller narrates
own or others real experiences (as in autobiography, or when expressing a reflection on a factual event), the universe that is created by this narration is still entirely
fictional, and so are the characters, real or imagined, that populate the story and the
voices that express their respective points of view. The reader collaborates with
the author in identifying (re-creating) the characters, as they are entering on the
literary scene and voicing their parts. But the voices do not always sound in perfect
unison or dramatic counter-point; when voices become mixed up and clash, the
reader becomes confused and may abandon the common literary enterprise altogether.
Thus, voice clashes (along with the related phenomena of transition between,
and confusion of, voices) offer an interesting indirect perspective on the roles of
authors and readers in the literary enterprise. These roles will be seen as intimately
interrelated, indeed dialectically presupposing each other: the literary enterprise is
a collaborative effort.
2.
Storytelling and stories
Contrary to what many people believe, storytelling is never a matter of getting the
facts, all the facts, and nothing but the facts. What makes a story interesting is the
way we consider those facts as they are presented to us by a narrator; this is what
makes a story different from a simple retelling of factual happenings, as in the familiar kind of grade school composition that starts each sentence with And then .
Stories are part of our lives: they extend into our lives and branch out from
there; but also, stories can only be understood if they are localized, fleshed out in
512
Jacob L. Mey
terms of our daily reality. A story that has no relation to my world is probably not
going to interest me very much. Stories of earlier centuries (like the few surviving
novels of classical antiquity or the interminable chevaleresque poems of the
Middle Ages) appear often boring to a modern reader by their monotonic, endless
repetition of the same themes: but these very qualities may have been what endeared this literature to the contemporary readership and contributed to its popularity. Similarly, the 18th century British author Samuel Richardsons Sir Charles
Grandison (an endless epistolary novel in seven ( ! ) volumes, called Series of
Letters) must have received some appreciation among its contemporaries; but already a century later both Jane Austen (in Northanger Abbey) and Aleksander S.
Pushkin in Evgenij Onegin saw fit to quote the hapless hero as the paragon of reading-induced boredom; compare Pushkins sneering reference to
Grandison, the precious bore
Whose gallant lines provoke our snore
(bezpodobnyj Grandison
kotoryj nam navodit son )
Evgenij Onegin (ch. III, verse ix; my translation)
Stories like Grandison and its predecessors Clarissa and Pamela seem end- and
pointless, precisely because we cannot relate to the reality from which they derive
and to which they point back.
On the other hand, we also find stories in older literature, which today seem as
fresh as when they first were written down. I already mentioned Austen and Pushkin: both furnish outstanding examples of stories that are as fascinating and pageturning today as they were when they were written, nearly two centuries ago. For
who among us has not chuckled over the pointed ironies of Austens characters, or
swooned with Tatiana in her romantic, unrequited love for Onegin? Or (to take an
even earlier case): Boccaccios Decameron is, after three quarters of a millennium,
still good for a thrill and a laugh, on a great number of exciting levels.
While it is difficult to pinpoint what makes a story succeed, it is perhaps easier
to spot at least a few of the causes that make a story fail, or let it fall into oblivion
after the story producers and the consumer generation that supported them have
passed away. In his Evgenij Onegin, Pushkin satirizes the ennui of the Russian genteel country nobility by pointing to the fabulous array of has-been authors who
constituted the daily spiritual bread of the country manor, and whose fame faded as
their pages yellowed (the incomparable Grandison, whose bore is matched by our
snore, is only one among many).
In this and similar cases, the interested readership is confined to students of literature or history; the regular consumers will find it difficult to relate to the characters and their voices, as these represent an entirely different social and societal
environment. The question thus is what makes for a story, and in particular, for a
513
good story. Literary scholars and the common readership have pondered about this
question for centuries, both theoretically, in studies of narrative, and practically, in
university courses of the Creative Writing 101 type and in popular do-it-yourself manuals.
In the following, I will first consider what we mean by narrativity: briefly, the
question is about the essential features of story-telling. Next, the voices of the
narrative will be brought into play, and I will show how they interact, and sometimes conflict. After that, I will identify some, but not all of the linguistic and pragmatic devices that authors use when they tell their stories and that readers rely on
for a successful interpretation and consumption of the narrative. And finally, I want
to go back to the key figures in the process of literary production and consumption:
the authors and the readers, as seen at work in their collaborative efforts, and as
perceived through their respective voices.
3.
Narrativity and voices
Narration is about storytelling. But a story has to be told by someone; let us call
him or her the narrator. Already, it is clear that the narrator is different from the
author: in every story, the author creates a world of fiction, the narrative, in which
the narrator plays an eminent role, even though not always discernible on the surface. Narrativity thus deals with the techniques and devices that a narrator has at
his or her command when telling a story; among these, the notion of character
and the accompanying voice are among the most important.
To see how character and voice connect and interrelate in sometimes rather oblique and inscrutable ways, consider the case of the homodiegetic novel, or
stories told in the first person. Clearly, we cannot assume that the speaking I is
identical to the author, who likewise cannot be held responsible for the narrators
claim as being his own. When the 19th century Russian Mikhail Yu. Lermontov
starts his classic story A hero of our times (1841) with the famous line I traveled
by stagecoach from Tbilisi , every reader will understand that this does not
mean that Lermontov himself did the traveling: the words are spoken by the narrator (referring to his conversation while sitting on top of the coach with a returning NCO, the ranking shtabskapitan serving in Chechenya).
So, the author is not the narrator, not even if the story is told in the first person
singular, as in the Lermontov case. The author creates the narrator as the embodiment of the storys authoritative competence; it is really not important whether this
authority assumes the I-role or pretends to be an outside voice: Trust the tale,
not the teller, as the old adage has it. What characterizes the story is its narrativity, its quality of being narrated and received as a story. The latter part is important, but it is often overlooked; Robert Scholes has pointed out that the very idea of
narrativity is dependent on the reader actively entering the story world and partici-
514
Jacob L. Mey
pating in the narration: a process by which a perceiver [author or reader, JM] actively constructs a story from the fictional data provided by any narrative medium
(1982: 60). Thus, on the part of the reader, there corresponds to the act of narration
an act of active collaboration, an active narrativity.
However we define narrativity, there is a need to distinguish between author
and narrator. A further necessary distinction is that between narrator and actor, and
this is where the concept of narrative level comes in. Following Bal (1985), I distinguish between a primary level, where the action unfolds, and a secondary (and
possibly further) level(s), where the actors come in and their voices are heard. On
the primary level, we have the narrators text; on the secondary (embedded)
level, we are confronted with the actors text. The latter text does not necessarily
contain only what the actors say or think in the context of the primary level; they
may start a story of their own, and this embedded narrative may then again have
several levels of its own, where the secondary story is related to the primary one in
various ways.
A good example is provided by the classic novella by the German author Theodor Storm, Der Schimmelreiter (The man on the white steed, 1888; Engl. transl.
The Dyke Master, 1996), where the narrator on the primary level is the authoritative retired school teacher, who frames the story by telling it to his friends in the
local pub (in German, this kind of narrative is called a Rahmenerzhlung, lit. a
story within a (narrative) frame). But the narrative is not about the teacher and his
audience; rather, it concerns the legendary local official in charge of dams and levees (in German: der Deichgraf, the dyke master), whose expertise and innovative efforts form the secondary narrative, which in fact is the story that we, the
readers, co-construct and remember. The teachers primary voice, as perceived in
his narration and buttressed by his memory, lends credence and local coloring to
the embedded story and its secondary voices.
Similarly, in one of the most famous collections of stories ever told, the 1001
Nights, the main character, Sheherazade, invents and orchestrates the stories,
which she then attributes to secondary narrators on the embedded level. These secondary narrators remain implicit, that is, they are heard, but not seen; in contrast,
another famous collection of stories, Boccaccios Decameron, has the actors on the
primary level narrate their own stories as narrators of secondary level-stories.
In what I have said so far, I have implicitly assumed that actors speak in narratives; in other words, I have attributed them a voice. But how does this happen?
More precisely: How do we know whose voices we are hearing in a narrative? How
do we decide if what we are hearing (and this includes of course reading) is
spoken by the narrator, by an actor (primary or secondary), or by some other
(third) party in a story? To better understand, and answer this question, let me introduce another notion: that of focus, or focalizing, as it is often called (Bal 1985).
Focalizing is what makes an uninteresting series of events into a potentially
fascinating story. To illustrate my point, I will recount a personal happening in
515
which some unrelated events somehow suddenly coalesced into something resembling a story.1
At one time during the winter semester of 1994, I was sitting in the sauna at the
Crown Aquatic Center of Northwestern University, Evanston, Ill. I was enjoying
myself, thinking about nothing in particular, paying not much attention to what
was going on either inside or outside of the sauna room. People came in and went
out again; some took a shower, some stayed longer, some shorter, some were reading their newspapers, others were talking, some were throwing water on the stove
to produce steam and get the temperature up.
All this had been going on for quite a while without my paying much attention.
But after a while, I became aware of a balding, middle-aged man sitting on a chair
outside the sauna and looking in through the big glass door. I remembered him
vaguely: he had been inside and gone out, then had come back in again; now, having taken another shower, he was sitting outside, watching. I am directly in his line
of vision, and suddenly my perspective is no longer that of one watching a scene
with some persons and objects in the background, but of me watching some person
watching me. In other words, I start focalizing.
My immediate, conscious reaction is, of course, that I am making things up.
The man may be there for a number of reasons, and why should he be watching
me? Still, something interesting happens the moment I realize I am no longer observing an inconsequential series of events, or looking at some irrelevant piece of
humanity, but at somebody who is somebody, even though I dont know him. When
I began to focalize, what happened is that I set in motion a narrative process: in
other words, I started making up a story in my head.
Of course, most of this story is just daydreaming, and it happens somewhere
below the explicit level of consciousness. For example, depending on my previous
experiences (or, if you want, my private hang-ups), I attribute a character to the
man, trying to figure out if this is somebody I know, or ought to remember. In a
more interesting scenario, I might consider him to be an elderly homosexual, waiting for me to get out of the sauna so he can make a pass at me.
Now, even though my story can be said to be pure, self-centered fantasy (and
even though I may know it to be exactly that), in terms of narrativity it is still the
case that I have given this person a role as a character in my story, an actor in my
little play; and the next thing I know is that the character starts speaking, starts having a voice. In my head, I even hear snatches of dialogue: the character is asking me
questions (Do I come here often; where do I live; what kind of music do I like; do I
have a phone number, and so on), and I listen to myself providing non-committal,
evasive answers (No, not really; somewhere in South Evanston; no particular
preferences; no, I dont have a phone, etc.). Thus, my focalization makes an indifferent piece of my environment come alive in my self-constructed narrative; from
being a silent prop, the person outside the sauna becomes a talking head. Focalizing implies vocalizing, that is, providing a voice to a character in a narrative.
516
Jacob L. Mey
No matter that the above ctional dialogue never took place, and never will, and
discounting any possible psychological explanations of my sauna daydreaming (earlier experiences, repressed feelings, anxieties, or what have you), what this little anecdote is supposed to show is that we cannot separate focus (and in general, the narrative perspective) and voice. Moreover, focalization presupposes a point of view,
inasmuch as it belongs to a character; hence, every focalization happens according to
a characters perspective. Similarly, every characters voice presupposes the existence of a focalizing perspective in which that character is vocalized. In the next section, I will look at how the characters voices interact, and how they sometimes clash.
4.
Characters and voices
In a story, the characters come to life through their voices. The characters are kept
apart, and their voices separated, by means of simple typographic means (such as
quotation marks or line shifts) or in other, more sophisticated ways; the point of
these techniques is to permit the reader or listener to keep track of what is going on,
to place and/or reposition the focus of his or her attention.
Consider now the following extract:
He [Jack Kemp] joined the Buffalo Bills after an injury.
I hit a helmet with my passing hand and dislocated a finger so severely that I
had to literally decide what shape I wanted it in. So I put my hand on a football
and they put a cast over it.
In those days football was pretty Darwinian. Wed do anything to survive.
I glanced at the finger. Yep.
(From an interview with the late former Republican U.S. Congressman Jack
Kemp, who had been a professional football player in the major leagues NFL/
AFL; Washington, D.C., October 1994; Parks 1994: 22)
In the above story, we have first the voice of the narrator, the journalist who is telling us the story of his interview with Jack Kemp and provides both a historical
frame (He joined the [minor league team] Buffalo Bills after an injury) and a
running comment (In those days football was pretty Darwinian).
Then there is Kemps own, quoted voice, telling us how he hit a helmet with
[his] passing hand and dislocated a finger. According to the conventions of journalism, we are expected to believe that this was how Kemp spoke while being interviewed; it also provides the background for the continued story (again, in
Kemp-speak) of what happened to the speakers hand (I had to literally decide
what shape I wanted it in. So I put my hand on a football and they put a cast over it).
But third, we encounter a voice that has not been heard so far, and a character
that unmistakably is neither Kemp nor the journalist-narrator of the interview. The
517
focus of the narration has changed: what we are hearing in the short utterance
Yep is the voice of the journalist himself, as an active participant in the narrated
scene who suddenly makes his presence in the interview explicit (as opposed to
earlier, when he was only the implicit narrating character telling us, in a running
comment, that He [Jack Kemp] joined the Buffalo Bills after an injury).
This Yep does not belong to the interview, as it is not part of any dialogue; it
represents the journalist Parkss comment to himself on what Kemp has just told
him about the consequences of his injury. This mental comment figures as a kind of
reality check: Parks looks at the finger in question (and Kemps football-curved
hand, as he calls it later on), and affirms the correctness of the story by inspecting
the evidence.
But notice that this affirmation is not part of the interview as narrated. It is the
authors internal, unspoken comment, uttered to himself: I looked, and there it
was all right: a football-shaped finger. Even though the tone of the commentators
voice belongs characteristically in spoken language (being recognizable as such by
its typically conversational nature: yep, unlike its formal equivalent yes, belongs in direct, spoken discourse). The comment is not spoken, and the person uttering it does not belong to the interview scene. One could perhaps say that Parks,
the journalist, intrudes upon the scene by offering his two-bits of commentary
aside, as in an Ibsenian stage direction. His voice is not heard directly, but rather
as a kind of ventriloquizing, as the Russian semiotician and linguist Mikhail M.
Bakhtin has called it.2 It is (self-)reported by the character as an unspoken confirmation of what just has been stated by the interviewee in the reported, uttered dialogue.
Summing up: In the extract above, we have first the (implicit) narrator, the
journalist: his (unheard) voice records the interview for the magazine publishing it,
The Smithsonian (the official publication of the Smithsonian Institution in Washington, D.C.). Then, we have the reported voice of the ex-football player-cum-congressman, Jack Kemp: he is quoted by his very words, in direct discourse. The
third voice is that of the journalist, now coming out so to speak: he sheds his protective gear and changes his status from implicit to explicit narrator: I glanced
at the finger; and he follows up by quoting himself in an internal, likewise unheard but clearly identifiable sotto voce: Yep.
From this little exercise in voicing, we see that in narrative, characters tend to
maintain their respective voices (and their corresponding stances, or points of
view) throughout the narration. They do not speak out of order or out of turn; their
vocalizations help us focalize, such that we do not lose the thread of the story, or
are uncertain of who is speaking or what any of the characters is saying.
In the next sections, we will see that voices not only collaborate in unison but
that they sometimes engage in intra-narrative (and sometimes even extra-narrative) conflicts: the case of the clashing voices.
518
Jacob L. Mey
5.
When voices clash
5.1.
Free Indirect Discourse (FID)
In the present section, I want to examine the phenomenon of vocalization under the
general angle of what I above called focalization (often referred to, as I have done
above, by the term point of view). Since the point from which a voice is heard is
directly dependent on the place of the voicer in the story, it will be useful to
examine this place (and its surrounding discourse) in some detail. I will illustrate
my findings with appropriate examples culled from literary and other discourse; in
particular, many of my examples will be taken from what is often referred to as
free indirect discourse, FID, or represented speech and thought, as it is also
called.
The difference between direct and indirect discourse on the one hand, and FID
on the other, is that in the former, we have explicit devices indicating the point
from which the stream of narration emanates. This point is either localized in the
narrative instance (as in most of traditional novel writing), or in one or the other of
the actors (as it happens in dialogue); or finally (and exceptionally), in the author
him- or herself. Whereas in both direct and indirect discourse, such origins are
marked with conventional means (e.g. quotation marks, or some form of a verbum
dicendi et sentiendi, or an inquit),3 in the case of FID we have no such formal
marks.
FID is especially favored among writers who exploit the technique known as
stream of consciousness, a technique often prototypically associated with modern writers (such as Joyce or Hemingway), but by no means a recent invention or
innovation. Here is an example from Jane Austen:
when [Catherine] thought of Henry, so near, yet so unconscious, her grief and agitation were excessive. The day which she had spent at that place had been one of the
happiest of her life. It was there, it was on that day, that the General had made use of
such expressions with regard to Henry and herself, had so spoken and so looked as to
give her the most positive conviction of his actually wishing their marriage. Yes, only
ten days ago had he elated her by his pointed regard had he even confused her by his
too significant reference! And now what had she done or what had she omitted to do,
to merit such a change?
Jane Austen, Northanger Abbey (Chawton edition, pp. 190191).
In this fragment, even though Catherine is not mentioned as saying or thinking,

there is no doubt that the discourse is Catherines. The personal pronoun she in
the second sentence and onwards refers univocally to the novels heroine. But
since there is no formal marking of this indirect discourse, we call it free, as opposed to regular indirect discourse, which is preceded by a verbum dicendi or sentiendi (such as Catherine said to herself, or Catherine thought).
5.2.
519
Voices clashing in discourse
The interesting thing about FID is that it seems so natural, so without effort, both
on the part of the reader and of the author; yet, this naturalness is an illusion. It does
not take much effort to provoke a clash of voices, wreaking havoc in the discourse, such that the true origin of a particular voice, the characters point of view
is obscured, or even made to disappear.
The French semiotician Roland Barthes has suggested using what one could
call a literary litmus test in order to determine the true provenience of a particular
voice in the discourse. The test consists in exchanging the pronominal reference
for a narrated third person with a (narrating) first person, and then observing how
the various voices get repositioned with regard to their origin in the discourse. The
technique is applied by Robert Scholes to one of Ernest Hemingways short stories;
below, I will repeat Scholes experiment and discuss its implications.
According to Scholes (1982: 116117), in Hemingways A very short story
(1965/1925), there are two third person characters of apparently equal consequence (1982: 116). Scholes goes on to show how the substitution of the first person for each of the two third persons in the narrative gives an absurd result in the
case where the substituted third person in reality was not the original voicer of
the FID. That is to say, the point of view represented by this character is not that of
the I character of the story, which actually represents a covert, first-person narration, as Scholes remarks; but the same thing could, of course, be said of much of
Hemingways work, making up its weakness as well as its strength.4
In the original story, the hero, an American soldier who was wounded in the
Great War (he), has to undergo an operation. While the hero is being readied for
the operation table, he wants to make sure he behaves himself well in the eyes of
the Italian nurse, Luz, with whom he has fallen in love; afterwards, during his convalescence, he takes pains to please her by performing some of her chores.
to take the temperatures so Luz would not have to get up from the bed As he
walked back along the halls he thought of Luz in his bed.
(Hemingway 1965: 141142),
Here is Scholes first transposition, with I replacing the occurrences of the he
third person:
I went under the ansthetic holding tight on to myself so I would not blab
about anything during the silly, talky time After I got on crutches I used to
take the temperatures so Luz would not have to get up from the bed As I
walked back along the halls I thought of Luz in my bed.
520
Jacob L. Mey
And here is the same text with I substituted for the Luz third person:
to take the temperatures so I would not have to get up from the bed As he
walked back along the halls he thought of me in his bed.
Clearly, the first transposition makes sense, and in fact has the same overall reader
effect as does the original story. The occurrences of the third person he keep their
true ring, even when transposed to the first person; they continue to be attributed to
the narrators point of view. The omniscient narrator is telling us what happens
with the hero; likewise, the hero is able to tell us what is happening directly, the
discourse emanating from himself rather than from the narrator. In the same vein,
his thoughts are accessible to the omniscient narrator; there is no clash between the
voice directly representing he, his, and him, and the voice of the narrator telling us the same things in narrative mode.
The second transposition, though, is very odd and voice clashes abound. The
overall point of view here is that of the I person, the woman called Luz but she
is not the omniscient narrator, so how could she possibly know that the male hero
thinks of her in his bed (which is in reality hers)? Similarly for the heros reservations about the blabbing: unless he has told her this (and unless we are informed
that he has done so!), she has no legal narrative access to this knowledge.
Clashes like these, where the voices of the characters and that of the narrative
attributer or voice provider apparently contradict one another, have their origins in the different points of view that the voices represent: clashing vocalizations
are due to incompatible focalizations. While in the first rewriting, the original FID
stands uncontradicted as the male protagonists discourse, a voice clash happens
when we introduce other points of view, and still want to keep alive the pretense of
an FID (as it is done in the second rewriting). In the first rewriting, the focalizer
and the voice are one and the same, just like in the original; in the second rewriting,
the voices are at war.
In our example, the clashes between the different voices were made visible by
the use of a syntactic device: a pronoun which is being switched around with respect to person reference. Clashes of this kind are, however, not bound to a particular syntactic device such as a referential pronoun; among others, the techniques of
changing tenses and moods are frequently used as indicating a change in point of
view.
A clash between different points of view occurs whenever the changes are not
made in accordance with the readers normal expectations. In this case, a shifted
point of view is usually rendered in a different tense and mood than what had been
used in the original discourse. This phenomenon is often called back-shifting: in
FID, a present tense in direct discourse is shifted back into the past, whereas an indicative mood becomes a subjunctive.
521
Compare the following (constructed) examples (due to Bal 1985: 141):

Direct Discourse:
[Elizabeth said] I refuse to go on living like this. (quotation marks; present
tense, indicative)
Free Indirect Discourse:
Elizabeth would not go on living like this. (no quotation marks; past tense, subjunctive)
In the latter case, the focalizing point is not Elizabeths (who is quoted in the first
example as speaking the words herself, using her own voice); rather, the voice is
that of the narrator (the narrative instance), crawling inside Elizabeths head, so
to say, and rendering her monologue as one who was privy to the characters
speech and thought. The words, they are the words of Elizabeth; but the voice is
that of the implicit narrator: the narrator acts as a Biblical impostor, taking on the
identity of the character.
As to the term back-shifting itself, this is somewhat misleading. Actually, the
point of view that is responsible for the FID may shift not only backwards in time
and mood, but also in a forward direction. Compare the following stretch of narrative (also due to Jane Austen):
[Catherine] travelled on for about eleven hours without accident or alarm,
and between six and seven oclock in the evening found herself entering Fullerton.
A heroine returning, at the close of her career, to her native village, in all the
triumph of recovered reputation, and all the dignity of a countess, is an event
on which the pen of the contriver may well delight to dwell; it gives credit to
every conclusion, and the author must share in the glory she so liberally bestows. But my affair is widely different: I bring back my heroine to her home in
solitude and disgrace; and no sweet elation of spirits can lead me into minuteness. A heroine in a hack post-chaise is such a blow upon sentiment as no attempt at grandeur or pathos can withstand. Swiftly, therefore, shall her postboy drive through the village, amid the gaze of Sunday groups; and speedy
shall be her descent from it.
Jane Austen, Northanger Abbey, Chawton edition, pp. 190191.
In this extract, the shifting points of view are clearly marked by the shifts in tense:
first, we have a shift from the narrative past tense ([Catherine] travelled on) to
some gnostic authorial comments offered in present tenses (A heroine returning is an event, and so on). Subsequently, the narrative is extended, in future
tense, to embody the authors own voice, when she is describing to her readers
what she is going to do to her heroine: swiftly shall her post-boy drive speedy
shall be her descent .
522
Jacob L. Mey
Admitting that such direct author interventions in the text are the exception in
modern literary prose (John Fowles being one, fairly recent exception; earlier authors such as Anthony Trollope literally reveled in it), we have to acknowledge
Austens skill in avoiding a clash of voices by making her presence on the scene
clear, among other things, by using a different tense. Readers usually have no beef
with this: if omniscience is the narrators privilege, then omnipresence (albeit in
differing tense garb) may well be the authors. But in order to be properly perceived, the textual roles and their voices have to remain clearly separated, not least
with the help of the linguistic devices that are at the authors and readers disposal.
The next section will discuss how meaning enters into the picture when it
comes to properly distinguishing between voices, and avoiding their clashes.
5.3.
Voicing and reality
Voice clashes often have a semantic background, as when an actor pretends to command a knowledge that is not her or his at the time of speaking: we witness the
author sneaking onto the scene, as it were, without any forewarning or contextual
restrictions. In cases like these, the rules of discourse organization and voice attribution are not observed: what we get is rampant omniscience and an improbable
authorial omnipresence. Often, the author omits to cloak her- or himself in the narrator role when describing events and sentiments that the actors cannot be familiar
with; since the readers depend on the narrator for their being properly introduced to
the specific parts of the literary scene where the narratives events take place, respectively the characters thoughts and sentiments are voiced, their entry into what
is called the fictional space becomes blocked or hindered. Alternatively, the narrator loses control of the narration, and misconstrues the state of the world-atlarge, as it is reflected inside the narrative; anachronisms and erroneous place attributions (along with other malapropisms) are then apt to occur, all resulting in potential voice clashes.
When this happens, the person to blame is ultimately the author who does not
respect the limitations that the fictional space imposes on the narrator and the addressees. Here is an example from a modern novel; the narrative is set in the year
1943, and the female protagonist is enrolled in flight training at a Home Guard base
somewhere in Texas.
As the summer deepened, the heat thickened. The only time she was ever
comfortable was aloft, where the wind and altitude provided natural air-conditioning.
Marge Piercy, Gone to soldiers. (New York: Ballantine Books, 1987, p. 302)
The voice heard here is that of the main character, a young airwoman volunteer;
however, there is another voice, clashing with hers, that of the author who (without
announcing herself or properly introducing her voice) comments on the heat in
523
terms that belong to a much later period. Air conditioners, as we know them today,
were not part of peoples natural reference points in the forties; the most one could
imagine in the line of space cooling technology were devices based on evaporation,
often in combination with a fan or stuck outside of a car window, blowing humid
air into the overheated space and lowering the temperatures by letting the excess
heat be absorbed by the evaporating water. Thus, air conditioners were not natural
comparison points for people in the early forties, and it is for this reason that the
protagonists voice in this passage becomes entangled in a reality clash (usually referred to as anachronism).5
Or, to take another example from the same book (the speaker is an American
soldier commenting on various places where he has been assigned):
They say Switzerland is the same scene on a cleaner scale. But you have to
rank higher than I do and have more top honchos for your friends to pull that
service.
(Piercy 1987: 182)
Here, the voice clash is blatant: the soldier, who is speaking in 1943, uses a word
that didnt come into use until well after the war was over: honcho.6 The characters voice reveals itself as containing a foreign element; the author (speaking in
her own voice-that-knows) intrudes on the narrators voice by putting inappropriate words into the mouth of the character (letting him speak with a voice-that-cannot-know). Its a bit like watching a thirties movie in which you suddenly spot a
vehicle having flashing turn indicators instead of the traditional, semaphore-type
directionals of the earlier period; or spotting a wrong period animal, such as a New
World monkey hanging by its tail, displayed in a presumed painting by Hubert van
Eyck, when all van Eyck could have possibly known in his days were Old World,
tailless monkeys.7
Other instances of this type of reality clash occur characteristically when authors impute knowledge, feelings, or wants to their characters that either are impossible to verify or that could not be imagined in the actual lives of those authors. To
take an example: in Susan Sontags novel, The Volcano Lover (1992: 182), a scene
depicts the deposed Queen of France, Marie-Antoinette, on the scaffold, waiting to
be executed, and feeling the guillotines yoke chafing around her neck. The narrative cannot be attributed to the queen herself, who is soon to die, and whose voice
is only heard internally and has left no record. The voice we do hear through the
implicit narrator is that of the queens sister, Charlotte, Queen of Naples, who
(across the waters) conjures up the horrid scene with the cart, the mob, the steps up
to the scaffold, even the executioners muddy boots and the creaking of the gruesome machinery as the axe starts to fall. Here, the narrator vicariously invests the
story with the feelings, observations, emotions, and horrors of the French queens
sister, as well as providing the reader with a self-styled internal monologue in
Marie-Antoinettes voice.
524
Jacob L. Mey
In this case, too, we are confronted with a shift in point of view that only can be
attributed to the twentieth-century authors invisible appearance, in the guise of
her implicit narrator, on the late eighteenth-century scene. Where earlier authors,
such as Henry Fielding, would think nothing of entering the narrative scene in person, to openly deliver their narrative homilies to the reader (a tradition that still was
alive in the days of Anthony Trollope, and even has been occasionally used by
modern authors), in the scene depicted above the author remains unseen, while
speaking through a variety of inconsistent mouthpieces. Apparently, however,
neither readers nor critics seem object to this kind of vicarious FID being
smuggled into the narrative.8
Contrary to the free indirect discourse that we find in texts like the Jane Austen
passages quoted above a discourse which is duly attributed to one of the characters (usually the hero/ine of the story), in the case of the quotes above, we have a
reality clash (either an involuntary one, as in the Sontag case, or one exploited for
a particular purpose, as it is done by Robertson Davies). In either case, the semantics of the story are not borne out by the reality of the speaking voices. In fact, we
are getting close here to the pragmatic aspect of voice clashing; I will deal with this
in the next section.
6.
The pragmatics of voicing
6.1.
Positioning the voices
Pragmatic voice clashes are bound up with the notion of narrative level; by this, I
understand the relative spatial positioning of the voices within the text. A voice
may be speaking from within the text, as in the case of the narrators or an actors
voice; or it can be heard as speaking from the outside, as it were, on an extra-textual metalevel. In the latter case, we may be listening to what the author or another
text-external authority has to tell us: their voices may emanate from an authoritative source, as it is often done in newspaper reporting, or they may be attributed
to a scholarly authority (propped up by the proper bibliographic references), and
so on.
Voice positioning is important for a proper understanding of the text; but also,
the proper attribution of a voice, no less than its proper use, is part and parcel of the
pragmatics of voicing. Referring to an event is not just a matter of naming place
and date; it requires an understanding of why this event is important for the interlocutors, and how a mention of the event should be voiced. As a clear case of the
need for proper voicing, consider the way ones condolences are expressed when
we meet with a friend or acquaintance who has gone through the painful personal
experience of losing a loved one. Here, correct voicing is so important that most
people will not take any chances, but prefer to use some pre-set formula, like
525
Sorry to hear about your loss, or My prayers and thoughts are with you (the
latter kind is usually restricted to written use).
In the following sections, I will discuss the pragmatics of voicing from a
number of angles, including that of territory of information and empathy. I will
conclude by offering some thoughts on the linguistic realization of voice positioning, and on some of the clashes that may occur when proper positioning is not observed.
6.2.
Territory of information and empathy
The late Japanese linguist Akio Kamio has, in a number of articles and books (e.g.
1994, 1995, 1997), offered a new perspective on the problem of where to properly
position a texts voice. His concept of territory of information implies that we all
have access to some information in ways that others normally do not; our territory
is so to speak sacrosanct in the sense that it cannot be intruded into or operated
upon unless the intruder, the interlocutor, has been properly ratified with regard
to his or her rights and duties.
In certain cases, the territory is defined so personally that we cannot even refer
to phenomena occurring within it without committing a linguistic or social gaffe.
In Japanese, for instance, bodily phenomena, such as being hungry or thirsty, belong to the territory defined by the speaking I. Only the speaker can pronounce
I about him- or herself, as Karl Bhler (1982: 103) remarked a long time ago. But
also, only the speaking I has access to his or her own feelings and sensory experiences; thus, saying You are cold to a person in Japanese is not proper voicing,
since the position of the speaker is outside of the territory where such information
is found. The interlocutor can only say something like You look cold to me, or
You seem to be cold; and similarly for other expressions having to do with the
ego whose territory we are dealing with (like being hungry, having a problem,
and so on). Here is another example, borrowed from Kamio (1995):
In Japanese, it is usually allowable (and hence pragmatically correct) to be
asked, and answer a question, about ones wifes age.9 But while it is OK for a husband to utter
Kanai wa 46 desu My wife is 46
(a husband is supposed to be familiar with the number of years that have
elapsed since his wifes entry into this world), the same husband cannot properly put himself outside this territory of information by adding a quotative
maker (in Japanese -tte or soo) to the utterance. Hence both
Kanai-wa 46 desu-tte I hear my wife is 46
and
Kanai-wa 46 da-soo desu My wife supposedly is 46
526
Jacob L. Mey
are pragmatically ill-formed, as they do not reflect the proper positioning of the
husbands voice, viz., inside his and his wifes shared territory of information.
Clearly, the fact that there are limits to ones territory of information is not particular to Japanese or any other language; but the specific ways in which the territorys boundaries are expressed linguistically are often strictly bound to the respective territorialities of cultures and languages. Something similar is valid in the
cases discussed by Heritage under the general labels of accessibility and the
right to empathy (Heritage 2007). Congratulations, condolences, and other expressions of empathy should not be offered out of place: a faulty positioning of the
speaking voice may result in a fatal clash. A student cannot properly congratulate
his or her professor on the latters brilliant lecturing (see Mey 2008 for a case in
point). Similarly, as Heritage and Sefi remark, advice giving carries problematic
implications about the knowledge or competence of the intended recipient (1992:
389); anybody who has witnessed or experienced cases of spontaneously alter-offered, well-intentioned but ill-received advice knows how toe-curlingly irritating
such information-territorial transgressions can be. To enter or not to enter another
persons territory is not just a matter of fact; it has eminently to do with the pragmatics of voicing, and in particular with the proper positioning of voices within the
proper territory.
The linguistic mechanisms that are at our disposal to effect correct voice positioning and avoid voice clashes that result from undue changes of position or unwarranted infraction of territories are numerous. But more often than not they are
insufficiently understood and clumsily managed by authors and readers; as a result,
a clash of voices may occur. In the following section, I will consider one of the
most widely employed mechanisms that are instrumental for a correct pragmatic
positioning, to wit: the discourse markers often referred to as sentence adverbials, and discuss how they function pragmatically.
6.3.
Sentence adverbials
Among the most frequent mechanisms used for positioning the discoursal voices
are the so-called sentence adverbs (often called sentence adverbials, since they
not strictly have to be what are usually called adverbs, but may be made up of
longer stretches, phrases, and even entire sentences functioning adverbially, that
is to say, specifying a verb or verbal expression). Such adverbial expressions have
as their scope not just the (immediately) preceding or following verb (as in I admire her greatly), but may extend over a whole phrase or sentence. Their name,
sentence adverbials, expresses this fact: they act as ad-verbs, added not just to a
single adjective or verb, but potentially to whole phrases or sentences. In this way,
they are often equivalent to full-blown adverbial clauses; for instance, in an utterance such as To be honest with you, your article is not good enough for our journal, I could replace the entire adverbial clause To be honest with you by a single
527
adverb or adverbial phrase: Honestly/In all honesty, your article is not good
enough for our journal.10
Sentence adverbials have the property of coloring or characterizing the entire
sentence they scope over. Thus, they may be used to express the utterers or
writers viewpoint on what the sentence says; they serve to position the speakers
or writers voice.
Consider the adverb happily in the following sentence (taken from a letter addressed by a prospective author to a journal editor): The paper is formatted in MS
Word; I will happily send you a diskette if necessary. In the context, this tells the
editor that the sender of the message is willing to provide a diskette with her paper,
and moreover that she will be happy to do so.
Alternatively, a sentence adverbial may be used to tell the addressee where the
message in question is to be placed in the ongoing discourse; again, the senders
voice is positioned properly thanks to the adverbial. An adverb such as finally is
often used to tell us that what follows is the last element in a series of arguments,
examples, and so on. In this way, finally is equivalent to a meta-statement by the
author about his or her text: Im telling you now that the next argument or example
is the last one. And consequently, the sentence adverbial may transmit the additional message that there is no more to talk about, from the speakers point of view.
Of course, such meta-statements may remain implicit in the text, or be expressed in other ways, as is often the case in highly ritualized text productions,
such as sermons, political speeches, or newscasts. A particular inflection of the
voice may accompany a summing-up statement (For National Public Radio, this is
Melissa Brown); in such cases we do not need a finalizing adverbial phrase of
the type at the end of the day, or when all is said and done. Alternatively, I may
make use of the multi-purpose adverb anyway to conclude a discussion by indicating my position as the (purportedly) final contributor to the debate, or to voice a
need to return to the original argument the same argument that was expressed in
full sentential array by the judge in the medieval farce of Matre Pathelin, who
coined the famous expression Revenons nos moutons (literally; lets go back
to our sheep), meaning: Anyway, enough of these digressions, lets get back to
the subject.
Sentence adverbials are extremely useful devices, not only for positioning the
voices of the discourse, but also for assuming or denying speakers responsibility
for an utterance. Thus, a sentence adverbial like realistically speaking may indicate a speakers attitude to what is being said; hopefully characterizes the speakers wishful expectations as to the success of the action announced in an utterance;
and so on and so forth. In the language of the press and the courts, the adverbial allegedly is universally used whenever there is talk of a person suspected of being
involved in a criminal act, but without the speaker wanting to make a formal accusation. The sentence adverb allegedly tells us that what we are going to hear does
not necessarily reflect the speakers beliefs or intentions, but something he or she
528
Jacob L. Mey
has heard somebody else say: the speaking voice is located at a safe distance from
the (recording or reflecting) alleging voice, which is only properly heard through
the mouthpiece of the speaker. Thus, on the level of the informative speech act of
assertion, nothing is said that could incriminate, or create a legal liability for, the
speaker person: the catch-all sentence adverbial allegedly acts as a disclaimer for
the whole utterance.
Consider the following constructed example, in which a newspaper reporter
quotes what transpired from the courtroom proceedings in the case of an armed
robbery:
Allegedly, the defendant entered the 7&11 convenience store on South Congress at around 10 pm, demanded to be handed over the money, and when the
store owner declined, proceeded to hit him over the head with an 11 in. pipe
wrench.
Here, the speaker implicitly states that he is not responsible for the truth of the allegations: allegedly is spoken with a borrowed voice; that is to say, somebody
(for whom the speaker is not responsible) is making he statement at his or her own
risk. The responsibility for the truth of the utterance found within the scope of the
sentence adverbial is lifted off the shoulders of the speaker; the expression allegedly marks the whole reported phrase (the defendant entered the 7&11 convenience store and so on) as something that is merely alleged, and therefore not
necessarily true as reported.
In other words, the sentence adverbial in question functions as a disclaimer.
Whatever is within its scope, is said to be outside the verbal jurisdiction of the
speaker, outside his or her territory of information; consequently, the speaker cannot be legally held responsible for the effects of his or her words.
Non-adverbial expressions may serve a similar distancing function. As
examples, consider the words Watch my lips, famously uttered by the first US
President Bush (George H. W.), when questioned about his promise not to raise
taxes, or the fully-fledged disclaimers of the type I didnt mean to say that. This
latter type can often result in voice clashes commonly known as paradoxes: a celebrated case is that of the so-called Moores paradox, by which one denies what
one just has uttered (classic example: The cat is on the mat but I dont believe it,
attributed to the philosopher G.E. Moore; Mey 1993: 156).
Since sentence adverbials are so unobtrusive and impersonal, as contrasted
with explicit, full sentence disclaimers, their voices are positioned in neutral territory, so to speak, while their anonymity prescribes the correct interpretation of the
words uttered even in the absence of a concrete utterer: a speaker or writer. We are
in a borderland of voices here: on the one hand, we have sentence adverbials that
straightforwardly represent the attitude of the speaker like in Strictly speaking,
your request is out of order, where the voice of the speaker is heard as strictly applying certain rules and regulations. On the other, we have cases where the speaker
529
distances him- or herself from what s/he is going to say, as in Strictly off the record, I think youre indulging in a frivolous interpretation of the law. Here, the
speaker wishes to be quoted off the record, that is, the words spoken are not to be
recorded, that is to say, made part of some official account. But the wish to not say
something is not equivalent to not saying that very thing; when I say I dont mean
to insult you, but I think youre not behaving like a gentleman, I cannot countermand the insult, as the words have been spoken by what is rightly assumed to be
my voice, even though that same voice now tries to distance itself from what it just
has said, or is going to say. The speaker dons a speaking mask, so to speak, a persona in the classical sense of the term (like the masks worn on the Greek and
Roman dramatic scenes): speaking on the authority of this voice (and possibly employing what is currently on the verge of becoming a standard disclaimer, viz.,
speaking on condition of anonymity), the voice of the speaker assigns the voice
of the alleged or quoted utterance its correct positioning. We are confronted here
with a change of voice that we may experience as normal and legitimate, but also
on occasion as representing a clash.
The question may be raised whose voices we really are hearing in cases like the
above. Most hearers and readers will not usually be in doubt in the normal course
of events: it does not take too much reflection on the part of the addressee to correctly position the voices he or she is hearing and attribute them to the proper characters in the narrative or discourse. But there is of course some rhetorical trickery
involved in this process. By forcing the addressee to reflect on the positioning of
the narrative voices and on the characters mutual relationships, the speaker may
gain time and improve his or her own position: while the addressees defenses are
down, the speaker is able to get another thrust in edgewise. The clashing voices of
the narrative recall the clanging swords of the classic duel, and it often takes an uncommon skill on the part of the attacked (s)wordsman to position himself correctly
in the face of the attack.
In less martial contexts, the clashing voices may limit themselves to spreading
confusion, as in the following excerpt from a contemporary mystery story, where
Viera, the protagonist, suddenly realizes that the character she had trusted, in reality is the main villain. The development of the narrative, as told by the narrator (the
story is in the homodiegetic, that is, the I mode, as opposed to the heterodiegetic, or he/she mode) leads us to Vieras discovery, and thereby opens up a new
vision on reality for the storys protagonist, causing her to reshuffle her voice
cards.
But what about Vieras own voice: where does it position itself? In particular,
what does the sentence adverbial subconsciously in the following fragment tell
us about this? (Viera is conducting an internal monologue):
Subconsciously, I realized
(Elizabeth Powers, All that glitters. New York: Avon Books, 1981, p. 43).
530
Jacob L. Mey
Vieras internalizing the changed state of affairs, the new reality, may aptly be described by the verb to realize, even when uttered by the narrative voice (recall
what I said earlier about the omniscient narrators position in the story); but how
are we going to deal with the sentence adverbial subconsciously? If the voice we
are hearing is the voice of the character Viera, then the utterance subconsciously
in her monologue is not appropriately attributed to this voice. A person cannot realize something without being conscious of it; hence Viera could not properly utter
the sentence fragment in question without provoking a clash of voices and an ensuing confusion in the readers mind.
This is a pragmatic clash, and we can only solve the dilemma by appealing to a
super-voice that is not heard directly in the dialogue, but is supposed to be implicitly present: this the hidden voice of the omniscient author, inserting itself into
the discourse of the character, where it subsumes and suppresses the clashing
voices of the monologue. We lend credence to this maneuver because we trust the
author; but as readers, we experience a certain uneasiness. This uneasiness may be
formulated in the following simple question: How could Viera realize what she
subconsciously was not even thinking, let alone expressing in words? It is as if the
author were telling the character what its subconscious thoughts would sound like,
when brought to the surface (where they of course would provoke a clash). By
keeping it under wraps, so to speak, the author may avoid the clash, but not eliminate the readers uneasiness, caused by the narrators attempt to pull the literary
wool over our eyes, as she tells us what to believe about the character Vieras unexpressed state of consciousness rather a tall order, even for an omniscient narrative instance.
7.
Conclusion: narration and conversation
From a pragmatic point of view, both the literary text and the text generated in conversation are manifestations of human language use. It may be useful to briefly
compare the two uses in this concluding section, to see if and how they are different, and if any observed differences can shed light on the use of the literary voices
that we have been dealing with so far.,
If one asks how language use differs from narration to conversation, one notices that the ways we use language in the two cases are rather diverse. Narration
may include conversation, and conversely, conversations usually include stretches
of narration; yet, their respective ways of doing things with words are vastly different. Compare the meticulous wording that is the hallmark of literary prose, with
the spontaneous utterances that characterize everyday conversation: in the first
case, we have authors judiciously assigning the various voices of the text to their
respective characters (where clashing voices should not normally, but do occasionally, occur); in the second, we are confronted with what sounds like a cacophony of
531
voices, some speaking in unison, some interrupting, some dropping out, some
dominating, some subordinated, and so on.
The conversation analysts have shown us that (despite apparent irregularity)
the business of conversation is conducted according to strict rules of selection,
turn-taking, and (non-)overlapping. Beginning with the ground-breaking work of
people like Sacks, Schegloff and Jefferson (1974), a new discipline, Conversation
Analysis, has seen the light of day and is enjoying great popularity among a crosssection of the linguistics and sociology fields.
At first blush, the question Whose voice are we hearing? may seem a bit inappropriate in the conversational context. In conversation, we will always be able
to identify the current speaker (the one who has the floor, as it is called). The current speaker is the one who can be singled out by his or her use of the first singular
(or sometimes plural) personal pronoun (the speaking I, as the German psychologist Karl Bhler used to call it; 1982: 103). By the same token, the passages of a
novel in which the characters conduct conversation should be the ones that are the
least problematic with respect to the positioning of those voices.
But on closer inspection, the question of where the voices are coming from
turns out to be more complicated. As we have seen in the preceding, voices are not
always what they pretend to be, or sound like; an implicit voice may represent an
extraneous instance, speaking with the authority of a superaddressee (as Bakhtin
calls this instance), namely: God, absolute truth, the court of dispassionate human
conscience, the people, the court of history, science, and so forth (1994: 126).
Such a voice may not be as easy to detect as, say, the voice of a person using a local
or social dialect; but it may manifest itself in numerous indirect ways, e.g. by the
authority it invokes, the appeals it makes to the listeners, the value-laden judgments it proffers or reproduces, and so on.
When Jane Austen rebels against the common cant which undervalue[s] the
labour of the novelist (1947: 23), we hear the authors voice deploring the poor
appreciation allotted to novelists on the part of those whose voices belong in the
specific discourse, common to a particular type of readers: those who will pick up
an Austen novel, but hide it in shame when caught in the act of reading. Similarly,
when Virginia Woolf lets Clarissa Dalloway muse about the important meetings
her husband Richard has to attend (Armenians, he said; or perhaps it was Albanians; Woolf 1990: 119), we hear, through the speaking voice of the protagonist, the implicit background voices of the societal discourse of the time, where the
Armenian question played such an important role but not important enough for
Mrs. Dalloway to be able to keep Armenians apart from Albanians.
Similarly in everyday conversation, voices may appear on the interlocutory
scene that are not explicitly heard or observed, but stay hidden in the conversational backdrop. For instance, a person may ask an interlocutor a question that is
not appropriate in the context, as it presupposes an authority on the part of the
questioner that he or she may not have. The implicit voices we are hearing when
532
Jacob L. Mey
such questions are denied an answer are those of societys powerful instances (such
as the police or the judiciary), that pervade the entire fabric of our society. While a
policeman or -woman can stop me in the streets of an affluent neighborhood and
ask me for my identification if he or she perceives something untoward, a regular
neighborhood dweller is not allowed to do the same, except maybe in an act of
neighborhood watching (or in the US context) of making a citizens arrest. In
the latter cases, such a person would be speaking with the (implicit or impersonated) voice of the police official, ventriloquating (to use Bakhtins colorful expression) the sounds emanating from the official instance, the police superaddressee. In conversation as in narration, the proper positioning of the speakers and
their voices is every bit as important as the produced text (narrative or conversational) itself.
In the pragmatic analysis of a text (narrative or conversational), correctly placing the voices in the discourse is as crucial for the readers as it is for the interlocutors. Both have to know, at any given moment of the (literary or conversational) interaction, where they stand in relation to the other participants, and where the
protagonists of the literary production: viz., the characters (including the author
and the narrator) of the narrative, respectively the interlocutors in the conversational situation, are located in the discourse.
Pragmatics, when applied to narrative as well as conversational texts, raises a
double question: From which point of view are we being addressed as hearers and
readers; and: Whose language is being spoken, whose voices are being heard? In
the preceding pages, I have tried to make it clear that, in order to answer these (not
unrelated) questions, we have to rely on the linguistic and pragmatic concepts that I
have referred to in the preceding pages, first of all, the central notion of voice and
its (harmonious or clashing) manifestations in literary and conversational discourse.
Notes
1. The story has been told earlier in my book When Voices Clash (Mey 2000: 148150).
2. The author speaks as it were through language, a language that has somehow become
objectified, that he merely ventriloquates [sic] it (Bakhtin 1992:299).
3. These terms are taken from traditional Latin (school) grammar. Verba dicendi are, e.g.,
to say, state, affirm, exclaim, ; verba sentiendi express the thoughts or feelings of the
subject (e.g. to think, mean, believe, ); inquit is the 3rd person singular of the defective Latin verb inqueo to say, speak.
4. Of course, as Scholes also remarks, [t]he third-person narration of the original text is a
disguise, a mask of pseudo-objectivity worn by the text for its own rhetorical purposes.
(1982:117)
5. According to the Encyclopaedia Americana, air conditioners for home use began to be installed in the U.S. in the late 40s-early 50s.
533
6. The words first official mention (as far as I have been able to ascertain) is in the
Oxford American Dictionary (1980). It is said to be derived from the Japanese hancho, leader of a band; it was first introduced into English, according to Wikipedia, in
1955.
7. And in this way, the clash reveals the true date of the forged painting, The Harrowing
of Hell, as the incident is described by the Canadian novelist Robertson Davies in
the second volume of his Cornish trilogy, Whats bred in the bone. (Penguin 1985,
pp. 350351)
8. Compare the British critic Natasha Walters comments on this episode, in her review of
Sontags work: there are scenes to remember in the book: Marie-Antoinette on the
scaffold, feeling the yoke choking her (The Independent, October 2, 1992, p. 27)
9. Or ones own, for that matter. Other cultures may have other, special socio-cultural conventions of propriety; compare: Women dont have ages, as Talleyrand supposedly remarked to Metternich at the Vienna Congress.
10. Sentence adverbials are among the lesser studied pragmatic discourse operators. The
fullest treatment (by now somewhat dated) is by the Canadian-Czech linguist Eva Koktov (1986).
References
Austen, Jane
1947 [1815] Northanger Abbey (Chawton edition). London: Allan Wingate.
Bhler, Karl
1982 [1934] Sprachtheorie: Die Darstellungsfunktion der Sprache. Stuttgart and New
York: Gustav Fischer.
Bakhtin, Mikhail M.
1992
Discourse in the novel. In: Michael Holquist (ed.), 259422. [First Russian
publication 1975].
Bakhtin, Mikhail M.
1994
The problem of the text in linguistics, philology and the human sciences. In:
Caryl Emerson and Michael Holquist (eds.), 103131. [First Russian publication 1976].
Bal, Mieke
1985
Narratology: Introduction to the theory of narrative. Toronto and London:
Emerson, Caryl and Michael Holquist (eds.)
1994
Mikhail M. Bakhtin: Speech Genres and other late essays. Trans. by Vern W.
McGee. Austin, TX: University of Texas Press. [First Russian publication
1986].
Hemingway, Ernest
1965 [1925] A very short story. New York: Scribner.
Heritage, John
2007
Territories of knowledge, territories of experience: (Not so) empathic moments
in interaction. Keynote speech at the XVth Symposium About Language and
Society, Austin (SALSA), Austin, TX, April 14, 2007.
534
Jacob L. Mey
Heritage, John and Sue Sefi

1992
Dilemmas of advice. In: Paul Drew and John Heritage, Talk at work: Interaction in institutional settings, 359417. Cambridge: Cambridge University
Press.
Holquist, Michael (ed.)
1992
The dialogic imagination: Four essays by M.M. Bakhtin. Austin, TX: The University of Texas Press.
Kamio, Akio
1994
The theory of territory of information: The case of Japanese. Journal of Pragmatics 21(1): 67100.
Kamio, Akio
1995
Territory of information in English and Japanese and psychological utterances.
Journal of Pragmatics 24(3): 235264.
Kamio, Akio
1997
Theory of territory of information. Amsterdam: Benjamins.
Koktov, Eva
1986
Sentence adverbials in a functional description. Amsterdam: Benjamins.
Mey, Jacob L.
1993
Pragmatics: An introduction. Oxford: Blackwell.
Mey, Jacob L.
2000
When voices clash: A study in literary pragmatics. Berlin: Mouton de Gruyter.
Mey, Jacob L.
2008
Impeach or exorcise? Or, whats in the (common) ground? In: Istvan Kecskes
and Jacob Mey (eds.), Intention, common ground and the egocentric speakerhearer, 254275. Berlin: Mouton de Gruyter.
Parks, Edward
1994
Around the Mall and Beyond. The Smithsonian 25(8): 2025.
Piercy, Marge
1987
Gone to soldiers. New York: Ballantine Books.
Powers, Elizabeth
1981
All that glitters. New York: Avon Books.
1974
Language 50(4): 696735.
Sontag, Susan
1992
The volcano lover: A romance. New York: Farrar Straus Giroux.
Scholes, Robert E.
1982
Semiotics and interpretation. New Haven, Conn.: Yale University Press.
Storm, Theodor
1996 [1888] The Dyke Master. London, Angel Books. (German original: Der Schimmelreiter. Transl. by Dennis Jackson,).
Woolf, Virginia
1990 [1925] Mrs. Dalloway (Harvest edition). San Diego, New York and London: Harcourt Brace and Company.
Part V
Methods and tools
19.

Monika Bednarek
1.
Introduction: approaching the data of pragmatics
As the editors of this series, Wolfram Bublitz, Andreas Jucker and Klaus Schneider
argue pragmatics is a transdisciplinary subject which has a substantial and
wide-ranging significance for nearly all areas of language description. From the
broad conceptualization of pragmatics that they envisage, pragmatics includes approaches that are traditional and contemporary, linguistic and philosophical, social
and cultural, text- and context-based, as well as diachronic and synchronic. In this
view, pragmatics shares interests with other related and not always easily differentiable linguistic sub-disciplines, such as discourse studies, semiotics, psychoand sociolinguistics, conversation analysis, or ethnography of speaking (see
Wodak 2008: 34 for a list of seven dimension these fields have in common), sharing for instance the focus on language use or linguistic (inter)action. There is thus a
common commitment of these linguistic sub-disciplines to language in use the
pragmatics of language.
It seems reasonable to propose that any discipline interested in language use
must have some kind of approach to its object of investigation the data of linguistic (inter)action, or simply, the data of pragmatics. This includes a methodological standpoint from which to approach this data, some kind of research
methodology. However,
the commitment to particular methods or procedures often fails to deal with a range of
important questions: where or from whom should the data be collected and how? (For
example, shall we use texts that we obtain in interviews or published self-descriptions?
Do we include the news on the notice board and the graffiti in the toilets? Are texts sufficient for our purposes or should we also carry out observations?) Equally undecided is
the matter of how the data collection is to be organized. (For example, do we generate
our texts in group interviews or individual interviews? In what order shall we do the interviews?) The question of how the material is to be stored also remains open. (For
example, will audio recordings be sufficient or do we need video recordings? What rules
of transcription shall we apply?) (Titscher et al 2000: 1314, see e.g. Kreuz and Riordan
in this volume on transcription systems).
Thus, while the focus of this chapter is on the use of different data in pragmatics
and related disciplines (compare first paragraph above) rather than a general discussion about research methodology, epistemology, or evaluation criteria such as
ermk 2001; Silverman 2004;
reliability, validity, replicability (see Taylor 2001; C
McNeill and Chapman 2005; Drnyei 2007), reference will be made to the insights
gained from different approaches and methodologies to data and the different per-
538
Monika Bednarek
spectives they provide on the linguistic data. The starting point of this chapter is
thus the differentiation between different kinds of data, rather than between different kinds of methodologies (e.g. quantitative vs. qualitative).
2.
Issues of data collection and analysis
Issues that have to be addressed by all researchers using data are whether all data
should be systematically collected before the analysis or whether further data can
and should be added after some data has already been analyzed. This question relates to the design of data collections or corpora (see e.g. Wynne 2005, Voormann
and Gut 2008). Another issue is the difference between inductive (data-driven) and
deductive (theory-driven) research. In other words, do we analyze linguistic data
according to some pragmatic theory such as politeness (Leech 1983, Brown and
Levinson 1987) or do we deduce some pragmatic theory from the analysis of the
linguistic data itself? This kind of distinction cuts across the use of different data,
and different methodologies, such as quantitative vs. qualitative research. For instance, speaking of large-scale corpus linguistics, Tognini-Bonelli (2001) makes a
distinction between corpus-driven and corpus-based linguistics. The term corpusbased is employed to refer to a methodology that avails itself of the corpus mainly
to expound, test or exemplify theories and descriptions that were formulated before
large corpora became available to inform language study (Tognini-Bonelli 2001:
65), and the term corpus-driven refers to an approach where the commitment of
the linguist is to the integrity of the data as a whole, and descriptions aim to be
comprehensive with respect to corpus evidence (2001: 84). Similarly, Bednarek
(2006a) speaks of a text-driven methodology, where texts are analyzed trying to
make as few a priori theoretical assumptions as possible before analyzing the text
data, and a text-based methodology, where texts are analyzed using a previously
established theoretical framework. In contrast to corpus-based/corpus-driven research, text-based/text-driven research is based on the manual (and more qualitative) analysis of small-scale text corpora rather than on automated quantitative
large-scale corpus analyses (Bednarek 2006a). However, what Halliday notes with
respect to corpus-based and corpus-driven research is also true for text-based and
text-driven analysis: the two approaches represent a continuum rather than a strict
dichotomy, and much research is situated somewhere in between (Halliday 2004:
24). This may also be the case more generally with respect to inductive vs. deductive approaches to data analysis in pragmatics.
Other issues that need to be addressed by all researchers include legal, ethical
and technical considerations, with many research proposals and projects needing
the approval of ethics committees at universities, and much data collection and
publication requiring sometimes costly copyright consent of respective copyright
owners. Most spoken data also require consent of speakers to use and publish these
539
data. Relevant technical issues concern questions such as: what technology is
needed to collect the data (e.g. audio and video recordings, keying in of data, conversion of internet data into other formats), to store the data (e.g. large databases,
large electronic storage facilities) and to analyze the data (e.g. computer software)?
Such issues can place real constraints on the analysis of linguistic data and should
in general be considered before data collection.
3.
Types of data
3.1.
Non-attested data
I use the term non-attested data to refer to any kind of data that has not been observed or recorded as having been produced by a particular speaker or writer. This
covers predominantly the use of introspection thinking and reasoning about language use based on ones own knowledge and experience of language. The term
non-attested does not, however, refer to recourse to anecdotal evidence data
that has been observed by the researcher at a particular time without necessarily
having information on when it was observed, by whom it was produced, what the
context of use was etc. This kind of attested but anecdotal data can sometimes be
useful, even if systematic data collection is to be preferred. For example, whereas
Bublitz (1998) uses corpus evidence to argue that totally has a negative semantic
prosody (Louw 1993), Bednarek (2008c) uses informal observations of the use of
totally on broadcasts on MTV to argue that, in AE young adult language, totally
does not exhibit such a prosody.
Much classic research in pragmatics seems to rely on introspection, particularly that related to logics and language philosophy (Austin 1975, Searle 1969,
Grice 1989, Kompa and Meggle in this volume). Adolphs (2008) goes as far as
stating that most research in pragmatics has not used attested examples of language-in-use (Adolphs 2008: 7). One example is speech act theory, although more
recent approaches in pragmatics study speech acts with respect to elicited (e.g. discourse completion tasks; see below) or corpus data. For instance, Adolphs (2008)
focuses on describing and exploring lexico-grammatical patterns that are associated with speech acts. Thus, why dont you can be related to the act of suggesting
and could you just can be related to the act of requesting. She also explores the distinction between direct and indirect speech acts with the help of corpus data.
In general, introspection can be useful for forming hypotheses, but these need
to be applied to attested language data for confirmation, rejection or modification.
Indeed, many corpus linguists would argue that what we think about language use
does not always correspond to language use. The focus of introspection also seems
to be more on what is possible in language use, rather than on what is actual, typical, salient or frequent.
540
3.2.
Monika Bednarek
Attested data: elicited and non-elicited
Attested data means language data which was actually produced by speakers/
writers on particular occasions. Such data can be elicited (by the researcher) or can
have occurred more naturally, that is, without being produced in a research context.
3.2.1.
Elicited data
Under the heading of elicited data we can include interviews (involving researchers), verbal reports, elicited narratives, discourse completion tasks, and experimental data.
3.2.1.1. Interviews
Interviews are a standard way of collecting data across the social sciences (see
e.g. Miller and Glassner 2004), including linguistics. There are various kinds of
interview data. Structured interviews use prepared questions, which are often
standardized, and thereby pre-define topics, whereas unstructured interviews
leave room for topics to be explored, and are more like an informal conversation (McNeill and Chatman 2005: 56). Semi-structured interviews feature a
combination of the two. While unstructured interviews thus may be more authentic, they are also more subjective, less representative, and tend to involve a
smaller number of participants. Depending on mode, interviews can be in written
or spoken form (with written interviews shading into questionnaires; see below);
depending on participants, they can involve only one participant or more than
one. Sometimes focus groups are set up featuring discussions in which one or
more researchers moderate the discussion and interaction between members of a
group.
There are at least two ways in which interview data are used in research. Either
what people say in interviews is seen as a reflection or description of what they
actually do/think or what people say in interviews is itself analyzed as a particular kind of linguistic construal. Benwell and Stokoe (2006) write:
Unstructured interviews and focus groups as data collection methods have been the
focus of much debate and discussion across the social sciences, largely due to the way
they are often treated as a transparent description of opinion or unmediated report of
practice. [] Some commentators argue that interview data must be treated as a researcher-driven, occasioned and situated account of identity and practices. (Benwell
and Stokoe 2006: 194)
There is a point in arguing that interviews may tell us what people think but not
necessarily what they do or even that interview data are too much influenced by the
interview situation, relations between interviewer and interviewee, social values
541
etc to tell us anything about the topic of the interview. Wetherell et al (2001) ask:
when a researcher interviews someone, how far can their talk be treated as a reflection of something else, such as established memories and ready-formed
opinions? Should it instead be analysed as the place and process in which memories and opinions are constituted? (Wetherell et al 2001: 7). Group discussions in
focus groups can raise new problems of about-ness: perhaps the interaction is
now about group relations rather than the apparent topic of discussion? (Wetherell
et al 2001: 28).
It may thus be more useful to use interview data to explore not what participants think about or how they handle the ostensible topic of the interview, but to
focus the analysis on other issues of the linguistic data. For example, Galasinski
(2004) makes use of interviews with 26 male and 29 female participants, where the
interview topic was a variety of issues related to middle age [], the experience
of its onset, and its relation to family and working relationships (Galasinski 2004:
24). He then explores the data with respect to emotion and gender, looking in particular at how emotion narratives (Galasinski 2004: 24) work in the construal of
masculinity. In particular, participants can choose to distance themselves from
emotions using a variety of distancing strategies, such as not directly describing
themselves as experiencing a particular emotion. This is the case in this interview
extract where a situation is described as being traumatic/dramatic rather than as directly and explicitly referencing the speakers emotion (as in the alternative I really
hated it or I was really sad):
(1)
I: and can you describe what it was like when your daughter moved out?
IF: well (.) permanently (.) it sort of crept up on us. I mean the night we left her
at university (.) she didnt go in halls, she went into a students house. And that
night was pretty traumatic, the drive back down the M6 was pretty dramatic.
[]
(Galasinski 2004: 53, my bold)
Such distancing strategies can be used by male speakers to maintain the pretence
of meeting the requirements of the dominant model of masculinity (Galasinski
2004: 73). Such use of data thus includes considerations of context and dominant
ideologies (e.g. prevailing models of gender), and seems a more interesting way of
looking into emotion narratives than interviews where participants are asked about
what they think about emotion and gender or where men and women are asked
about their emotional experience.
In general, interviews have a number of methodological advantages, as they are
argued to be
a natural and socially acceptable way of collecting information that most people feel
comfortable with and which can be used in a variety of situations and focusing on di-
542
Monika Bednarek
verse topics to yield in-depth data. The interviewers presence allows for flexible approaches, probing into any emerging new issue, while the interview guide helps to
maintain a systematic coverage of the domain. (Drnyei 2007: 143)
However, interview data are clearly not naturally occurring; further, to set up and
run interviews and to analyze the resulting data can be challenging and time-consuming; and there are also questions concerning the nature of the data with respect
to the interview situation (as discussed above). As already noted, it is problematic
to consider interview data as reflecting reality or producing reliable stories of experience; rather, interviews concern the repetition of familiar cultural tales (Miller and Glassner 2004: 125) and the interview is obviously and exclusively an interaction between the interviewer and interview subject in which both participants
create and construct narrative versions of the social world (Miller and Glassner
2004: 125).
3.2.1.2. Verbal reports
Verbal reports are self-reflections on participants language use, and are primarily
used in second language studies, interlanguage studies and cross-cultural research
(Woodfield 2007: 2). Verbal reports are employed by researchers to reconstruct
cognitive processes in language use, and to find out about speakers pragmatic
knowledge and their choice and formulation of speech act strategies (Woodfield
2007: 2). Verbal reports can be concurrent (on-line) e.g. as think aloud, or they
can be retrospective (subsequent). However, it is necessary to point out that when
interviewing participants after interactions, participants might not themselves be
able to recollect their own interpretations at a later stage (Adolphs 2008: 41), and
that trying to access cognitive processes is problematic because such processes
may be unconscious and thus not amenable to self-reporting. As with interviews, it
is worth problematizing whether or not verbal reports can be a reflection of actual behavior, cognitive processes and knowledge structures. Further, the influence
of the researcher and fellow informants on the resulting verbal report cannot be neglected (Woodfield 2007, citing Haastrup 1987).
3.2.1.3. Discourse completion tasks
Much cross-linguistic research in pragmatics (e.g. the Cross-Cultural Speech Act
Realization Project) or research in Second Language Acquisition into speech acts
uses oral or written discourse completion tasks (DCTs). DCTs were first introduced
by Levenston (1975) for testing communicative competence and then employed by
Blum-Kulka (1982) for speech acts (Rose and Ono 1995: 192). Written DCTs are
questionnaires that involve a description of different situations, potentially some
dialogue, and empty slots for subjects to produce a particular speech act or to fill in
appropriate responses (e.g. to a compliment). For example, to elicit English re-
543
quests, the following description may be used: You have been asked by your lecturer to give a seminar paper but you need more time. You go to see her to ask for
extra time to prepare your paper. Ask for an extension (from Woodfield 2007, who
also gives an overview of different request situations used in research). Participants in DCT resist such tasks to some extent, commenting on the non-interactive
and inauthentic nature of DCTs and their lack of contextual data (Woodfield 2007).
Although only limited efforts have been made to evaluate the validity of this
method (Yuan 2001: 272), it has been suggested that both written and oral DCTs
have a number of limitations in that they do not capture dynamic discourse features
such as turn-taking, conversational structure, speaker-listener interaction, and
pragmatic features (Kasper 2000, cited in Woodfield 2007; see also Yuan 2001).
Thus, responses given in DCTs are shorter in length, simpler in wording, less
face-attentive and negotiatory in the negotiation process, and less emotionally involved than naturally-occurring speech (Yuan 2001: 272). They may capture
pragmatic knowledge rather than usage. In as far as utterances in DCTs are more
decotextualized, more decontextualized and less multimodal than natural discourse, they can be argued to neglect the complexity of language usage. Adolphs
(2008) notes that the fact that data generated through DCTs is based around single
utterances distorts the overall picture of a speech act which is often negotiated and
developed over a number of turns in a dynamic discourse event. It is therefore unclear whether DCT or role play data display the same patterns and attributes as naturally occurring data (Adolphs 2008: 22).1
A further disadvantage of DCTs is that they are difficult to construct, and there
is a need for rigor in questionnaire construction and for extensive pilot-testing of
DCTs to control for various possible effects for prompt, rejoinder, and so forth
(Rose and Ono 1995: 198). As with other questionnaires, general issues that need
to be considered concern the sampling of the population of the participants (on
sampling see e.g. Drnyei 2007: 95100),2 its design (e.g. formulation of situation
in terms of appropriate detail, multimodality) and administering the DCT.
On the other hand, DCTs, like other questionnaires, allow for control and quantification; they are replicable; they are not time-consuming; and they are versatile.
For a more detailed discussion of DCTs in pragmatics see Woodfield (2007), who
proposes that DCTs represent highly constrained instruments of data collection in
terms of the degree to which the data is predetermined by the elicitation instrument (Woodfield 2007: 5). With respect to the distinction between oral and
written DCTs, the former share more features of natural speech than the latter
(Yuan 2001), and thus may be preferable. Yuan concludes that an oral DCT
generates data that are, to a certain extent, representative of natural speech; it allows researcher control; it is efficient in terms of the amount of data it can elicit: and it is ethical. Nevertheless, [] the oral DCT suffers similar drawbacks as do the written DCT
[sic] in that it cannot elicit elaborated negotiations and indirect compliment exchanges
seen in everyday conversations. (Yuan 2001: 289)
544
Monika Bednarek
3.2.1.4. Experimental data

Experimental data in pragmatics and related areas are used in educational research
(see Drnyei 2007), with respect to child language acquisition as well as in cognitive linguistics. For example, experimental research on understanding and recall
asks subjects to reproduce sentences with the help of specific cues. Such research
has suggested that understanding may proceed with the help of mental models
(frames, schemas etc, see Bednarek 2005). With respect to the sentence The fish attacked the swimmer, the word shark worked better as a recall cue than fish, suggesting that readers constructed a mental model for the sentence (Brown and Yule
1983: 251, reporting on research by Anderson et al 1976). Experimental pragmatics bridges pragmatics and psycholinguistics. The 4th Experimental Pragmatics conference in April 2009 included research on topics such as semantic approaches, language development, impaired communication, non-verbal and
non-human communication, non literal language, negation, presupposition, reference, conditionals, sentence processing. (http://experimentalpragmatics.org/, last
accessed 17 October 2008).
With respect to experimental research it is questionable whether it is possible to
find out about how people behave in real life rather than in the artificial and simplified lab situation. Other problematic issues are related to how many subjects are
used and how representative they are, and to what degree they are influenced in
their language use by the experimental situation and the presence of the experimenter. Some of these factors may be less problematic in field experiments that do
not take place in a lab but rather in a real-world environment such as a classroom. With the help of a carefully constructed experiment and control groups, it is
argued that the advantage of experimental data is that they allow for the testing of
hypotheses, the control of variables and the analysis of cause-effect relationships.
(For a more detailed discussion of experimentation-based pragmatics see contributions to Noveck and Sperber 2004, Breheny in this volume.)
3.2.1.5. Elicited narratives
An important strand of research on language use analyzes elicited narratives, employing various techniques (including the use of non-linguistic data) to elicit narratives from subjects. For example, participants are shown a story represented in a
sequence of pictures or a film narrative and are then asked to retell the story. Thus,
Bamberg (1991) shows subjects a 24-picture story called Frog, where are you?
and subsequently asks them to tell the story depicted in this sequence of pictures.
The data are then analyzed in terms of narrative structure and narrative activity.
Similarly, Labovs classic analysis of the structure of narratives (abstract, orientation, complicating action, evaluation, result/resolution, coda) is based on elicited
narratives. Tannens (1993) data stem from a project in which small groups of
545
young Greek and American women watched a film (the pear film),3 and afterwards talked about its contents to another woman, who they were told had not seen
it (Tannen 1993: 21). Tannen analyzes the womens accounts of the film and categorizes sixteen types of linguistic evidence such as negatives and repetitions
which reveal their expectations or, in cognitive linguistic terms, their schemas (see
above). Thus, elicited narratives allow a variety of linguistic enquiries about topics
including narrative structure, the construal of narratives, and cognitive phenomena
to name but a few. However, they share some disadvantages with interviews, for
instance that they are not naturally occurring, and that it can be very time-consuming to elicit the data in the first place.
With respect to all elicited data discussed in the above sections (interviews,
verbal reports, discourse completion tasks, experimental data, and elicited narratives) as well as other kinds of elicited data not discussed here (such as role plays
and different types of questionnaire,4 e.g. open or multiple choice questionnaires)
the disadvantage is that they are elicited rather than naturally occurring. In
contrast, the following sections look at different types of non-elicited data as used
in pragmatics and related research into language use.
3.2.2.
Non-elicited data
Under the heading of non-elicited data we can include observational ethnographic

data and different kinds of text/discourse data, including corpora.
3.2.2.1. Ethnographic data
Ethnography has its roots in anthropology, in particular research by Bronislaw
Malinowski and Franz Boas as well as in the linguistic works of Edward Sapir (Titscher et al 2000: 90). It is self-reflexive, culturally founded and based on participant observation (Titscher et al 2000: 90). Ethnographic methods have emerging
research foci, based on what evolves from the researchers immersion in a particular context for a period of time (at least 612 months). Thus, participant observation is a crucial method in ethnography, which may be complemented by other
techniques such as interviews or document analysis (McNeill and Chapman 2005:
89). However, not all participant observation is ethnographic: A study becomes
ethnographic when the fieldworker is careful to connect the facts that s/he observes
with the specific features of the backdrop against which these facts occur, which
are linked to historical and cultural contingencies (Baszanger and Dodier 2004:
12). In other words, texts are interpreted in terms of cultural structures or used to
reconstrue such structures (Titscher et al 2000: 91).
The participant in participant observation can be actively involved in the social
group/activity which is the focus of research (covertly or overtly) but need not be.
His/her observations can be recorded in observational schemas, reports or field-
546
Monika Bednarek
notes. Data can be recorded and then transcribed or jotted down in notebooks as it
occurs. Yuan (2001) points out that the latter type of linguistic data is not very accurate in capturing the details of language use. When participant observation is
overt, the observers paradox (Labov 1972, cited in Wardhaugh 1986: 18) may play
a role. This paradox is that the aim of linguistic research is to find out how people
talk when they are not being systematically observed, but the data are available
only through systematic observation (Wardhaugh 1986: 1819). Ethnographic
data can be argued to show the complexity and dynamics of authentic behavior;
ethnographic methods are flexible and open to the observation of new phenomena.
However, ethnographic research is time-consuming, not replicable or quantifiable,
and not necessarily representative. It may, however, be ideal for generating initial
hypotheses about something totally unknown (Drnyei 2007: 133). While different types of ethnographic research exist (Baszanger and Dodier 2004: 19) in linguistics it is usually associated with John Gumperz and Dell Hymes and their ethnography of speaking or communication (e.g. Gumperz and Hymes 1972). For
instance, Hymes is interested in speech communities, speech events, speech styles,
and ways of speaking as units of analysis (Hymes 1972: 5358). Within speech
acts he proposes to look at various components such as setting, scene, participants,
outcomes and goals, key (tone/manner), genres, norms of interaction and interpretation. Key questions in such research are: Who speaks, with whom, when, where
and in what code, about what? (Titscher et al 2000: 95).
3.2.2.2. Texts/discourse data
With respect to the use of spoken and written text/discourse data, we can identify a
continuum in terms of size, with analyses ranging from individual case studies of
one or just a few texts, to the use of small-scale corpora consisting of a range of
texts to the recourse to large-scale corpora of hundreds of thousands or millions of
words.
Continuum of text/discourse data
Individual text(s):
case study/ies
small-scale
corpus
large-scale
corpus
Figure 1
Discourse data also vary in terms of their content and context of production such
that analyses can focus on political documents, speeches, media texts, conversations, radio interviews, reality TV talk, books, academic articles, advertisements, songs, poems, television series, university prospectuses, chat room interaction or websites to name but a few. These types of data offer different advantages.
For example, written language is usually easier to collect and to study than casual
547
conversation. Public spoken language (e.g. radio or internet broadcasts) is easier to

get hold of than conversations among friends or partners. Printed text
is infinitely easier to collect than audiovisual data and more permanent than most Web
material. It is already out there, ready to be gathered, and does not require time-consuming transcription before analysis. Also, the observers paradox is not an issue. Unlike, say, informants elicited responses or structured interviews, printed material is not
in danger of changing just because it is being observed. (Mautner 2008: 32)
In contrast,
[t]he variation in the types of contexts recorded as part of different spoken corpora, as
well as in the parameters used for collecting data, makes it difficult to accumulate a
large enough sample of spoken data that follows the same design principles. As a result,
research based on spoken data is often limited in terms of the number of examples of a
particular linguistic feature in the corpus, especially once contextual filters have been
applied. (Adolphs 2008: 33)
With respect to the recording of spoken data, the observers paradox may again be
relevant, since people may talk differently when they are being recorded, although
there are techniques for minimizing this problem, such as recording people for
long periods of time. Spoken data also needs to be transcribed according to specific
transcription systems (see Kreuz and Riordan in this volume). Since the emergence
of the internet and other technologies, data from the new media (such as emails,
chats, social forums, message boards etc) are also increasingly studied in pragmatics (Bublitz 2008; Gruber 2008).5 The majority of the new media incorporate
semiotic systems other than language, and multimodal research becomes important. Wodak (2008) even speaks of the visual turn (Wodak 2008: 2) in linguistics,
necessitating new toolkits and theories. Multimodality is a very vibrant field in linguistics nowadays, with many different approaches and notation/transcription systems (e.g. Hbler 2007; Jewitt 2009). In general, multimodal analysis, whether of
traditional or of new media, can become highly complex and time-consuming
(see Pollak 2008 on film analysis).
These and other types of spoken and written discourse data may be used in case
studies, small-scale corpus studies and discourse analyses, and large-scale corpus
linguistics. (Detailed descriptions of different kinds of corpora are given in most
introductions to corpus linguistics.) The way in which these data are analyzed
varies widely (see below and Andersen as well as Ruehlemann in this volume), but
all methodologies have in common the focus on the actual (rather than the possible) and all value authentic language-in-use. There are a variety of approaches that
usually analyze individual texts or small-scale collections of texts (including
specially designed corpora) such as ethnomethodological conversation analysis
and membership categorization analysis (Schegloff 2006; see also Holly and
Domke in this volume), interactional sociolinguistics, politeness theory, discursive
psychology, critical discursive psychology, content analysis, grounded theory, eth-
548
Monika Bednarek
nographic methods (discussed above), narrative semiotics, SYMLOG (system for

the multiple level observation of groups), critical discourse analysis, (German)
functional pragmatics, distinction theory text analysis, objective hermeneutics,
and systemic functional discourse analysis. See Taylor (2001), Stubbe et al (2003),
and Titscher et al (2000) on evaluations of and comparisons between these approaches to discourse data.
The advantages of looking at single texts or a limited number of texts or smallscale corpora are that the texts can be analyzed in depth through close-reading and
attention to manifold aspects of language use. Such an analysis allows researchers
to focus on the localized construal of social phenomena such as identity in particular contexts, often resulting in a complex, rich, interpretive, dynamic and flexible
analysis of micro-contexts and capturing the dynamic and negotiatory nature of
much language use. If video data are used (Heath 2004), the multimodal context can
be taken into account and attention can be paid not only to intonation and prosody
but also to gesture, posture, body movement and facial expression. However, while
individual case studies allow a focus on significant or important texts and events,
they lose out on generalizability and representativeness. They raise the question
of how far statements made about one particular text are transferable to other contexts. Case studies can also be time-consuming, relatively subjective, and are not
easily replicable. Analyses that make use of small-scale corpora or collections of
texts are more representative but can be just as subjective depending on the features
analyzed and the method of analysis. In addition to the advantages of case studies,
they offer observation of recurring regularities, practices or patterns, at least to
some extent. Thus, conversation analyses have uncovered regularities of turn-taking and sequencing using collections of spoken data. Small-scale corpus studies
have, for instance, been undertaken by Semino and Short (2004) with respect to reported speech and thought, by Bednarek (2006b) focusing on evaluative meaning in
a 70,000 word corpus of hard news stories, by Martin and White (2005: 165) on a
systemic functional linguistic appraisal analysis in 85 items from journalistic discourse and by Bednarek (2008a) focusing on emotion talk in an 85,000 word corpus
comprising conversation, news reportage, fiction and academic discourse. The corpora used in this research are small enough for manual, context-sensitive analysis
but large enough to show at least some patterns and allow some generalizability.
Coding can be supported through computer assisted manual annotation (CAMA),
where a human annotates the text in terms of patterns that generally computers
cannot recognize (ODonnell 2007). The latter makes use of smaller corpora because manual coding takes up time but it allows the researcher to explore semantic
and pragmatic patterning. ODonnell has designed computer software that facilitates such analyses: the UAM CorpusTool (http://www.wagsoft.com/CorpusTool/),
but other coding/annotation software also exists for use across the social sciences.
Finally, a few words on large-scale corpus data. Corpus linguistic applications
include discourse analysis, lexicography, stylistics, forensic linguistics, language
549
variation studies and language teaching (Baker 2006: 23). Although corpora have
in the past been neglected in pragmatics, corpora are increasingly used in studies of
language use, such as e.g. Adolphs (2008) and contributions to Skaffari et al
(2005). Compare also the three sub-issues on corpus linguistics in the Journal of
Pragmatics (2004: 36/2, 36/6 and 36/9).
A crucial advantage of corpus linguistic investigations lies in the representativeness of the data and the use of empirical, systematic, and sophisticated data. The
analyses are often reliable and replicable, especially when focusing on formally defined items. Further, corpus linguistics can uncover features of language that are inaccessible to intuition or a close reading type analysis of individual texts. This concerns patterning, typicality of usage, and quantification (e.g. type-token ratios,
frequency lists, lexical density) and particular kinds of semantic-pragmatic meanings such as semantic prosody (Louw 1993, Bednarek 2008c). Corpus data also
give researchers access to shared connotations (Coffin and OHalloran 2005) and
the experience of language users (Thompson and Hunston 2000: 15). They allow us
to study typicality adequately through access to large amounts of actually occurring
discourse, and enable easy application of tests of statistical significance through
user-friendly software. Thus, hypotheses can be verified, falsified or modified and
new language features can be uncovered with the help of large scale corpora.
However, there are also some disadvantages6 in that researchers are only able
to search the corpus for formally defined items, unless the corpus data have previously been annotated manually or semi-automatically (in which case there is a
limitation on the size of the corpus because this process can be quite time-consuming). Social categories such as gender or age are often treated as stable variables
rather than as discursively construed (Butler 1999), and many (though not all) corpus analyses fail to take into account reception, the dynamic nature of discourse, its
context or textual structure. As Stubbs (2008) notes with respect to keyword analysis, since the texts have been ripped apart into lists of individual words and/or
n-grams, the patterns ignore text segmentation. They are a feature of global textual
cohesion, but not textual structure (Stubbs 2008: 5). It has also not yet been discussed sufficiently to what extent statistical tests are adequate for data of authentic
language-in-use, to what extent linguists can evaluate and understand the information provided by statistics correctly, and to what extent they are able to
choose the most appropriate statistical tests for their analyses.
The majority of corpora are mono-modal, and the incorporation of semiotic
systems other than language is not yet wide-spread. With respect to multimodality,
Bauldry and Thibault (2006) stress the importance of using multimodal corpora in
semiotic analysis while noting that their design is very much in its infancy
(Bauldry and Thibault 2006: 181). Johnston (2008) has forcefully argued for the
need for sign language corpora.
Corpus analyses are also usually interested in the repeated, the typical, the frequent, rather than the individual and outstanding (but see e.g. Hoey 2005 on cre-
550
Monika Bednarek
ativity). For example, some texts are highly valued in the community or []
[have] special significance in some domain such as history or politics. Here we
treat texts as artefacts objects of study in their own right (Matthiessen 2006:
108). An emphasis on repeated patterns of semiotic behavior neglects the importance of outstanding singular texts that might have an impact on phylogenesis and
ontogenesis that goes beyond the impact of ordinary singular texts that achieve
their impact through repetition. An example from environmental discourse would
be the documentary An Inconvenient Truth produced and presented by Al Gore,
which has been widely credited with responsibility for a seismic shift in the attention of politicians, governments and the media towards environmental matters and
in policy changes around issues of environmental sustainability. In fact, its impact
was recognized in the awarding of the Nobel Peace Prize to Gore in 2007 for efforts to build up and disseminate greater knowledge about man-made climate
change, and to lay the foundations for the measures that are needed to counteract
such change (http://nobelprize.org/nobel_prizes/peace/laureates/2007/). The impact of such extraordinary texts, which are more powerful than normal texts produced countless times by speakers over time, presumably depends on diverse factors such as the profile of the author/speaker, the context of reception, and in the
case of An Inconvenient Truth, the sophisticated use of semiotic resources (Bednarek, Caple and Hood 2008). Studying such unique texts, then, also provides important insights into the construal of categories, and is an argument for qualitative
analyses of individual texts.
Further, some publications are more read than others and thus might have
greater impact on phylo- and ontogenesis:7 According to claims, the most likely
document that an ordinary English citizen will cast his or her eyes over is The Sun
newspaper (Sinclair 2005: 7). While it is not clear what exactly the impact of such
(valued and/or wide-spread) texts is, this has at least two implications for largescale corpus data. On the one hand, it needs to be acknowledged that special texts
become lost in the corpus where they have only as much weight as any other text:
[F]requent patterns of language do not always necessarily imply underlying hegemonic
discourses. Or rather, the power of individual texts or speakers may not be evenly distributed. A corpus which contains a single (unrepresentative) speech by the leader of a
country or religious group, newspaper editor or CEO may carry more weight discursively than hundreds of similar texts which were produced by ordinary people. (Baker
2006: 19)
On the other hand, this question clearly needs to be discussed in relation to corpus
design. For example, with respect to the quote above, Sinclair asks if we should include more texts from The Sun than from other newspapers in a corpus of British
English on account of its importance (Sinclair 2005: 7). In fact, many texts that are
included in corpora such as the British National Corpus differ in circulation status
(size of readership/circulation level) (Lee 2001: 68). Sinclair notes that issues of
551
representativeness and balance in corpus design are far from resolved at present
(Sinclair 2005: 79) and recommends that in the meantime corpus design should
be documented fully with information about the contents and arguments in justification of the decisions taken (Sinclair 2005: 8). With respect to corpus design
(Wynne 2005) other problematic issues include the thorny notion of representativeness (Mahlberg 2004), procedures of sampling (see Mautner 2008, Voormann
and Gut 2008), and practicalities of coding and inputting as well as technological
exigencies. It may also be difficult to find a corpus that has enough natural occurrences of individual pragmatic phenomena that the researcher is interested in (such
as speech acts, see Yuan 2001: 275) and not all pragmatic phenomena can be formally defined necessitating the use of small-scale corpora at least in a pilot study
or as complementary data.
4.
Conclusion: triangulating the data of pragmatics
In view of the above observations, it seems that all data of pragmatics and associated methodologies provide certain insights and perspectives, without necessarily
being able to capture the complexity of language in use in its entirety. The use of
different data and methodologies contributes variability to the field of pragmatics.
With respect to this, Kasper and Dahl (1991: 215) speak of a double layer of variability: variability that reflects the social properties of the speech event, and the
strategic, actional, and linguistic choices by which interlocutors attempt to reach
their communicative goals and the variability induced by different instruments
of data collection (Kasper and Dahl 1991: 215, cited in Woodfield 2007).
It thus seems reasonable to argue for a mixed-methods approach (Drnyei
2007) in pragmatics, combining quantitative and qualitative research. Triangulation, or the use of a variety of methods, allows researchers to cross-check and
verify the reliability of a particular research tool and the validity of data collected
(McNeill and Chatman 2005: 23). Researchers arguing for a combination of
methods include Holmes (1997), Taylor (2001), Holmes and Schnurr (2005),
Baker (2006) and Wodak and Krzyzanowski (2008).
If there is a team of researchers with various backgrounds, triangulation may be
less difficult than if there is one individual researcher only. Large research projects
with research teams are thus at an advantage, as they can combine expertise in different research methodologies and distribute the analyses among researchers. An
individual researcher may not be able to gain sufficient expertise in various research methods, and there are limitations to the extent to which s/he can undertake
different types of analysis for a single research project. One possibility of triangulation for discourse analysis is the three-pronged approach suggested by Bednarek (2008a, b, 2009). In essence, such an approach extends previous linguistic research that involves both corpus and discourse analysis, for instance corpus-based
552
Monika Bednarek
discourse analyses (Baker 2006), recent sociolinguistic research (e.g. Holmes and
Schnurr 2005), studies in Critical Discourse Analysis (Fairclough 2000; Barker
and Galasinski 2001: 26; Mautner 2008), and Matthiessens (2006) two-pronged
approach (Matthiessen 2006: 110). The three-pronged approach combines a
large-scale quantitative approach with a small-scale corpus analysis to be complemented by qualitative case studies. The large-scale corpus studies use a large (~
more than 500,000 or 1 million words) corpus and computer software to investigate the frequency and distribution of lexico-grammatical items. If the corpus design is reasonable, it can provide a relatively representative sample of language
that allows us to formulate hypotheses about general functions of the respective
lexico-grammatical items. On account of the large size of the corpus, however, the
description is less detailed than would otherwise be possible. In contrast, a smallscale corpus analysis uses a corpus of a size that is amenable to manual or semiautomated analysis, e.g. of semantic or pragmatic meanings, and that still shows us
intertextual patterning of such meanings. On account of its smaller size, such a corpus will be less representative, but its analysis will be more detailed and more interpretive. Finally, undertaking one or several case studies allows for qualitative
discourse analysis, which can take into account textual development and the sociocultural context (e.g. of production and reception) of the discourses at hand. It will
be least representative, in that researchers will be least able to generalize from findings, but its analysis will be most detailed and most interpretive. Thus, applying
these three methodologies can provide us with various perspectives on the discourse(s) at hand. However, even with this three-pronged discourse analytical approach, a time span of at least two to three years is necessary for the development
of theory, methodology, corpus design, software expertise, and analysis. In terms
of its focus on discourse, the analysis is also limited with respect to the various
ways in which this discourse can be taken up by its readership or audience. Generally speaking, the development of other mixed-method approaches in discourse
analysis and pragmatics is called for, and there is also a need for further evaluation
of how to approach the data of pragmatics, as there are only few studies that evaluate data-gathering and analysis in pragmatics (cf. Yuan 2001: 273 on the dearth of
such studies within empirical pragmatics). As Rose and Ono (1995) put it: We are
not claiming that one source of data is better than any other; we know too little to
make such claims and, frankly, we are sceptical of their value. It would be more
productive to ask better for what purpose? (Rose and Ono 1995: 207).
Notes
1. Research that compares DCT and other data includes Bardovi-Harlig and Hartford
(1993), Golato (2003), and Yuan (2001). Rose and Ono (1995) give a useful overview of
research on the methodology of DCTs in speech act studies.
553
2. Sampling procedures are relevant to many approaches to the data of pragmatics: different
types include ad hoc sampling, cluster sampling, multi-stage sampling, quota sampling,
random sampling, and stratified sampling (Titscher et al 2000).
3. About the history of the pear film:
In the mid-1970s Professor Wallace Chafe, a specialist in Native American languages,
decided to test how much a simple story will vary from language to language. With his
research team, he designed a very simple film to elicit stories from speakers around the
world, including uneducated speakers in rural areas. []
The Pear Stories film was designed to tap into universal experiences, by showing a
man harvesting pears, which are stolen by a boy on a bike. The boy has some other adventures with other children, before the farmer discovers that his pears are missing. The
film is six minutes long, in colour, with sound effects but no words. It was filmed in
northern California, near the University of California, Berkeley. The man who plays the
farmer is a Cuban.
The story line is deliberately loose and bland, to avoid imposing a strong U.S. cultural
bias. But the film makers deliberately inserted a short scene of a man leading a goat to
test descriptions of a background event with no later significance. The scene of falling
off the bike and spilling the pears can measure language for cause and effect. And the
unusual ping-pong toy tests how people describe an unfamiliar object. The final scene,
when the farmer discovers his fruit has been stolen, re-introduces a character who had
been off-screen for most of the film, and stimulates speakers to describe emotions and
state a moral. (http://www.pearstories.org/docu/ThePearStories.htm, last accessed
14 October 2008)
4. Questionnaires differ with respect to their object of analysis (e.g. factual/behavioral/attitudinal), their use or non-use of scales (e.g. the well-known Likert scale), whether they
are open-ended or closed-ended etc (Drnyei 2007).
5. Computer mediated communication (CMC) can be studied or can itself be made use of as
a research tool, with certain implications. On web-based studies compare e.g. Drnyei
(2007: 121122).
6. Baker (2006: 1721) discusses some limitations of corpus data and analysis, and Paltridge (2006: 174176) provides a useful, if brief, summary of and response to criticisms
of corpus studies.
7. Additionally, there may be differences in impact depending on who says what to whom
how in what context, and mental factors such as cognitive biases (Baker 2006: 11) are
also involved in discourse production and reception.
References
Adolphs, Svenja
2008
Corpus and Context. Investigating Pragmatic Functions in Spoken Discourse.
Anderson, Richard C., James W. Pichert, Ernest T. Goetz, Diane L. Schallert, Kathleen V.
Stevens and Stanley R. Trollip
1976
Instantiation of general terms. Journal of Verbal Learning and Verbal Behaviour 15: 667679.
554
Monika Bednarek
Austin, John L.
1975
How to Do Things with Words. 2nd ed. London: Oxford University Press.
Baker, Paul
2006
Using Corpora in Discourse Analysis. London/New York: Continuum.
Bamberg, Michael
1991
Narrative activity as perspective taking: the role of emotionals, negations and
voice in the construction of the story realm. Journal of Cognitive Psychotherapy 2: 275290.
Bardovi-Harlig, Kathleen and Beverly S. Hartford
1993
Redefining the DCT: Comparing open questionnaires and dialogue completion
tasks. Pragmatics and Language Learning 4: 143165.
Barker, Chris and Dariusz Galasinksi
2001
Cultural Studies and Discourse Analysis: A Dialogue on Language and Identity. Thousand Oaks: Sage.
Baszanger, Isabelle and Nicholas Dodier
2004
Ethnography: relating the part to the whole. In: David Silverman (ed.), Qualitative Research. Theory, Method and Practice, 934. London: Sage.
2006
System and Corpus. Exploring Connections, 164183. London: Equinox.
Bednarek, Monika
2005
Frames revisited: the coherence-inducing function of frames. Journal of Pragmatics 37: 685705.
Bednarek, Monika
2006a
Epistemological positioning and evidentiality in English news discourse a
text-driven approach. Text and Talk 26/6: 635660.
Bednarek, Monika
2006b
Evaluation in Media Discourse. London/New York: Continuum.
Bednarek, Monika
2008a
Emotion Talk across Corpora. Basingstoke/New York: Palgrave Macmillan.
Bednarek, Monika
2008b
What the hell is wrong with you? A corpus perspective on evaluation and
emotion in contemporary American pop culture. In: Ahmar Mahboob and
Naomi Knight (eds.), Questioning Linguistics. 95126. Newcastle: Cambridge
Scholars Press.
Bednarek, Monika
2008c
and Linguistic Theory 4/2: 119139.
Bednarek, Monika
2009
Corpora and discourse: a three-pronged approach to analyzing linguistic data.
In: Michael Haugh, Kate Burridge, Jean Mulder and Pam Peters (eds.), Selected Proceedings of the 2008 HCSNet Workshop on Designing the Australian
National Corpus: Mustering Languages, Sommerville, MA: Cascadilla Proceedings Project.
Bednarek, Monika, Helen Caple and Sue Hood
2008
Communicating a sustainable Australia: discourses of persuasion. Unpublished manuscript, Faculty of Arts and Social Sciences, University of Technology, Sydney.
555
Benwell, Bethan and Elizabeth Stokoe

2006
Discourse and Identity. Edinburgh: Edinburgh University Press.
1982
Learning to say what you mean in a second language. Applied Linguistics 3:
2959.
1987
1983
Bublitz, Wolfram
1998
I entirely dot dot dot: copying semantic features in collocations with up-scaling intensifiers. In: Rainer Schulze (ed.), Making Meaningful Choices in English, 1132. Tbingen: Narr.
Bublitz, Wolfram
2008
Sailing the islands or watching from the dock: the treacherous simplicity of a
metaphor. How we handle new (electronic) hypertext versus old (printed)
text. In: Andrea Gerbig and Oliver Mason (eds.), Language, People, Numbers
Corpus Linguistics and Society, 249273. Amsterdam: Rodopi.
Butler, Judith
1999
Gender Trouble: Feminism and the Subversion of Identity. New York: Routledge.
Cermk, Frantisek
2001
Research Methods in Linguistics. Essential Principles, Based on a General
Theory of Science. Prague: Karolinum.
Coffin, Caroline and Kieran OHalloran
2005
Finding the global groove: theorising and analysing dynamic reader positioning using APPRAISAL , corpus, and a concordancer. Critical Discourse Studies
2/2: 143163.
Drnyei, Zoltan
2007
Research Methods in Applied Linguistics. Oxford: Oxford University Press.
Fairclough, Norman
2000
New Labour, New Language? London: Routledge.
Galasinski, Dariusz
2004
Men and the Language of Emotions. Basingstoke/New York: Palgrave Macmillan.
Golato, Andrea
2003
Studying compliment responses: a comparison of DCTs and recordings of naturally occurring talk. Applied Linguistics 24/1: 90121.
Grice, Paul
1989
Gruber, Helmut
2008
Analyzing communication in the new media. In Ruth Wodak and Michal Krzyz anowski (eds.), Qualitative Discourse Analysis for the Social Sciences,
5476. Basingstoke/New York: Palgrave Macmillan.
1972
Directions in Sociolinguistics. The Ethnography of Communication. New
York: Holt, Rinehart and Winston.
556
Monika Bednarek
Haastrup, Kirsten
1987
Using thinking aloud and retrospection to uncover learners lexical inferencing procedures. In: Claus Faerch and Gabriele Kasper (eds.), Introspection in
Second Language Research. 197212. Clevedon: Multilingual Matters.
2004
The spoken language corpus: A foundation for grammatical theory. In: Karin
Aijmer and Bengt Altenberg (eds.), Advances in Corpus Linguistics. Papers
from the 23rd International Conference on English Language Research on
Computerized Corpora. 1139. Amsterdam/New York: Rodopi.
Heath, Christian
2004
Analysing face-to-face interaction: video, the visual and material. In: David
Silverman (ed.), Qualitative Research. Theory, Method and Practice. 266
282. London: Sage.
Hoey, Michael
2005
Lexical Priming. A New Theory of Words and Language. London: Routledge.
Holmes, Janet
1997
Women, language and identity. Journal of Sociolinguistics 1/2: 195223.
Holmes, Janet and Stephanie Schnurr
2005
Politeness, humor and gender in the workplace: Negotiating norms and identifying contestation. Journal of Politeness Research 1: 121149.
Hbler, Axel
2007
The Nonverbal Shift in Early Modern English Conversation. Amsterdam: Benjamins.
Hymes, Dell
1972
Models of the interaction of language and social life. In John J. Gumperz and
Dell Hymes (eds.), Directions in Sociolinguistics. The Ethnography of Communication, 3571. New York: Holt, Rinehart and Winston.
Jewitt, Carey
2009
The Routledge Handbook of Multimodal Analysis. London: Routledge.
Johnston, Trevor
2008
Signed languages and corpus linguistics. Paper presented at the 2nd International Free Linguistics Conference, University of Sydney, 1112 October
2008.
Kasper, Gabriele
2000
Data collection in pragmatics research. In: Helen Spencer-Oatey (ed.), Culturally Speaking. Managing Rapport through Talk Across Cultures, 316341.
London/New York: Continuum.
Kasper, Gabriele and Merete Dahl
1991
Research methods in interlanguage pragmatics. Studies in Second Language
Acquisition 13/2: 215247.
Labov, William
1972
Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press.
Lee, David
2001
Genres, registers, text types, domains, and styles: clarifying the concepts and
navigating a path through the BNC jungle. Language Learning & Technology
5: 3772. Retrieved from http://llt.msu.edu/vol5num3/lee/ on 20 October
2005.
557
Leech, Geoffrey
1983
Levenston, Edward
1975
Aspects of testing the oral proficiency of adult immigrants to Canada. In: Leslie Palmer and Bernard Spolsky (eds.), Papers on Language Testing 1967
1974, 6774. Washington, DC: TESOL.
Louw, Bill
1993
Irony in the text or insincerity in the writer? In: Mona Baker, Gill Francis and
Elena Tognini-Bonelli (eds.), Text and Technology. In Honour of John Sinclair,
Mahlberg, Michaela
2004
Corpus design and the words in a dictionary. Lexicographica 20: 11429.
Martin, James R. and Peter R.R. White
2005
The Language of Evaluation: Appraisal in English. Basingstoke/New York:
Palgrave Macmillan.
Matthiessen, Christian M.I.M.
2006
Frequency profiles of some basic grammatical systems: an interim report. In:
Susan Hunston and Geoff Thompson (eds.), System and Corpus: Exploring
Connections, 103142. London: Equinox.
Mautner, Gerlinde
2008
Analyzing newspapers, magazines and other print media. In Ruth Wodak and
Michal Krzyzanowski (eds.), Qualitative Discourse Analysis for the Social
Sciences, 3053. Basingstoke/New York: Palgrave Macmillan.
McNeill, Patrick and Steve Chapman
2005
Research Methods. 3rd ed. London: Routledge.
Miller, Jody and Glassner, Barry
2004
The inside and the outside: finding realities in interviews. In: David Silverman (ed.), Qualitative Research. Theory, Method and Practice, 125139.
London: Sage.
2004
Experimental Pragmatics. Basingstoke/New York: Palgrave Macmillan.
ODonnell, Mick
2007
The UAM Corpus Tool. Paper presented at ICAME 28, Stratford-upon-Avon,
UK, 2327 May 2007.
Paltridge, Brian
2006
Discourse Analysis. London/New York: Continuum.
Pollak, Alexander
2008
Analyzing TV documentaries. In: Ruth Wodak and Michal Krzyzanowski
(eds.), Qualitative Discourse Analysis for the Social Sciences, 7795. Basingstoke/New York: Palgrave Macmillan.
Rose, Kenneth and Reiko Ono
1995
Eliciting speech act data in Japanese: the effect of questionnaire type. Language Learning 45/2: 191223.
Schegloff, Emanuel
2006
A tutorial on membership categorization. Journal of Pragmatics 39: 462482.
Searle, John
1969
558
Monika Bednarek
Semino, Elena and Mick Short

2004
Corpus Stylistics: A Corpus-based Study of Speech, Thought and Writing in a
Corpus of English Writing. London: Routledge.
Skaffari, Janne, Matti Peikola, Ruth Carroll, Risto Hiltunen and Brita Wrvik (eds.)
2005
Opening Windows on Texts and Discourses of the Past. Philadelphia: Benjamins.
Silverman, David (ed.)
2004
Qualitative Research. Theory, Method and Practice. 2nd ed. London: Sage.
Sinclair, John
2005
Corpus and text basic principles. In: Martin Wynne (ed.), Developing Linguistic Corpora. A Guide to Good Practice, 116. Oxford: Oxbow.
Stubbe, Maria, Lane, Chris, Hilder, Jo, Vine, Elaine, Vine, Bernadette, Marra, Meredith,
Holmes, Janet and Ann Weatherall
2003
Multiple discourse analyses of a workplace interaction. Discourse Studies 5/3:
351388.
Stubbs, Michael
2008
Three concepts of keywords. Revised version of paper presented to the conference on Keyness in Text, Certosa di Pontignano, University of Siena, June
2007. Retrieved from http://www.uni-trier.de/fileadmin/fb2/ANG/Linguistik/
Stubbs/stubbs-2008-keywords.pdf on 7 August 2008.
Tannen, Deborah
1993
Whats in a frame?: surface evidence for underlying expectations. In: Deborah
Tannen (ed.), Framing in Discourse, 1456. Oxford: Oxford University Press.
Taylor, Stephanie
2001
Evaluating and applying discourse analytic research. In: Margaret Wetherell,
Stephanie Taylor and Simeon Yates (eds.), Discourse as Data: A Guide for
Analysis, 311330. London: Sage in association with The Open University.
Thompson, Geoff and Susan Hunston
2000
Evaluation: an introduction. In: Susan Hunston and Geoff Thompson (eds.),
Evaluation in Text. Authorial Stance and the Construction of Discourse, 127.
Titscher, Stefan, Michael Meyer, Ruth Wodak and Eva Vetter
2000
Methods of Text and Discourse Analysis. Translated by Bryan Jenner. London:
Sage.
2001
Voormann, Holger and Ulrike Gut
2008
Agile corpus creation. Corpus Linguistics and Linguistic Theory 4/2: 235251.
Wardhaugh, Ronald
1986
An Introduction to Sociolinguistics. Oxford: Basil Blackwell.
Wetherell, Margaret, Stephanie Taylor and Simeon J. Yates (eds.)
2001
Discourse as Data: A Guide for Analysis. London: Sage in associated with The
Open University.
Wodak, Ruth
2008
Introduction: Discourse studies important concepts and terms. In: Ruth
Wodak and Michal Krzyzanowski (eds.), Qualitative Discourse Analysis for
the Social Sciences, 129. Basingstoke/New York: Palgrave Macmillan.
559
Wodak, Ruth and Michal Krzyzanowski (eds.)

2008
Qualitative Discourse Analysis for the Social Sciences Basingstoke/New York:
Palgrave Macmillan.
Woodfield, Helen P.
2007
Problematizing discourse completion tasks: voices from verbal report. Evaluation and Research in Education 20/2. Retrieved from
http://hdl.handle.net/1983/943 on 14 October 2008.
Wynne, Martin (ed.)
2005
Developing Linguistic Corpora. A Guide to Good Practice. Oxford: Oxbow.
Yuan, Yi
2001
An inquiry into empirical pragmatics data-gathering methods: Written DCTs,
oral DCTs, field notes, and natural conversations. Journal of Pragmatics 33/2:
271292.
20.
Experimental pragmatics*
Richard Breheny
1.
Introduction
Pragmatics is a mature discipline, focussed largely on the use of language or communication in context. Experimental psychology is a mature discipline, focussed in
part on language and communication. It may come as a surprise then to find that
Experimental Pragmatics, that is, the development of pragmatics as a theoretical
discipline via experimental methods, is more of a fledgling field. While there had
been a small but influential tradition in psychological research informed by pragmatic theory (see for instance, Clark 1996), until recently there was comparatively
little theoretical pragmatics research founded on the experimental methods of the
psychological sciences. Likewise, while there is a strong tradition of psycholinguistic research on the interpretation of discourse in context (see Graesser et al
2003 for a recent handbook on the topic), until recently little of this research had
been much interested in the subject matter of pragmatics. Why this state of affairs
persisted for so long and how the situation is changing are questions that will be
addressed in this entry. In addition, the current state of experimental pragmatics
will be reviewed and future prospects surveyed.
Recent experimental pragmatics has focussed on Gricean pragmatics, which
considers phenomena such as implicature, presupposition, figurative language and,
more generally, speaker meaning. Thus the focus of this entry will reflect this bias.
2.
The foundations of pragmatics and experimental methods
Modern experimental psychology aims to develop scientific theories of individual

mental, cognitive or psychological capacities; normally of how certain functions,
such as vision, planning or memory, are realised. Contemporary pragmatics, particularly as it interfaces with formal linguistics, stems largely from the tradition
of analytic philosophy through the work of Austin, Grice, Carnap, Lewis among
others. It is not straightforward to apply the methods of experimental psychology
to this field of inquiry since thinkers like Grice and Austin were not in the business
of proposing accounts of individual pragmatic competencies. For analytic philosophy, questions about how we cognise language in context are separate from critical philosophical questions about meaning in context. If we see pragmatics as a
continuation of the analytic tradition, there would seem to be little scope for the application of experimental psychological methods in pragmatics. Things are not that
562
Richard Breheny
different if we consider more socially oriented pragmatics, which takes inspiration

from social sciences such as anthropology, sociology and economics. While these
disciplines are more empirically grounded, their methods rarely overlap with those
of the experimental psychologist.
Given the philosophical and social scientific foundations of pragmatics, perhaps we should not be surprised to find that Experimental Pragmatics has been a
minority pursuit, at least until relatively recently. However, pragmatics has also
flourished not so much as a branch of social sciences like anthropology or of analytic philosophy but as a branch of contemporary linguistics. Given the nature
of the more dominant formal end of this discipline, it is more surprising to find
the lack of experimental pragmatics. After all, most of contemporary formal linguistics takes itself to be part of the cognitive scientific enterprise, where theory is
ultimately aimed at explaining the cognitive psychological underpinnings of linguistic and pragmatic phenomena.
3.
Introspection and its limits in formal pragmatics
Since its inception, the main method for collecting empirical data in formal linguistics has been to introspect on intuitions (and check with a small number of informants if judgments are not sharp). While this methodology has served the discipline well in syntax and semantics, it might be questioned in the case of
pragmatics. For pragmatics interfaces with formal linguistics mostly in the domain
of generalised pragmatic inference. Probably the best known example of generalised phenomena are the scalar implicatures formalised (to a greater or lesser degree) in the neo-Gricean works of Gazdar (1979), Horn (1989), and Levinson
(Atlas and Levinson 1981, Levinson 2000).1 In these works, Grices informal account of implicatures is recast so that generalised implicatures become default.
I.e. the implicatures can be derived solely on the basis of the structure of the sentence used and only become unavailable (cancelled) under specific contextual
conditions. While neo-Gricean theory provides a computationally more tractable
description of scalar implicature than Grices original informal accounts, and
hence better meets the desiderata of formal linguistics, the empirical basis for this
description is the intuition that these implicatures are default. To take a concrete
example, the assumption is that an implicature as in (1b) is available from the use
of a sentence like (1a) unless context overrides this (for example by the speaker
saying, In fact all of them did):
1. a. Some of the consultants had a meeting with the director.
b. Not all of the consultants had a meeting with the director.
Note however that any introspective evidence of the defaultness of (1b) depends on
implicit assumptions to the effect that implicatures are available independently of
563
an utterance situation (real or imagined). To the contrary, one could assume that no
implicature becomes available to introspection without a real or imagined utterance situation. In that case, one could observe that it would be dubious to try to ascertain introspectively whether an implicature is available when ones interest is in
whether the implicature in question is available. Although neo-Gricean theory has
been around for thirty years or more, no argument has been put forward to defend
the implicit assumption behind introspective evidence for the defaultness of certain implicatures. In a recent important set of studies, Geurts and Pouscoulous
(2009) have established that introspection on the availability of an implicature introduces a substantial bias in favour of judgements of availability. In their design,
Geurts and Pouscoulous had participants perform two tasks. Their inference task
mirrors the task undertaken by the theorist when she introspects: participants were
asked about whether a putative scalar implicature, like that in (1b), followed from a
simple sentence containing the implicature trigger, like (1a). Using separate items
in a verification task, participants had to verify the truth of a sentence like (1a)
when in fact all consultants had a meeting with the director (contrary to the implicature in (1b)). In a number of different kinds of construction, Geurts and Pouscoulous found that participants were significantly more likely to base their response on
an interpretation that included the implicature in the inference task than the verification task. Their conclusion is that, no matter how consistent and robust an introspected inference may be, that doesnt say anything about the frequency with
which it is drawn in practice (Geurts and Pouscoulous 2009 p.4.18). These results
reinforce the view that it is very important to establish the defaultness of a given
pragmatic effect (like scalar implicature) by methods other than introspection. As
we will see in the next section, much recent work in experimental pragmatics has
focussed on scalar implicatures and is informative on this issue.
4.
Testing for defaultness and preferred readings
One clear statement of a cognitively oriented neo-Gricean theory can be found in

Levinson (2000). Levinson argues that there exist utterance-type meanings that include default implicatures, and that these implicatures are actually generated upon
encountering the appropriate linguistic stimuli. This thesis has been directly and
indirectly tested in experimental work. One set of studies builds on a sentence
verification paradigm established in Noveck (2001). In those studies, participants
are asked to verify statements like Some elephants have trunks. Noveck (2001)
found that a substantial proportion of participants responded negatively, suggesting that those participants had based their judgement on the meaning of the sentence plus the implicature, not all elephants have trunks; while a smaller but still
substantial number responded yes, apparently basing their response on the plain,
logical meaning of the sentence. In subsequent studies, Noveck and Poseda
564
Richard Breheny
(2003) and Bott and Noveck (2004), measured response times to these questions
and found that it was significantly greater when participants responses were based
on the enriched meaning the one that included the implicature. In a follow up
study, Bott and Noveck (2004) also found that a group that was put under time
pressure to respond as quickly as possible gave significantly more logical responses than the group that was not. These results are relevant to the question of
whether generalised implicatures are generated by default. According to Levinson
(2000), where default implicatures are unavailable this is the result of cancellation a process akin to revision when a sentence processor has been garden
pathed. As such, responses based on an interpretation of a stimulus without an implicature ought, other things equal, to manifest a greater cost than those based on
an interpretation with an implicature. As just discussed, in the speeded verification
tasks, Noveck and colleagues found the opposite pattern.
In a more direct measure of on-line access of implicatures, Breheny et al (2006)
tested Levinsons theory employing a reading time methodology. In one experiment, participants read short discourses which provided either a context which
heavily biased a scalar implicature (as in (2)) or one in which the implicature is
generally felt to be unavailable due to lack of relevance (see (3)). The latter kind of
contexts are referred to as implicitly cancelling contexts:
2. Mary asked John where he would accommodate guests for the weekend of the
wedding. John replied that /some of his relatives/ would stay with him. /The
rest/ would book into hotels.
3. Mary asked John why he was cleaning his apartment. John replied that /some of
his relatives/ would stay with him. /The rest/.would book into hotels.
Reading times on the segment the rest confirmed that the and not all implicature
was highly salient in the biasing contexts but not so in the context where the implicature was irrelevant. In line with results in Bott and Noveck (2004), Breheny et
al (2006) found that reading times for the segments around the quantifier some
were significantly longer in the biasing context, like (2). These results again run
contrary to the predictions of Levinsons theory which holds that in the implicit
cancellation context, extra work is required to suppress an activated implicature.
By contrast, all of these results are in line with a more Gricean view whereby implicatures are only available and drawn where warranted by specific features of the
context (as per Sperber and Wilson 1986).
Breheny et al (2006) present a further study to directly compare the predictions
of Levinsons strong interpretation of neo-Gricean theory and a more fundamentally Gricean view. In this study, there was a simple manipulation of the position in
the sentence in which an implicature trigger occurs. For example, participants read
one of the discourses in (4a,b):
565
4. a. Some of the consultants had a meeting with the director. The rest did not
manage to attend.
b. The director had a meeting with some of the consultants. The rest did not
manage to attend.
Citation examples presented out of the blue in theoretical discourse or indeed in
experimental settings are sometimes referred to as neutral contexts (cf. Levinson
2000, Chierchia 2004). There is no agreed definition for neutral in this kind of
discourse but the suggestion seems to be that the examples are interpreted without
any presumptions about the utterance situations in which the items are produced. A
more Gricean view would argue that in as far as the examples are given an interpretation by the reader or experimental participant, there ought to be some imagined utterance situation in which these are produced. That is, they are treated as
utterances (albeit in written form), produced by a speaker with a purpose. Breheny
et al reasoned that if experimental participants treated these items as occurring in
imaginary utterance situations, then other things equal, they would accommodate
context incrementally. Thus it is more likely with (4a) than (4b) that the utterance
is addressing some concern about the consultants, and so in (4a) and not (4b) it is
more likely that participants would derive the implicature. So, by the Gricean account, we would expect to see this effect manifested in response times to the anaphor the rest, which is easier to interpret in contexts where the not all inference
has been drawn. Stronger versions of neo-Gricean theory, including Levinsons,
predict that in these neutral contexts the implicatures are equally likely to be generated. Thus there should be no difference. In line with the Gricean account, Breheny et al found a difference, suggesting that sentence position has an effect on implicature.
In a further test of the strong neo-Gricean view, Grodner and Sedivy (forthcoming) adapted a visual-world design (see Cooper 1974, Tanenhaus et al 1995) from
Sedivy et al (1999) to be discussed below which involves participants hearing
instructions such as in (5):
5. Pick up the tall glass.
Modified referential phrases like the tall glass often presuppose that there is more
than one object that satisfies the nominal description glass. In a series of papers,
Sedivy and colleagues have established that this extra presupposition is effectively
a Gricean implicature. Grodner and Sedivy (forthcoming) tested the claim in Levinson (2000) that this inference only requires the appropriate linguistic context to
be generated. One group of participants in their study were given information that
the speaker producing instructions like (5) was unreliable in certain relevant ways.
Grodner and Sedivy reasoned that if the inference is triggered solely by the linguistic form, even though there was reason to think that the speaker may not have
intended the implicature, in early eye-movements to the visual array, one should
566
Richard Breheny
see evidence that the implicature was accessed before it could be cancelled. By
contrast, if the information about the speaker is allowed to modulate interpretation,
then in contrast to the control group, one should see no effect of the implicature in
early eye-movements. In line with the off-line and reading time studies reviewed
above, Grodner and Sedivy again found that the contextual information about the
speaker does modulate the on-line access of implicatures.
The findings of these studies point toward the conclusion that even apparently
common implicatures are not generated solely on the basis of linguistic context in
on-line comprehension. While these findings tend to disconfirm Levinsons particularly strong version of neo-Gricean theory, other versions of neo-Gricean pragmatics could be adapted as the basis of a cognitive theory of pragmatics, for instance, those found in Horn (1984, 1989), Gazdar (1979) or Sauerland (2004).
These accounts presume the default availability of generalised implicatures, rather
than actual default generation. The difference is important for predictions in empirical studies that seek to measure participants responses based on the inclusion of
an implicature in the interpretation of an utterance. For, just as certain logically derivable entailments of a discourse may not necessarily be generated or accessed in
a given circumstance, so it may be that default implicatures may not necessarily be
derived or generated in comprehension. The parallels with deductive inferences
are important for another reason since it is quite reasonable to view pragmatic description as operating at a normative, rather than cognitive level. The relation between the normative level and the level of individual psychological competence
may be very indirect (see Saul 2002). The more indirect relation between neoGricean pragmatics and cognition is suggested in Horn (2006) and simply manifests a more specific version of the position held by Grice (see Warner 2001, Petrus
2010) and others in the analytic tradition.2 Adopting this stance results in a disengagement of pragmatics from empirical, experimental research of the kind undertaken in contemporary psychology. While it remains a very interesting question
whether it is viable to take either the normative stance or the more neutral a priori
stance towards pragmatics, it is a question which is rarely addressed. For the most
part, as mentioned above, those interested in Gricean pragmatics are not philosophers but linguists who work on the assumption that linguistics is in the business
of explaining individual behaviour via theories of the cognitive processes that
underlie that behaviour. From this perspective, pragmatic theory should be seen as
a theory of individual competence, and pragmatic description ought ultimately to
be specified in such a way that it becomes open to empirical (dis-) confirmation.
Thus an account of pragmatic competence based on neo-Gricean theory according to which implicatures merely become available by default cannot be tested
unless we make some additional assumptions about when available pragmatic inferences are accessed. For instance, neo-Gricean theory could be implemented in a
constraint-based framework (MacDonald and Seidenberg 2006) so that generalised implicatures would be accessed in the presence of certain cues. However, it
567
then falls to the proponent of default implicatures to specify in some more detail
under what condition implicatures are accessed.
It has become clear through the studies in Geurts and Pouscoulous (2009) and
the various studies testing Levinsons strong version of neo-Gricean theory that
there is very little evidence that even the most proto-typical scalar implicatures are
default. Only when reflecting on the availability of these implicatures does one get
a judgement that the implicature is reliably available. In other kinds of task, there is
much evidence in support of an alternative, more fundamentally Gricean, view that
implicatures are derived if and only if warranted by evidence in the utterance situation. So while it is still perhaps an open empirical question whether neo-Gricean
pragmatics offers insights for experimental pragmatics, it is a question which will
require a more thorough engagement with experimental research into utterance interpretation to answer.
With evidence supporting the idea that implicatures are only derived where
warranted by information about the speakers likely intentions in context, we can
ask if there are any theories of pragmatic competence that take a more fundamentally Gricean stance on utterance interpretation. One theory which is notably
Gricean in this respect is proposed in Sperber and Wilson (1986), which argues
that inferring the speakers informative intention is greatly facilitated by the existence of an underlying Relevance mechanism. By contrast, Geurts (2010) is an advocate of a psychologically oriented Gricean position, which adopts Grices original ideas about pragmatic inference based on the Co-operative Principle and
maxims as the starting point of a theory of pragmatic competence. Similarly Clark
(1996) and Tomasello (2008) adopt elements of Grices work in their joint-action
approach to language use.
The viability of more Gricean theories of pragmatic competence has been recently questioned in Sedivy (2003) where it is argued that the rapidity of access to
implicatures on line is difficult to square with the apparent complexity of the reasoning that lies behind these implicatures. Breheny et al (under review a) directly
test whether very rapid access of implicatures in on-line comprehension could be
the result of full-blown pragmatic reasoning on line. The studies involved quantity
implicatures that have no linguistic trigger, as where (6a) below implicates (6b) or
(6c):
6. a. The student put a book on the dusty shelf
b. The student put nothing else on the dusty shelf
c. The student put a book nowhere else
According to both the standard Gricean account and the Relevance Theory account
(see Sperber and Wilson 1995), implicatures such as those in (6b) and (6c) are derived on the basis of assumptions about the purpose (or source of Relevance) of the
utterance and on considering whether the speaker could know alternative, more informative propositions. Breheny et al (under review a) report a series of visual-
568
Richard Breheny
world studies based on the look and listen paradigm first explored in Altmann
and Kamide (1999). In this paradigm, participants are presented with a visual display and listen to spoken discourse. It has been consistently found that eye-movements around the visual scene reflect participants understanding of the text. In
particular, participants looks around the display show that they can anticipate reference to an object on the basis of their understanding of the previous parts of the
utterance but before reference is explicitly made. For instance, in Altmann and Kamide (2007) participants hear about an agent who has drunk or will drink certain things. Where the visual display contains a full glass (of beer) and an empty
wine glass among other items, participants look to the empty glass when the verb is
in perfect tense and at the full glass when the verb is in the future tense. In both
cases, anticipatory looks precede mention of the critical noun, wine or beer.
Breheny et al (under review a) adapt this paradigm to explore whether a quantity
implicature like (6b) is accessed by participants on-line, given the right biasing
context. In a visual display containing a representation of the student putting a
book and another item to one location (a table) and a book and nothing else to the
shelf, participants do anticipate reference to the shelf before they hear the noun
shelf, given the appropriate context that biases the and nothing else quantity
implicature. It is also shown that, at an earlier stage in processing, participants are
sensitive to what the purpose of the utterance is and that they are aware of more informative alternatives. Thus there is evidence that we can and do access quantity
implicatures in incremental interpretation of utterances and that we do so on the
basis of sensitivity to the likely purpose of the utterance and reasoning about the
speakers mental state in the process.
5.
Pragmatic theory meets psycholinguistic theories of the architecture

of on-line comprehension
Independently of any question about the correct theoretical account of the availability and/or derivation of implicatures, we can ask questions about the cognitive
mechanisms that underpin semantic and pragmatic processing, how these mechanisms interact and how the whole architecture of comprehension develops through
infancy and childhood. We will consider developmental research in the next section. In this section, we look at recent research relevant to the architecture of adult
comprehension. As we will see, in this domain, higher level pragmatic theory and
psycholinguistic theory are able to inform one another to their mutual benefit.
As mentioned above, it has been established that we are capable of accessing
quantity implicatures in on-line, incremental interpretation of utterances. While
Breheny et al (under review a) have shown this for quantity implicatures that have
no linguistic trigger, as in (6a-c), Breheny et al (2006) and Breheny et al (under
review b) have demonstrated this for the more familiar case where saying some
569
implicates not all. As we will see in this section, there is already a tradition of
research in psycholinguistics that has looked at another kind of pragmatic implicature, having to do with establishing the reference of a definite expression. While
this research has been relevant to the question of whether any pragmatic implicatures are default, it is equally relevant to architectural issues.
Sedivy et al (1999) have established that, in a visual context where there is
more than one glass, participants already infer that the referent of the tall glass in
(5) is the taller of the glasses before they hear glass, even in the presence of another tall object. Sedivy (2003) reports a study where the scalar adjective tall is
replaced by a material adjective (plastic, wooden etc) and the same effect is
found. Grodner and Sedivy (forthcoming) have established that the effect disappears in an unreliable speaker condition. All of this evidence points to the conclusion that the inference about the referent based on the modifier alone is a
Gricean one: comprehenders assume that speakers will avoid redundancy in their
utterances. When they encounter an adjectival modifier in a definite expression in a
visual context containing two potential referents of the same type, participants can
infer that the likely reason for the presence of the modifier is to distinguish between the two referents.
As participants in these eye-tracking studies seem to be able to access this inference very rapidly (between the onset of tall and the onset of glass), we can
ask how this is possible. As we have seen above, Sedivy and colleagues have ruled
out the theory that rapid access to generalised implicatures only requires formbased cues. As also noted above, one well-motivated alternative hypothesis is that
pragmatic inferences are accessed in on-line interpretation via a constraint-based
processing mechanism. Such a mechanism is interactive, frequency-based and
automatic, meaning that at any point in language processing, the currently favoured hypothesis is based on its likelihood given information from all levels of
processing (including contextual information). While a constraint-based approach
that simply implements neo-Gricean theory would work for common generalised
implicatures, it would have to be extended considerably to work for the ad hoc implicatures studied in Breheny et al (under review a) and in the studies reported in
Huang and Snedeker (2009) and Grodner et al (forthcoming) to be discussed
below. Still, Breheny et al (under review a) argue that a constraint-based architecture could serve to implement more classical Gricean theory, where inferences
about the likely purpose of an utterance, and other aspects of the speakers epistemic state are made automatically and probabilistically.
An alternative to a constraint-based approach to pragmatic processing is one
according to which pragmatic implicatures are accessed only once an interpretation of the linguistic expressions is made, based solely on the compositional semantic meaning of those expressions. This alternative is proposed in Huang and Snedeker (2009), who report on similar visual-world studies to those in Sedivy et al
(1999), but which have different results. In Huang and Snedeker (2009), partici-
570
Richard Breheny
pants are given instructions like in (7) where the visual display has a girl with socks
and a girl with soccer balls:
7. a. Point to the girl that has some of the soccer balls.
b. Point to the girl that has all of the socks.
While participants can disambiguate the referring expression on hearing all in
(7b), they cannot disambiguate the referent for (7a) until after the recognition point
of soccer unless they are able to make a pragmatic inference. Huang and Snedeker found that participants were able to anticipate the correct referent in the
some condition before the recognition point but that, compared to the all condition, they did so at a delay. Huang and Snedeker argue that this delay reflects both
that implicatures are accessed only after the compositional meaning of the sentences is activated and that this is a reflection of the architecture of comprehension
whereby implicatures belong to a different, pragmatic level of representation than
the conventional meaning of a sentence.
The findings in Huang and Snedeker (2009) are difficult to square with those of
Sedivy and colleagues which suggest the pragmatic inference is accessed very
rapidly. So, the question of the time course of accessing implicatures is still open.
This is particularly the case as Grodner et al (2010) found very rapid access to the
pragmatic implicature in a study almost identical to that of Huang and Snedeker.
Breheny et al (under review b) argue that a good Gricean account is available to explain the difference in timing among these studies and present look and listen visual-world studies showing that participants can access scalar implicatures of the
some/not all variety as rapidly as information based on the meaning of all.
Although it is certainly too early to tell for sure, there is mounting evidence that
the results of Gricean pragmatic reasoning can be accessed as rapidly as the results
of deriving information about the interpretation of an utterance from the compositional meaning of the sentence uttered. This evidence suggests that processes that
underpin access to pragmatic implicatures can in some cases operate automatically
or spontaneously. As mentioned above, this kind of evidence is in line with constraint-based approaches to language processing, although other heuristics-based
accounts cannot be ruled out. The evidence does not decide definitively against the
idea that there are two levels of representation, semantic and pragmatic, but only
against the conjunction of that idea and a modular processing architecture according to which all and only Gricean pragmatic processes are set apart from processes
that establish a semantic representation.
Having found evidence that, in some cases, genuine Gricean pragmatic inferences are accessed via what appear to be automatic processing mechanisms, we
can ask to what extent are such inferences operationalised in this way. Gricean
pragmatics explains utterance interpretation as a matter of inferring the speakers
intentions in the utterance situation and much recent research has explored the
question of whether inferences about the speakers intentional state in context can
571
rapidly affect interpretation. Earlier work in this field by Keysar and colleagues
tended to support a strategic egocentrism model according to which first-pass
processes effectively ignored information about the speakers mental state (see
Keysar and Barr 2005). In a typical study, Keysar et al (2000) have participants
play a game where a speaker gives orders to participants to manipulate objects in a
grid and where some objects are obscured from the speaker. In critical items, an instruction like, Pick up the tape is referentially ambiguous if one cannot ignore
items that the speaker is ignorant of, but unambiguous if one only takes the speakers perspective into account. Keysar et al (2000) report severe interference from
the object the speaker cannot see, as predicted if hearers are initially egocentric in
their language processing. Keysar and Barr (2005) argue that the strategic egocentrism strategy is motivated by the apparent difficulty we have in reasoning about
the mental states of others, especially manifest when beliefs diverge from one another (see Apperly and Butterfill 2009).
More recent research into speaker perspective has explored other factors which
may have led to a delay in incorporating speakers perspective. For instance,
Hanna et al (2003) argues that the use of a recording in Keysar et al (2000) is detrimental to the process of monitoring perspective. In Hanna et al (2003), the instructions were given by a confederate who engages the participant in face to face
interaction. The results show much more sensitivity to speaker perspective, virtually without delay. However there is still some small interference. Similar results
are reported in Heller et al (2008) which eliminates a global ambiguity from the
Hanna et al studies that may have encouraged participants to overcome any strategic egocentrism bias. Still better evidence that perspective can be integrated immediately comes from Brown-Schmidt et al (2008) who use a different type of discourse to previous studies. In Keysar et al (2000) and most other perspective
research, the speaker gives orders to a participant. In Brown-Schmidt et als study,
the speaker and hearer are engaged in a joint action and the critical sentences are
questions. That contextual factors such as type of discourse can affect results in
this way supports a middle position between one according to which information
about speakers mental states is completely ignored and one according to which
such information is always automatically incorporated. One version of the middle
position is the Constraint-Based view (Hanna et al 2003, Heller et al 2008, BrownSchmidt et al 2008, among others), according to which cues to mismatched perspective are treated with greater or lesser degrees of certainty and so perspective
information is incorporated only where these cues overcome the very strong constraints provided by the form of the speakers referring expression. One interpretation of this Constraint-Based view is that processing mechanisms are always attempting to integrate the relevant information about the speakers mental state
because interpretation systems are geared toward inferring the speakers intended
meaning. This would be a manifestation of the one-level-of-representation view. A
second interpretation of the C-B view would be to see information from the level of
572
Richard Breheny
speakers meaning interacting to affect a separate semantic level of representation.

Until now, no research has been done to tease these positions apart.
A more modular alternative is offered in Barr (2008) where it is proposed that
while speaker perspective may always be monitored and may immediately affect
anticipation of referents, integration of perspective information with information
from the linguistic input is not automatic. Using a similar design to other studies
mentioned above but a new method of analysis which discriminates anticipation
from integration effects, Barr finds that bottom-up, linguistic information affects
early referential hypotheses independently of whether participants manifest awareness of contrary evidence of speaker perspective. As these studies used a recording
of a speaker and lacked genuine speaker/hearer interactions, and as they involved a
director giving orders, it remains to be seen whether the anticipation-without-integration effect is maintained in more genuinely interactive scenarios across different types of discourse.
Supplementary to research on how a speakers ignorance may affect reference
assignment, recent on-line research has studied the time course of so-called entrainment effects, long documented in off-line interactive research (see Garrod and
Anderson 1987, Brennan and Clark 1996). Entrainment is the phenomenon where
interlocutors fix on a single description for an object or event throughout a conversation. So the same object may be referred to as a metal pipe or silver tube, but once
such a mode of reference is established, subsequent reference among interlocutors
tends to maintain this perspective on the object. The analysis of entrainment
stemming from Brennan and Clark (1996) is that it is the result of a conceptual
pact among interlocutors who (tacitly) agree on a mode of presentation of a referent for the purposes of a conversation. Off-line research has shown that such entrainment effects can last for days (Markman and Makin 1998) and that speakers
do adjust their descriptions according to the knowledge of their audience (Brennan
and Clark 1996).
The off-line research suggests strongly that entrainment is a partner-specific effect but it is still open whether it results from an intentional act by interlocutors, as
Brennan and Clark maintain, or whether some other factors are at play. In an early
on-line study, Metzing and Brennan (2003) had participants play a game with an
experimenter moving objects around a grid. The experimenter (speaker) would fix
on a description for a certain object (silver tube) for three moves and then the experimental conditions manipulate both whether the pact was maintained and who
the speaker was. In measures of visual responses around the grid, it was found that
when the description was changed, significantly more interference was created
with the same speaker than with a new one. However, when descriptions were
maintained, there was no effect of speaker. That is, participants fixed on the correct
referent as rapidly with a new speaker as with the old. While the results of the different description condition support the conceptual pact account, it would be expected that there should be an effect of speaker both in the case where the pact is
573
broken and when it is maintained. Kronmuller and Barr (2007) offer an account of
Metzing and Brennans results according to which the partner-specific effect is the
combined result of an automatic pre-emption of other descriptions due to a link
made in memory between the referent and the description, and a somewhat delayed
application of inference based on pragmatic principles. This account explains the
lack of difference in same-description condition of Metzing and Brennans study
and predicts that the interference effect for the same-speaker in the different-description condition occurs at a delay. This is the result Kronmuller and Barr found
in a visual world study similar to that of Metzing and Brennan. Their conclusion is
that there are two mechanisms at play in the maintenance of descriptions in conversation. However, it is somewhat mysterious why an automatic pre-emption
mechanism would develop, except as a result of something like Gricean pragmatic
principles. So even if Kronmuller and Barr are right, it may still be the case that entrainment is the result of a heuristic that serves joint interaction in discourse. On
the other hand, Brown-Schmidt (2009) argues that the lack of interactivity in Kronmuller and Barrs studies could explain the weakness of their entrainment effects.
Brown-Schmidt employed a slightly different design to the Metzing and Brennan
(2003) study, employing critical items that would encourage entrainment effects
and using more participants. She found both an effect of speaker in the same- and
different-speaker condition. In a second study, she employed the same items but
with a recording rather than a live confederate and found that the effects of speaker
were diminished greatly. The conclusion seems to be that entrainment effects that
are the product of genuine real-time interactions among speakers can occur spontaneously or automatically. Whether, in addition, there exist pre-emption effects,
and if so why they arise, are still open questions.
To sum up this section, we have considered a sample of psycholinguistic research which sheds light on the mechanisms that underpin access to pragmatic inferences. We have seen that there is no strong support for the idea that genuinely
Gricean pragmatic inference processes are somehow set apart from processes that
establish a separate semantic representation. The idea that Gricean inferences
could not automatically or spontaneously be derived because of the complexity of
the mental state reasoning involved also needs to be reconsidered, for two reasons.
Firstly, across a number of different domains, it has been demonstrated that genuine intentional pragmatic inferences are accessed as rapidly as any other information relevant to an interpretation. Secondly, the idea that inferences about
mental states are per se especially difficult compared to any other kind of non-demonstrative inference ought to be re-examined. Indeed, as we will see in the next
section, there is good reason t to think that in the domain of language use and communication, humans develop a keen ability to infer the contents of goal states from
the earliest possible stage.
Given the rapidity of Gricean inference in some cases and the relative open-endedness of their confirmation, the challenge for both psycholinguistics and prag-
574
Richard Breheny
matic theory is to explore the mechanisms that deliver them. We have ruled out the
idea that we only get rapid pragmatic inferences when these can be associated with
particular forms (like some or adjectival modification). In fact we have seen evidence that any ad hoc quantity implicature can be accessed very rapidly. Thus, if
we are to adopt a Constraint-Based model or any other frequency-based model for
pragmatic inference, we need to specify which elements of an utterance situation
are being automatically monitored by hearers. Breheny et al (under review a) suggest that one important place to start is with expectations of the conversational purpose or, more generally, the source of relevance for an utterance. In the case of
quantity implicatures, this could be coupled with a mechanism that monitors alternatives. One way to explore the latter mechanism is via the more general mechanisms that anticipate the completion of an utterance. These are evident in the visualworld research mentioned above (see Altmann and Kamide 1999, Barr 2008) but
much more so in ERP (Event Related Potential) research focussed on the N400 effect.3
Generally speaking, it seems clear that future psycholinguistics research into
the mechanisms that underpin utterance interpretation would benefit from insights
of theoretical pragmatics. At the same time, pragmatic theory (at least the cognitively oriented variety) can greatly benefit from research into on-line processing.
In particular, one important question for recent philosophical and linguistic research is whether there is justification for positing a separate semantic level of representation of the truth conditions of an utterance in context which is independent
of the level at which the speakers intentions are represented. If we treat this theoretical question as also a cognitive question (as do Borg 2004, Carston 2002), then
the type of research reviewed above could be informative to the question. However, while much evidence supports a one-level view, the two-level view cannot be
ruled out given the current state of research.
6.
The development of Gricean pragmatics in language use
Any good theory of pragmatic competence ought not just inform the modelling of
on-line processing in adults but should also be able to explain how the relevant
mechanisms involved develop. This section will review research into the application of intentional pragmatic principles in early language use. We will see that,
while our understanding of child pragmatics has been transformed in the recent
past, the focus has been on getting certain facts right about what, if any, limitations
children have with regards to common pragmatic inferences. The emerging picture
is of children being capable of deriving any pragmatic inferences that adults derive
as early as anyone can ascertain. Where there has been apparent divergence between child responses in experimental studies and adults, this has been partly due
to inadequate methodology. The interesting theoretical question regarding devel-
575
opmental research is how children so rapidly achieve full pragmatic competence

and what lies behind any genuine differences between children and adults regarding the implementation of pragmatic abilities in conversation.
Perhaps in contrast to research into adult comprehension, child language research has, for a long time, focussed its attention on childrens intentional pragmatic abilities. The watchword for most work on childrens syntactic and semantic
abilities has been pragmatic plausibility. For instance, when asking a child whether
she finds a statement about an acted-out situation true or false, a good test for
whether a child is capable of ruling out ungrammatical construals of a string involves scenarios where grammatical construals are false but an ungrammatical
construal is true. For example, in Crain and McKee (1985) experimental scenarios
made He ate pizza and Ninja Turtle danced true only on the ungrammatical construal according to which the turtle ate pizza and danced, while the grammatical
construal is false. The methodological principle of pragmatic plausibility in child
studies is to make the grammatical construal relevant. In this case, by having another character consider eating pizza. The theory is that children in these studies
operate with a principle of charity whereby they are liable to ignore divergences
from proper usage if not doing so does not lead to a plausible construal. Whether
children should operate with such a principle more than adults is something of an
open question. Indeed, as we will see below, the real question now for child pragmatic research is why childrens responses often diverge from adults. Any explanation based on a principle of charity would have to be motivated more fully than
hitherto, while there may be alternative explanations for such divergences.
While early work in syntax and semantics using Truth Value Judgement Tasks
(TVJT) tended to assume some pragmatic competence on the part of toddlers, early
research directly testing childrens access to common pragmatic inferences suggested that pragmatic competence might not be fully developed in some respects.
Here we will focus on childrens understanding of scalar quantity implicatures and
the pragmatic dimension of Principle B (cf. below) of binding theory.
In a landmark paper, Noveck (2001) reports studies using a sentence verification task involving quantificational expressions like All elephants have trunks
or Some elephants have trunks. As mentioned in section 4, adult participants in
these studies tend to favour a response to the some items that is based on an interpretation of the utterance that includes the quantity implicature Not all elephants have trunks. Although numbers vary from study to study, the proportion of
responses that incorporate the implicature rarely falls below two thirds. Noveck
(2001) reports that childrens response rates do not approach that of adults until
7 years in some cases and later with more complex items. Rather, children tend to
favour the logical response, saying that Some elephants have trunks is true. In a
follow-up study, Papafragou and Musolino (2003) used act-out scenarios and a
truth-value judgement task to see if the task demands of sentence verification
(requiring access to long-tem memory) might have been a factor. Indeed, Papafra-
576
Richard Breheny
gou and Musolino (2003) found younger children were performing like adults in
rejecting statements like Some of the horses jumped the fence but still five year
old children showed little sign of having incorporated the implicature into their
understanding of such statements.
In an attempt to determine whether it is the inherent difficulty of the Gricean inference that is at issue, Papafragou and Musolino ran a felicity judgement task
which required the children only to evaluate whether the puppet had described the
scenario well. They found significantly more children were able to evaluate an
underinformative utterance appropriately. The conclusion that could be drawn
from this is that, while children do have expectations of informativity appropriate
to drawing quantity implicatures, something about the TVJT causes the children to
diverge from the adult response pattern. This conclusion is bolstered by the results
of another kind of felicity judgement task (Guasti et al 2005), where children are
given a forced choice between two images, one that depicts a scenario where all of
the characters perform the action and one where some but not all do. In this kind of
study, five year-old children robustly choose the scenario consistent with expectations of informativity. Katsos and Bishop (under review) find that five year old
children succeed at the latter felicity judgement task where the implicature is not
associated with any lexical item, but ad hoc. Kastsos and Bishop also find that
childrens own descriptions of scenarios always respect Gricean expectations of informativeness.
Pouscoulous et al (2007) explore the idea that children do not readily incorporate the implicature into their understanding of the experimental items due to the
task demands resulting from the combination of the metalinguistic nature of the
TVJT and the extra effort required to derive quantity implicatures. In a series of experiments that were aimed at successively lightening task demands, Pouscoulous
et al did find a marked difference in performance. In particular, by changing the
task from truth evaluation to an act-out task, they found that even four-year-old
childrens rate of with-implicature response was above chance. They also found
that childrens with-implicature response rates increased when the items involved
the apparently less complex quantifier quelques rather than the more complex
certains.
One important conclusion from these child studies looking at quantity implicature seems to be that there really is no principled limitation on young childrens
ability to derive quantity implicatures. They seem clearly aware of expectations of
informativeness, they produce appropriately informative utterances and they can
interpret utterances with quantity implicatures. So we could say that the basics of
even four-year-old childrens pragmatic competence is in place. However, even in
the act-out task of Pouscoulous et al (2007) there are still differences across age
groups which look like a trend toward typical adult responses. Pouscoulous et al
(2007), following Noveck (2001), propose that this is due to the cost of the extra inference involved in the enriched interpretation. This view is supported by the adult
577
studies reported in Bott and Noveck (2004) where it is found that adults put under
time pressure tend to respond more often without the implicature. When considering the difference between performance in act out tasks and TVJTs, Katsos and
Bishop (under review) argue that processing cost may not be the whole story but
that children may be more inclined to assent to underinformative utterances in
TVJTs due to the operation of a principle of charity. In light of the fact that even
four-year-old children can derive quantity implicatures, it is clear that they must operate with such a principle, as do adults when faced with a statement that clearly
could be construed as true or false. However, the puzzle with the developmental
trend found in these tasks lies more with adult responses. Note that it is not as
though the adults tend to give a correct response while children deviate from this.
In some respects, the childs logical response is more correct than that of the
adults. If it is true that the principle of charity operates both with adults and
children, then the question really is why adults tend not to see the alternative interpretation. We may be able to find the answer to this question from the adult studies
reviewed above. There we found very rapid access to quantity implicatures in adults
and speculated that adult access to this type of implicature rests on a combination of
automatic, frequency based heuristics. So it may be that what develop between
childhood and adulthood are the heuristic mechanisms that both lighten processing
load and make judgements more certain when it comes to valuation tasks.
The tale of childrens abilities with quantity implicature has played itself out
with only minor modifications when it comes to the development of the so-called
Principle B effects studied by syntacticians. According to the favoured theory of
these effects, due to Reinhart (see Reinhart 1983, 2006), informants judge that the
noun phrases in (8a) cannot be co-referential due to a pragmatic parsimony principle similar to the one we might invoke to account for the contrastive implicatures
studied by Sedivy and colleagues:
8. a. Mama Bear washed her.
b. Mama Bear washed herself.
The idea is that if the speaker wished to express the proposition that Mama Bear
washed Mama Bear, then she would have used the reflexive form as in (8b), as it is
unambiguous. Many child studies of the Principle B phenomena have shown that
children abide by the syntactic principle restricting genuine binding but seem to
fail to appreciate that (8a) should not normally be used to mean (8b) for pragmatic
reasons (see Chien and Wexler 1990, Grodzinsky and Reinhart 1993, Thornton and
Wexler 1999). However, Conroy et al (2009) have challenged certain methodological aspects of these studies and show that 4 year old children are capable of rejecting the pragmatically inappropriate interpretation of (8a) given the right kind of
contextual support. Conroy et al conclude, contra Chien and Wexler (1990) that
children do have the requisite pragmatic abilities to rule out co-reference. But as
with the implicature studies, they do recognise that there is a kind of developmen-
578
Richard Breheny
tal trend. Conroy et al (2009) follow Reinhart (2006) in proposing that the divergence in child responses from the adult norm may have to do with the processing
resources involved. But, again, it may be that the full developmental story turns on
the development of the heuristics which make adults confidently reject the infelicitous utterance.
To sum up this section, we have seen two examples of research on pragmatic
development where, after initial indications to the contrary, it seems children have
full Gricean pragmatic competence as early as anyone can tell. We have also see
that the methods used in child studies have thrown up some intriguing questions as
to why childrens behaviour in some studies deviate from that of adults. Finding
out what develops and how, may shed light on the steady-state adult competence,
while research on the mechanisms that underlie adult processing may in turn shed
light on child pragmatic development.
7.
The ontogenesis of pragmatics Studies in pre-linguistic

communication
One conclusion we may draw from studies of toddlers pragmatic abilities is that as
soon as they have access to the syntactic binding principles, they can rule out pronominal co-reference on pragmatic grounds, and as soon as they grasp the meaning
of indefinite quantifiers like some, they can derive quantity implicatures. We may
then wonder how these pragmatic abilities are ever-present through the development of linguistic abilities. The answer coming from pre-linguistic research seems
to be that it is because infants are fully competent Gricean communicators before
they acquire language.
In a series of interactive pointing production and comprehension tasks, Tomasello and colleagues (see Tomasello 2008, Tomasello et al 2007) have shown that
children treat communicative gestures as involving the specifically reflexive intention first brought to light in Grices work on speaker meaning (see Grice 1957).
That is, not only do infants see communication as involving an intention that another person obtain information but as doing so on the basis of recognising the intention to inform. This overtness condition has been analysed in a number of ways
by philosophers and cognitive scientists but in practice it requires certain cues on
the part of the communicator, such as eye-contact or interactive, contingent behaviour, and certain cues on the part of audience to signal uptake. In elicited production tasks, Liszkowski et al. (2004) elicited infant points to an interesting event
and, in one condition, the adult experimenter simply attended to the event in response, rather than also acknowledge interest to the child. In a baseline condition,
the experimenter both attended to the event and acknowledged uptake with the
child. In the no-uptake condition, as compared to the baseline, children repeated
the communicative act until the experimenter responded appropriately.
579
In a comprehension experiment, Behne et al (2005) engaged infants in a game

where a prize is hidden in one of two containers. In one experimental condition, the
experimenter engages the infant with ostensive cues (particularly, eye contact) and
points to a container. Behne et al found that at 14 months, infants are above chance
in choosing the indicated container, while at 18 months they are virtually at ceiling.
In another condition, the experimenter forms the same hand gesture, fashioned in
the shape of a pointing, but does not engage the child in eye contact or use other ostensive cues. In this condition neither age group favoured the indicated container.
In more recent research, evidence has accumulated that infants understand
communicative acts not simply in terms of what is relevant to them but in terms of
what the communicator would think is relevant (see Liebal et al 2009, and particularly, Southgate et al 2009).
The overall picture is that by the time children approach the one- or two-word
phase, they are already employing sophisticated pragmatic abilities. It now seems
clear that these abilities appear to develop throughout infancy, from around
4 months of age. For instance, Farroni et al (2003) show that 5 month-old infants
tend to follow shifts in eye direction only if the shift is preceded by a period of eye
contact with the infant. Similar research, summarised in Csibra and Gergely (2009)
and Csibra (2010) points to the conclusion that the specifically human mode of
communication involving the recognition of complex intentions develops along
pre-determined pathways prior to the development of fuller social cognitive abilities at the end of the first year. Thus human pragmatic competence may have an innate basis and involve specific neural structures (see Grossman et al 2008). How
this competence relates to more general human social cognitive abilities and to
general linguistic abilities remain intriguing open questions (see Tomasello 2008,
Breheny 2006, Csibra 2010).
8.
Summary
In the past decade, there has been a significant upsurge in experimental research in
the domain of intentional pragmatics, that is, research into our understanding of
speakers meaning based on inferences about intentions in context. This upsurge
has come from different sources, more or less independently. On the one hand, Experimental Pragmatics has attempted to apply experimental methodology to illuminate theoretical debates among linguists.4 On the other hand, psycholinguists
have become increasingly interested in how information about speakers intentions
inform rapid on-line interpretation processes. Meanwhile, there has been a revolution in attitudes towards toddlers linguistic pragmatic abilities; and finally research originating mostly in the infant social cognitive domain has raised the exciting possibility that intentional pragmatic abilities are part of the basic human
makeup.
580
Richard Breheny
Taking this research together, we are by far better informed about intentional
pragmatic abilities now than ten years ago. But there are still very many empirical
questions that remain open. This paper has offered a small sample of recent work
and highlighted some of those open questions.
Notes
* This work was carried out with the support of a grant from the UK Arts and Humanities
Research Council (Ref: AH/E002358/1) and from the Centre for the Study of Mind in Nature (Oslo), Linguistic Agency project.
1. For a general overview of (types of) implicature cf. Huang this volume.
2. Although it is argued in Saul (2002) that Grices account of conversational implicature
should be understood to be a normative account, there is very little textual evidence for
this and far more evidence that Grice would have wished to remain neutral on this issue
(see Petrus 2010).
3. See van Berkum in press for an overview; N400 refers to a negativity that appears in an
ERP reading 400 ms after the offset of a stimulus. It is widely thought of as an indicator of
the processing of meaning. Larger N400 suggests extra semantic processing work (as
when there is semantic anomaly).
4. Here we have reviewed work focussed on quantity implicatures, but there is other work
on metaphor (see Glucksberg 2003, Rubio-Fernandez 2007), presupposition (see
Schwarz 2007, Chemla and Bott 2010) and other areas of pragmatics (see Noveck and
Sperber 2004 for an early collection).
References
1999
Incremental interpretation at verbs: restricting the domain of subsequent reference. Cognition 73: 247264.
2007
The real-time mediation of visual attention by language and world knowledge:
Linking anticipatory (and other) eye movements to linguistic processing. Journal of Memory and Language 57: 502518.
Apperly, Ian A. and Stephen A. Butterfill
2009
Do humans have two systems to track beliefs and belief-like states? Psychological Review 116(4): 953970.
Atlas, Jay and Stephen C. Levinson
1981
It-clefts, informativeness and logical form: radical pragmatics (revised standard version). In: Peter Cole (ed.), Radical pragmatics, 161. New York: Academic Press.
Barr, Dale
2008
Pragmatic expectations and linguistic evidence: Listeners anticipate but do not
integrate common ground. Cognition 109: 1840.
581
Behne, Tanya, Malinda Carpenter and Michael Tomasello

2005
One-year-olds comprehend the communicative intentions behind gestures in a
hiding game. Developmental Science 8: 492499.
Berkum, Jos J.A. van
in press The neuropragmatics of simple utterance comprehension: An ERP review.
To appear in: Ulrich Sauerland and Kazuko Yatsushiro (eds.), Semantic and
pragmatics: From experiment to theory. New York: Palgrave.
Borg, Emma
2004
Bott, Lewis and Ira A. Noveck
2004
Some Utterances are Underinformative: The Onset and Time Course of Scalar
Inferences. Journal of Memory and Language 51(3): 437457.
Breheny, Richard
2006
Communication and folk psychology. Mind and Language 21(1): 74107.
Breheny, Richard, Napoleon Katsos and John Williams
2006
Are Scalar Implicatures Generated by Default? Cognition 100: 434463.
under review, a How quantity implicatures are accessed in incremental utterance interpretation: Evidence for full-blown Gricean inferencing.
under review, b On the timecourse of access of conversational implicatures in incremental interpretation.
1996
Conceptual pacts and lexical choice in conversation. Journal of Experimental
Psychology: Learning, Memory, and Cognition 22: 14821493.
Brown-Schmidt, Susan
2009
Partner-specific interpretation of maintained referential precedents during interactive dialog. Journal of Memory and Language 61: 171190.
Brown-Schmidt, Susan, Christine Gunlogson and Michael K. Tanenhaus
2008
Addressees distinguish shared from private information when interpreting
questions during interactive conversation. Cognition 107: 11221134.
Carston, Robyn
2002
Thoughts and utterances. Oxford: Blackwells.
Chemla, Emmanuel and Lewis Bott
2010
Processing presuppositions: dynamic semantics vs pragmatic enrichment. Ms
Institut Jean Nicod (http://www.emmanuel.chemla.free.fr/Material/ChemlaBott-ProcPresProj.pdf)
Chien, Yu-Chin, and Kenneth Wexler
1990
Childrens knowledge of locality conditions in binding as evidence for the
modularity of syntax and pragmatics. Language Acquisition 1: 225295.
Chierchia, Gennaro
2004
Scalar implicatures, polarity phenomena and the syntax/pragmatics interface.
In: Adriana Belletti (ed.), Structures and beyond, 39103. Oxford: Oxford
University Press.
Clark, Herbert H.
1996
582
Richard Breheny
Conroy, Anastasia, Eri Takahashi, Jeff Lidz and Colin Phillips

2009
Equal Treatment for All Antecedents: How Children Succeed with Principle B.
Linguistic Inquiry 45: 446486.
Cooper, Roger M.
1974
The control of eye fixation by the meaning of spoken language: A new methodology for the real-time investigation of speech perception, memory, and language processing. Cognitive Psychology 6: 84107.
Crain, Stephen and Cecile McKee
1985
The acquisition of structural restrictions on anaphora. In: Proceedings of
NELS 16. Ed. by Stephen Berman, Jaewoong Choe and Joyce McDonough,
94110. Amherst: University of Massachusetts, GLSA.
Csibra, Gergely
2010
Recognizing communicative intentions in infancy. Mind and Language 25:
141168.
Csibra, Gergely and Gyorgy Gergely
2009
Natural Pedagogy. Trends in Cognitive Science 13: 148153.
Farroni, Teresa, Eileen M. Mansfield, Carlo Lai and Mark H. Johnson
2003
Infants perceiving and acting on the eyes: Tests of an evolutionary hypothesis.
Journal of Experimental Child Psychology 85: 199212.
Garrod, Simon and Anthony Anderson
1987
Saying what you mean in dialog: A study in conceptual and semantic co-ordination. Cognition 27: 181218.
Gazdar, Gerald
1979
Geurts, Bart
2010
Quantity Implicature. Cambridge: Cambridge University Press.
Geurts, Bart and Nausicaa Pouscoulous
2009
No scalar inferences under embedding. Semantics and Pragmatics 2(4): 134.
Glucksberg, Sam
2003
The psycholinguistics of metaphor. Trends in Cognitive Sciences 7: 926.
Graesser, Arthur C., Morton Ann Gernsbacher Susan R. Goldman
2003
Handbook of discourse processes. Mahwah, NJ: Lawrence Erlbaum.
Grice, H. Paul
1957
Meaning. Philosophical review 66: 377388. Reprinted in Grice (1989,
213223).
Grice, H. Paul
1989
Studies in the way of words. Cambridge, MA: Harvard University Press.
Grice, H. Paul
2001
Aspects of Reason. Oxford: Oxford University Press.
Grodner, Daniel and Julie Sedivy
in press The effects of speaker-specific information on pragmatic inferences. In: Neal J.
Pearlmutter and Edward Gibson (eds.), The Processing and Acquisition of Reference, Cambridge, MA: MIT Press.
Grodner, Daniel J., Natalie M. Klein, Kathleen M. Carbary and Michael K. Tanenhaus
2010
Some, and possibly all, scalar inferences are not delayed: Evidence for immediate pragmatic enrichment. Cognition.
583
Grodzinsky, Yosef and Tanya Reinhart

1993
The innateness of binding and coreference. Linguistic Inquiry 24: 69101.
Grossmann, Tobias, Mark H. Johnson, Sarah Lloyd-Fox, Anna Blasi, Fanni Deligianni,
Clare Elwell and Gergely Csibra
2008
Early cortical specialization for face-to-face communication in human infants.
Proceedings of the Royal Society, London B, 275: 28032811.
Guasti, Maria Teresa, Gennaro Chierchia, Stephen Crain, Francesca Foppolo, Andrea Gualmini and Luisa Meroni
2005
Why children and adults sometimes (but not always) compute implicatures.
Language and Cognitive Processes 20(5): 667696.
Hanna, Joy, Michael K. Tanenhaus and John C. Trueswell
2003
The effects of common ground and perspective on domains of referential interpretation. Journal of Memory and Language 49: 4361.
Heller Daphna, Daniel Grodner and Michael K. Tanenhaus
2008
The role of perspective in identifying domains of reference. Cognition 108:
831836.
Horn, Laurence
1984
A new taxonomy for pragmatic inference: Q-based and R-based implicature.
In: Deborah Schiffrin (ed.), Meaning, Form and Use in Context, 1142.
(GURT 84). Washington: Georgetown University Press.
Horn, Laurence
1989
A Natural History of Negation. Chicago, IL: University of Chicago Press.
Horn, Laurence
2006
The border wars: a neo-Gricean perspective. In: Klaus von Heusinger and Ken
Turner (eds.), Where Semantics Meets Pragmatics, 2148. Amsterdam: Elsevier.
Huang, Yi Ting, and Jesse Snedeker
2009
On-line Interpretation of Scalar Quantifiers: Insight into the Semantic-Pragmatics Interface. Cognitive Psychology 58: 376415.
Katsos, Napoleon and Dorothy Bishop
under review The development of informativeness from the speakers and the comprehenders perspective.
Keysar, Boaz, Dale J. Barr, Jennifer A. Balin and Jason S. Brauner
2000
Taking perspective in conversation: The role of mutual knowledge in comprehension. Psychological Sciences 11: 3238.
Keysar, Boaz, and Dale J. Barr
2005
Coordination of action and belief in communication. In: John C. Trueswell and
Michael K. Tanenhaus (eds.), Approaches to Studying World Situated Language Use: Bridging the Language-as-Product and Language-as-Action,
Kronmuller, Emundo and Dale J. Barr
2007
Perspective-free pragmatics: Broken precedents and the recovery-from-preemption hypothesis. Journal of Memory and Language 56: 436455.
Levinson, Stephen
2000
Presumptive Meanings. Cambridge, MA: MIT Press.
Liebal, Kristin, Tanya Behne, Malinda Carpenter and Michael Tomasello
2009
Infants use shared experience to interpret pointing gestures. Developmental
Science 12(2): 264271.
584
Richard Breheny
Liszkowski, Ulf, Malinda Carpenter, Anne Henning, Tricia Striano and Michael Tomasello
2004
12-month-olds point to share attention and interest. Developmental Science 7:
297307.
MacDonald, Maryellen C. and Mark S. Seidenberg
2006
Constraint satisfaction accounts of lexical and sentence comprehension. In:
Matthew J. Traxler. and Morton Ann Gernsbacher (eds.), Handbook of Psycholinguistics. 2nd ed., 581611. London: Elsevier.
1992
Inference during reading. Psychological Review 99: 440466.
Markman, Arthur B. and Valerie S. Makin
1998
Referential communication and category acquisition. Journal of Experimental
Psychology: General 127: 331354.
Metzing, Charles, and Susan E. Brennan
2003
When conceptual pacts are broken: partner-specific effects on the comprehension of referring expressions. Journal of Memory and Language 49: 201
213.
Noveck, Ira A.
2001
When children are more logical than adults: Investigations of scalar implicature. Cognition 78/2: 165188.
Noveck, Ira, and Andres Posada
2003
Characterising the time course of an implicature. Brain and Language 85:
203210.
2004
Experimental Pragmatics. New York: Palgrave.
Papafragou, Anna and Julien Musolino
2003
Scalar implicatures: Experiments at the Semantics-Pragmatics interface. Cognition 86: 253282.
Petrus, Klaus
2010
Introduction: Paul Grice, Philosopher of Language, but more than that.
In: Klaus Petrus (ed.), Meaning and Analysis: New Essays on H. Paul Grice,
130. New York: Palgrave.
Pouscoulous, Nausicaa, Ira A. Noveck, Guy Politzer, and Anne Bastide
2007
A developmental investigation of processing costs in implicature production.
Language acquisition 14: 347376.
Reinhart, Tanya
1983
Anaphora and semantic interpretation. London: Croom Helm.
Reinhart, Tanya
2006
Interface strategies. Cambridge, MA: MIT Press.
Rubio-Fernandez, Paula
2007
Suppression in Metaphor Interpretation: Differences between Meaning Selection and Meaning Construction. Journal of Semantics 24: 345371.
Saul, Jennifer
2002
Speaker meaning, what is said and what is implicated. Nous 36: 228248.
Sauerland, Ulrich
2004
367391.
585
Sedivy, Julie C.
2003
Pragmatic versus form-based accounts of referential contrast: Evidence for effects of informativity expectations. Journal of Psycholinguistic Research 32:
323.
Sedivy, Julie C., Michael K. Tanenhaus, Craig G. Chambers and Gregory N. Carlson
1999
Achieving incremental semantic interpretation through contextual representation. Cognition 71: 109147.
Schwarz, Florian
2007
Processing Presupposed Content. Journal of Semantics 24(4): 373416.
Southgate, Victoria, Coralie Chevallier and Gergely Csibra
2009
Sensitivity to communicative relevance tells young children what to imitate.
Developmental Science 12(6): 10131019.
1986
1995
Postface to 2nd Edition. In: Dan Sperber and Deidre Wilson, Relevance: Communication and cognition, pl. add+ rede. 2nd ed. Oxford: Blackwell.
Tanenhaus, Michael K., Michael J. Spivey-Knowlton, Kathleen M. Eberhard and Julie C.
Sedivy
1995
Integration of visual and linguistic information during spoken language comprehension. Science 268: 16321634.
Thornton, Rosalind and Kenneth Wexler
1999
Principle B, VP-ellipsis, and interpretation in child grammar. Cambridge,
MA: MIT Press.
Tomasello, Michael
2008
Origins of Human Communication. Cambridge, MA: MIT Press.
Tomasello, Michael, Malinda Carpenter and Ulf Lizskowski
2007
A new look at infant pointing. Child Development 78: 70522.
Warner, Richard
2001
Introduction: Grice on reasons and rationality. In: H. Paul Grice (2001),
viixxxviii.
21.
Corpus-based pragmatics I: qualitative studies1

Gisle Andersen
1.
Contextualising corpus-based pragmatics
This paper concerns the use of computational methods for collection and analysis
of language data as a basis for research in pragmatics. The aim is to describe advantages and problems associated with the use of corpora, and to provide both a
birds-eye view of the field as well as an in-depth account of a limited set of topics
of particular salience. A main claim to be argued here, with reference to several
papers, is that, as a method for empirical research, corpus-based pragmatic analysis
is essentially inductive and interpretative (Mosegaard Hansen 1998: 214) and
can never be performed without the reliance on the analysts intuition (cf. especially section 3). Corpus-based studies are conceptually related to other ways of
doing empirical pragmatics, such as anthropological or ethnological field work, sociolinguistic interviews and the use of experimental techniques and elicitation tests
(Chafe 1992). The insights offered by the studies reviewed in the current survey
have emerged via meticulous analysis of corpus data, as opposed to other empirical
methods. The studies discussed here are further distinguished from those discussed
in Christoph Rhlemanns contribution to the current volume, which focuses on
quantitative studies. However, corpus-based studies are rarely exclusively qualitative or quantitative but commonly describe pragmatic functions as well as quantifiable distributional aspects, involving for instance comparisons between users
and user groups within or across corpora. The current study, then, describes that
body of pragmatics literature that has its main focus on functional aspects but focuses less on measurable distributional differences.
The term discourse linguistics has been used as an umbrella term for discourse analysis, discourse studies, text linguistics, pragmatics, conversation analysis and other related approaches to the study of discourse (Virtanen 2009: 49), a
list of topics which is broadly consistent with the topics to be reviewed in this
chapter. The application of corpus-linguistic method in this knowledge field has
been relatively limited, especially compared to its application in the study of more
formal topics such as syntax, morphology and lexis (cf. McEnery and Wilson
2001; Kennedy 1998). Explorations of corpora have nevertheless been fruitful in
pragmatics, as evidenced by a series of studies which investigate how interaction is
sequentially and interactionally organised, how the attitudes and perspectives of
speakers and hearers are manifested in spoken discourse, how texts are coherently
structured beyond the sentence level, etc. Among the topics studied are turn-taking
phenomena, pauses and repairs, discourse markers, hedging and vague language,
588
Gisle Andersen
small talk, speaker attitudes and so on. (cf. section 2). Despite the variety of topics,
there are many aspects of pragmatics that have not been subjected to corpus-based
investigations to the extent they could have been, and I shall argue in favour of an
extended application of corpora in pragmatics, to capture a wider range of phenomena within the cognitive, social and cultural science of language and communication (Verschueren 2009: 1).
Corpus linguistics and pragmatics can be seen as two related, but historically
distinct research traditions. I endorse the generally accepted view expressed by
corpus linguists such as Leech (1992), Kennedy (1998), McEnery and Wilson
(2001) and Virtanen (2009), that pragmatics and corpus linguistics are ontologically different academic pursuits, in that the former is a branch of linguistics albeit many-faceted and complex, while the latter is not. Rather, corpus linguistics is
regarded as a certain way of performing empirical studies in linguistics, a research approach (Biber 1988) and a methodological basis for doing linguistic
research (Leech 1992: 105; cf. also Kennedy 1998: 7). Empirical pragmatics crucially depends on authentic language data, and corpus linguistics is one methodological approach which may be used for pragmatics research. Like other kinds of
empirical language data, corpora are documentations of the choices made by language users and surface manifestations of the underlying communicative competence of the speakers whose language the corpus represents (de Beaugrande 1999;
cf. also section 3). In the words of Chafe (1992:88), corpora provide a particularly
useful means of access to the language-related capabilities of the mind.
A notable recent development in corpus linguistics has been the increased
focus on corpus-driven research (as opposed to corpus-based research; TogniniBonelli 2001: 8487; Sinclair 2004), which differs from the standard practice of
linguistics in that it makes minimal a priori assumptions regarding the linguistic
constructs that should be employed for the analysis (Biber 2009: 276). In corpusbased research, the researcher approaches the data equipped with a set assumptions
as to its linguistic structure and expected findings. Corpus-driven research, on the
other hand, explores linguistic structures that emerge from the corpus, thus inductively exploiting the potential of a corpus to identify linguistic categories and
units that have not been previously recognised (Biber 2009: 278). One of the main
claims that could be made about pragmatics research using corpora is that it is by
and large corpus-based and not corpus-driven (Sandford 2008). Most studies
have as their starting point a previously recognised structure, where the main value
of the empirical work results from the investigators meticulous analysis of the
range of different functions associated with particular linguistic forms, as evidenced by the naturally occurring discourse contained in the corpus. Examples
would be a structure like the it-cleft in English (e.g. Collins 2005) or an individual
word form like really; see e.g. Stenstrm (1986b), who provides a detailed account
of this form and its various functions in the London-Lund Corpus. The corpus
method principally involves inspecting data via a search term that yields relevant
589
results, i.e. the method known as one-to-one searching (del and Reppen 2008:
2). Often, however, the inspection of extracted corpus data may well result in interesting observations of functional categories not previously accounted for or
even recognised, such as the observation made by Andersen (2001: 139ff) that the
invariant British English form innit (from isnt it or aint it) has an extended function as a marker of mutual manifestness (common ground) directed towards the
previous speakers utterance, in addition to its more generally recognised use as a
tag question which modifies a proposition of the current speaker. Most corpusbased pragmatics studies are thus form-based. The computerised corpus lends
itself well to this kind of approach, in which relevant instances of a type can be
identified as tokens in actual use.
The methodological counterpart is the function-based approach, which takes as
its basis a particular pragmatic function and describes its possible realisations in
actual discourse. Functions that have been investigated include conversational
openings and closing, topic transition, making a request, expressing disagreement
or the like. An example of a study where this method is successfully applied is
Mauranens (2004) study of hedges, items which make propositions more or less
fuzzy and add indeterminacy of one kind or another to an utterance (Mauranen
2004: 194). She shows that in a contextually constrained corpus, the Michigan
Corpus of Academic Spoken English (MICASE), this function is typically performed by a category of items kind of, sort of, or something, or so, somewhat, just,
and a little bit, items which share some functional characteristics as hedges but
which perform specialised functions in different contexts of use.
In both the form and function-based approaches, the role of the corpus is essentially to provide a repository of linguistic examples to be accounted for. Until recently, it has been less customary to carry out corpus-driven studies in pragmatics,
despite the fact that this approach has been applied widely and successfully in related fields, particularly phraseology and lexicography (cf. Sinclair 1991; Partington 1998; McEnery et al. 2006; Biber 2009). Notably, the corpus-based and corpus-driven approaches should be seen as complementary rather than conflicting.
As Biber argues, corpus-driven research is not in any way superior to corpusbased research. The two require radically different methods, and thus they offer the
possibility of uncovering radically different perspectives on language structure and
use (Biber 2009: 279). Studies in pragmatics are likely to gain from using corpora
in manners which go beyond the dominating pattern of one-to-one search used in
corpus-based studies. That this combination may be fruitful can be demonstrated
by a recent paper by Walsh et al. (2008). Their study of vagueness markers in academic discourse is essentially a synthesis of two methodologies not always seen
as complementary, corpus linguistics and conversation analysis. These have much
to offer each other as they provide both quantitative and qualitative insights respectively (Walsh et al. 2008: 16). Using this combinatory approach, the authors
show that vague language varies considerably according to mode of discourse
590
Gisle Andersen
(managerial, classroom, conversational) and that these qualitative differences are

a valuable means of understanding the complex relationship between language and
learning (Walsh et al. 2008:9). Furthermore, discourse functions are often associated with processes of language change, involving the emergence of new discourse
items, realised either as new forms, such as the English phrase yeah right used as
an irony marker in Norwegian (Andersen 2006), or as new uses of old forms, such
as the new quotative marker BE like (Tagliamonte and Hudson 1999; Andersen
2001; Rhlemann, this volume). These cases exemplify processes of change (borrowing vs. structural reanalysis) that are observable using corpus-driven methods.
Studies in pragmatics should investigate changes in vocabulary and in collocational and colligational patterns, with a view to revealing new ways of organising discourse, structural change, categorical reanalysis and emergent grammaticalisation (Traugott and Heine 1991; Lindquist and Mair 2004). In other words,
there are good reasons for complementing the well-tested and approved method of
corpus-based pragmatics with a corpus-driven approach.
The current chapter focuses on studies where corpora have been applied successfully in pragmatics research. The chapter is organised as follows. Section 2
provides an historical overview of the field of corpus-based pragmatics, giving a
brief description of a wide range of studies in order to show the breadth of this
field. Section 3 concerns method and has the two-fold purpose of arguing in favour
of a corpus-based approach as well as problematising the hermeneutical aspects
of this method. Section 4 provides an in-depth account some particularly salient
topics and contributors in corpus-based pragmatics. These include discourse
markers, prosody, listener activities (backchannels) and corpus-based models of
discourse structure. The final section reiterates some of the main claims and observations made in the previous sections.
2.
An overview of the field
Corpora are compilations of naturally occurring spoken or written language that

can be accessed on a computer. Such compilations may be monolingual or multilingual and may represent general language or specific domains (professional/academic corpora). This study will show that corpora form a well-prepared basis for
systematic, descriptive studies of instances of actual speech, of language variation
and of how social context constrains communicative practices. Spoken corpora
have a favoured position in pragmatics. A spoken corpus is a collection of speech
data made accessible via a computer, containing at least transcriptions of speech
but often also audio and/or video files (for problems and types of transcription, cf.
Roger J. Kreuz and Monica A. Riordan, this volume).
Corpora have been available for linguistic research from the 1960s. Since the
early days of corpus linguistics, there has been an interest in using corpora as a
591
means of exploring functional and contextual aspects of language use. However,

following the compilation of the first-generation written corpora, the Brown University Corpus (Francis 1967) and the Lancaster-Oslo/Bergen corpus (Johansson
and Hofland 1994), the bulk of studies were within other fields of linguistics, particularly syntax, semantics, morphology and lexicology (cf. Altenberg 1986 for
an overview). Though few and far between, early pragmatics studies include linguistic stylistics (Enkvist 1973) and aspects of text linguistics, addressing such issues as information structure, fronting and topicalisation (Gustafsson 1982, 1983,
1985; Davidson 1977). The parallel structure and content of the Brown and LOB
corpora enabled reliable comparative studies of American and British English that
focused on lexical and grammatical differences (e.g. Hofland and Johansson 1986;
Algeo 1988).
The ground-breaking compilation of the first spoken corpus, the London Lund
Corpus of Spoken English (LLC, Svartvik and Quirk 1980), triggered a range of
studies in pragmatics (Altenberg 1986). The corpus was based on recordings of
spoken data from the Survey of English usage (SEU), which in its pre-computerised version had resulted in Crystals detailed empirical work on intonation (Crystal 1972, 1975; Crystal and Quirk 1964), Haegemans (1984) empirically based account of pragmatic conditionals like the story, if so it may be termed, is weak and
loose and Kussmauls (1978) early study of Indikatoren von Sprechakten (speech
act markers). LLC inspired research into phenomena relating to speech production
and performance, such as slips of the tongue (Garnham et al. 1981), repairs (Hedstrm 1984) and hesitation (Orestrm 1977).
Several corpus-based studies were collected in what must be seen as a seminal
volume on conversational discourse, namely Enkvist (1982). This volume addresses an array of phenomena relating to the study of speech, including turn-taking, feedback and repair (Faerch and Kasper 1982; Orestrm 1982; Stenstrm
1982), intonation and prominence (Brazil 1982), pragmatic particles (stman
1982), referential expressions (Thavenius 1982) and functional aspects of negation
(Tottie and Paradis 1982). The book contains pioneering work on theoretical and
methodological topics such as transcription, segmentation and annotation of conversational data (Loman 1982; Svartvik 1982) and the validity of contrastive
methods (Sajavaara 1982). Also notable is the variety of languages represented by
the corpus data described in this volume (English, French, German, Hebrew, Russian and Swedish).
A topic which received early attention is discourse markers (dealt with more
extensively in section 4). The early contributions include Ermans work on pragmatic expressions like you know, you see and I mean (1986, 1987), Stenstrms
work on really (1986b) and Svartviks investigation on the various functions of
well (1980). Moreover, Stenstrms (1982, 1984) extensive work on question/response sequences (cf. section 4) dates from this period, as do some accounts that
deal specifically with tag questions (Nsslin 1984; Bublitz 1979). The availability
592
Gisle Andersen
of both spoken and written corpora also inspired comparative work, particularly
Bibers (1986, 1987, 1988) work focusing on the lexical, grammatical and pragmatic differences between these two modes of communication. In addition, LLC
formed the basis of an at least partly corpus-based account of conversational phenomena, Wardhaugh (1985), which has been criticised for its superficiality and
lack of an analytical apparatus and technical terminology (Garnham 1986).
A second wave of corpus-based studies came with the development of additional English language corpora, such as the Kolhapur Corpus of Indian English
(Shastri 1988), the Birmingham Collection of English Texts (Renouf 1986; Sinclair 1987), the Tosca Corpus (van den Heuvel 1988), The Nijmegen Corpus (Aarts
and Meijs 1986) and The Helsinki Corpus of English Texts, representing historical
English (Rissanen 1986; Kyt and Rissanen 1988). Moreover, the introduction of
formal annotation schemes for intonation in speech corpora was a timely and relevant innovation that became highly influential to pragmatics research (Svartvik
1987, 1988; Knowles 1986; Knowles and Lawrence 1987; Taylor and Knowles
1988; Knowles et al. 1992). This sparked research on the pragmatic aspects of
prosody (Altenberg 1987; Nevalainen 1987a) and studies exploring the correlation
between intonation and type of speech act (Geluykens 1986, 1987a, 1988) and between intonation and information structure (Nevalainen 1987b). The annotated
corpora enabled studies of the functional characteristics of pauses and other types
of prosodic segmentation (Stenstrm 1986a, 1988). The range of pragmatic phenomena accounted for increased in the 1980s and came to include investigations of
dislocation (Geluykens 1987b; Aijmer 1989), evidentiality and epistemic modality
(Westney 1986; Holmes 1988; Coates 1987; Biber and Finegan 1989), turn-taking
(Brodda 1988) and direct vs. indirect speech (Aijmer 1986).
There has been a gradual increase in the range of domains of language use that
are represented by corpora, and especially in the period from the 1990s, the growth
in size and coverage has been remarkable. The second generation of corpora includes large, general-purpose corpora like the 100 million word BNC corpus
(Crowdy 1995; Aston and Burnard 1998) and the 20 million word Birmingham
Corpus. A notable coordinated effort is the International Corpus of English (ICE,
Greenbaum 1991, 1992, 1996; Banjo 1996), which has been used for cross-varietal
studies of the English spoken in many Anglophone societies. The Santa Barbara
Corpus of Spoken American English (CSAE, Chafe et al. 1991) constitutes the
American spoken component of ICE. The family of corpora also came to include
corpora for specific purposes, such as foreign learner corpora like the International
Corpus of Learner English (ICLE, Altenberg 1997; Granger 1997) specialised LSP
corpora (Faber and Lauridsen 1991; Fang 1993) such as corpora of doctor-patient
interactions (Thomas and Wilson 1996), and parallel corpora such as the EnglishNorwegian Parallel Corpus (ENPC, Johansson and Hofland 1994) and the EnglishSwedish Parallel Corpus (ESPC, Altenberg and Aijmer 2000). Finally, corpora that
focus on regionally and socially restricted varieties of English include the Bergen
593
Corpus of London Teenage language (COLT, Haslerud and Stenstrm 1995; Stenstrm et al. 1988).
The growth in availability of corpora of various kinds has led to a growth in
corpus-based studies in pragmatics, although the relative growth has been larger in
other fields of linguistics. Several important contributions can be mentioned. With
respect to contrastive studies, Granger and Tyson (1996) and Granger (1997) compare discourse aspects of how native and non-native writers use connectives and
participle clauses in English academic writing. Work on phonology and intonation
advanced greatly through efforts in the SEC/MARSEC corpora (Bald 1990; Bauer
and Holmes 1996; Knowles et al. 1996; Wichmann 1996; Nevalainen 1992). Finally, studies in pragmatics came to cover an ever wider range of topics, including
amplifiers (Altenberg 1991), thanking, apologies and requests (Aijmer 1996), register (Biber 1993, 1994, 1995), conversational humour (Fill 1992), expletives and
taboos (Stenstrm 1991, 1995, 2006) and backchannels and response signals (Bublitz 1988; Tottie 1991).
In its initial phase, corpus linguistics was practically synonymous with English
corpus linguistics. The establishment of the ICAME (International Computer
Archive of Modern and Medieval English) organisation in 1979, and its coordinated development and distribution of many English language corpora, have provided a particularly fruitful environment for corpus-based research. Recently,
however, corpora for a number of other languages have been compiled, including
large, national corpus-building initiatives such as the Croatian National Corpus,
the Czech National Corpus, the Spoken Dutch Corpus, the Gothenburg Spoken
Language Corpus (Swedish), and a number of other, more targeted corpora developed in connection with specific research projects, especially in sociolinguistics
(Andersen 2010). For example, de Haan (2002) uses the Uppsala corpus of contemporary Russian texts to explore different means of expressing modality and negation in Russian and the ambiguity of different markers of modality. Similarly,
Schelfhout et al. (2005) explore the Dutch National Corpus, observing that speakers use of interjections varies significantly with textual genre as well as with
grammatical, functional and prosodic factors of the utterances in which they o
ccur. Among the international corpora available, the C-ORAL-ROM project has
launched comparable spoken corpora in four Romance languages, French, Italian,
Portuguese and Spanish (Cresti and Moneglia 2005). A study like Schneider
(2007) draws on no less than 22 corpora of spoken French, Italian and Spanish in a
comprehensive comparative account of parenthetical clauses like je crois, je dirais,
je pense, je sais pas and je trouve, and its equivalents in the other Romance languages. The Spanish Corpus Oral de Lenguaje Adolescente (COLA) enables comparative studies of the varieties of Spanish used by adolescents in major Hispanophone cities, and has been explored extensively in pragmatics studies (Stenstrm
and Jrgensen 2009). In contrast, the Corpus de Referencia del Espanol Actual
(CREA) represents Spanish spoken by adults. Downing (2006) explores CREA in
594
Gisle Andersen
a contrastive study of the stance marker surely and its Spanish cognates. Furthermore, the Italian spoken corpus Lessico dellitaliano parlato (LIP, De Mauro et
al. 1993) forms the basis of Waltereits (2006) study on the role of speaker motivations in the sense of abstractions of communicative intentions (Waltereit 2006:
62) in the grammaticalisation from lexical items to discourse markers of Italian
forms such as insomma and diciamo. The Lancaster Corpus of Mandarin Chinese
(McEnery and Xiao 2004) is explored by Ming and Chen (2010) in a recent study
of how the discourse function of noun phrases affects the word order in noun
phrases with a relative clause. Factors such as the information status, animacy, and
high versus low discourse salience of head NPs have a bearing on whether a Mandarin Chinese relative clause immediately precedes the head NP or whether it is
separated from its head by a numeral-classifier sequence consisting of yi (one)
and a classifier. Noun phrases representing abstract entities with a low discourse
salience are contrasted with concrete (animate or inanimate) entities with a high
discourse salience, and the relative order of the components are explained with reference to the principle of relevance (Sperber and Wilson 1995), as it seems to indicate the direction in which the relevance is to be sought (Ming and Chen 2010:
188). Used with a fully realised and highlighted discourse function of introducing a
salient participant, the numeral classifier tends to precede the relative clause;
otherwise it tends to follow the relative clause. The authors argue that this shows
that clause level structural analysis does not suffice to explain the underlying factors that govern the choice of alternative means of expressions in a language. By
contrast, corpus data can be tremendously useful in helping us arriving at a realistic
account of the structure of human language (Ming and Chen 2010: 188).
Despite notable studies like the ones mentioned, there is still a considerable
bias towards English-based studies in pragmatics in the literature (McCarthy and
Carter 2004b).2 Importantly, however, it should be pointed out that the availability
of corpora representing different languages, including parallel corpora consisting
of original and translated texts, has triggered significant comparative work. A recent volume that reports a number of multilingual studies is Aijmer and SimonVandenbergen (2006). The authors argue that comparative work is interesting for
its potential to lead to new knowledge about what is universal and what is language-specific (Aijmer and Simon-Vandenbergen 2006: 3), for its potential in investigating etymological or semantic cognates, and for revealing the effects of
translation on pragmatic aspects. These general claims are substantiated by the individual papers in the volume, such as Johansson (2006) and Hasselgrd (2006),
who jointly demonstrate the applicability of parallel corpora in determining the
meaning and distribution of the Norwegian English cognate words vel/well and n/
now, respectively. Moreover, Altenbergs (2006) account of adverbial connectors
shows how cross-linguistic similarities are largely discourse-driven. In the languages compared, Swedish and English, contrastive and exemplifying connectors
often serve to highlight the initial element, syntactically and prosodically, thereby
595
indicating a significant break or shift in the discourse. Hence, despite the syntactic
differences between English and Swedish, there are interesting parallels between
the two languages in the way that connectors adjust to, and signal, the progression
of discourse (Altenberg 2006: 35f). Similarly, Foolen (2006: 70) compares Dutch
toch and German doch, showing that the multilingual approach makes it clear
more than a monolingual study could have done, that the language specific uses of
the same item are indeed rather specific, but at the same time that in all these applications a single schematic meaning was preserved. The distributional differences of the two items are well accounted for; e.g., judging by translators choices
when handling doch/toch as an adversative connective, [a]pparently, German does
not like a bare doch in first sentence position (Foolen 2006: 65). This might be
explained by German dochs status as a conjunction. As modal particles, only toch
can occur in sentence-final tag position, where it functions as a checking device,
inviting an affirmative response from the hearer. As a whole, the collective volume
shows that the contrastive corpus-based approach provides a new perspective on
fundamental properties of discourse markers, such as their multifunctionality and
core meanings.
The most recent advances in corpus linguistics are connected with increased
data processing and storage capacity, the emergence of web-based corpora, and the
(semi-) automatic compilation of large monitor corpora which are constantly updated with new texts representing an unbroken, chronological data flow (Renouf
2007: 36). These recent advances have changed the face of corpus linguistics
(Renouf and Kehoe 2006: 3) to a much richer cycle of events, consisting not just
of small, medium and large primary corpus building, but of specialised and multidimensional secondary corpus building; not just of corpus analysis but also of corpus evaluation; not just of an initial application of theory but of self-reflection and
a new concern with theory in the light of experience (Renouf and Kehoe 2006: 3).
The tangible outputs of these processes are super-corpora containing hundreds of
millions of words, like the Bank of English at the University of Birmingham, The
Corpus of Contemporary American English (COCA; cf. Davies 2009) and the Norwegian Newspaper Corpus (Andersen 2005). These continually growing corpora
are particularly apt for studying change in progress and for revealing innovations
in language use, including the development of new and borrowed lexical items
(Renouf 2007; Furiasso and Hofland 2007). Web-based corpora, aka. cyber-corpora (Renouf 2007) use web crawler technology to harvest data which are subsequently filtered, annotated and transformed into a corpus. This has resulted in a
growing body of corpus-based studies using the web as its prime source of data
(Kilgarriff and Grefenstette 2003; Hundt et al. 2007). Although the exploitation of
these large corpora in pragmatics has so far been limited (but see section 3), the recent developments forecast a promising future for corpus-based pragmatics.
596
3.
Gisle Andersen
Methodological issues
An important event in the history of corpus linguistics was the 1991 Nobel Symposium on Corpus Linguistics, which is comprehensively reported in Svartvik
(1992). This volume does not focus on issues in pragmatics in particular, but it contains an impressive combination of papers that justifies its mention in this context.
It presents a theoretical and methodological basis with great scientific validity for
any type of corpus linguistic work, despite subsequent major technical developments that go well beyond what the pioneers of corpus linguistics could have foreseen (cf. above). The influential contributors, among them Chafe, Nelson Francis,
Fillmore, Halliday, Johansson, Greenbaum, Rissanen and Quirk, discuss a number
of theoretical issues and issues in observational methodology, corpus design and
corpus use. For instance, Chafe provides a strong argument for corpus linguistic
method in cognitive linguistics: the acceptance of the value of corpora is closely
tied to the acceptance of the view that language cannot be separated from what the
mind is and what the mind does and how the mind interacts with the universe in
which it is located. The modular view of language can get along without corpora;
the integrated view requires them (Chafe 1992: 81). Chafe further contextualises
corpus linguistics as an observational technique, placing the use of corpora within
a larger repertoire of observational alternatives (Chafe 1992: 84). Unlike introspective techniques that involve making judgements regarding invented language,
corpus linguistics involves the observation of overt behaviour. Further, corpus linguistics involves natural language, which distinguishes this method from artificial
methods like experimental techniques and elicitation tests. Corpora thus reflect
what people actually say and write. Ontologically, corpus linguistics shares its
characteristics with ethnography, in that it is both a natural and behavioural observation technique.
However, the introspection/observation dichotomy is problematic, especially
from the point of view of pragmatics. As Milroy (1987) points out, [i]n practice,
generalizations based on analyses of corpuses (sic!) are usually derived using an
additional method either introspection or some kind of experimental procedure. entirely corpus-based linguistic description of the kind attacked by
Chomsky (1957: 15; 1965: 2) is extremely rare and probably represents a concept
whose value is largely polemic (Milroy 1987: 4f). The corpus-based approach to
pragmatic phenomena relies on supplementing observations of overt linguistic behaviour with the intuitive knowledge of the analyst. Pragmatics concerns meaning
arrived at on the basis of inferential processes. The empirical researcher uses
spoken data as his/her primary stimuli on the basis of which he/she makes assumptions as regards the communicative intention of the observed interlocutors. Observable forms in a corpus, such as tokens of the word please, represent instantiations of pragmatic acts performed by speakers, such as the speech act function of
requesting somebody to do something, the attitudinal function of signalling tenta-
597
tiveness or the politeness function of face-threat mitigation. Naturally, it can be

hard to judge regarding speaker intentions. This methodological challenge is well
accounted for in Mosegaard Hansens (1998) comprehensive account of discourse
particles in French, where she argues that [a]n interpretative approach of course
has an inescapably subjective dimension, which means that one cannot in any
sense prove the correctness of ones analysis, but one can only hope to demonstrate
it as convincingly as possible (Mosegaard Hansen 1998: 214).
The complexity of spoken language data poses a challenge to the compilers and
users of corpora. The study of linguistic phenomena can be enhanced by annotations of various types of morpho-syntactic and discourse-structural information,
and contributions like Edwards and Lampert (1993) and Leech et al. (1995) offer
systematic accounts on the art of transcription and annotation of spoken discourse.
The availability of prosodically annotated corpora has provided uniform methods
for studies of intonation and its pragmatic effects (cf. e.g. Stenstrm 1984; Aijmer
1996). One objection that has been raised against the first generation of corpusbased studies is the fact that they regard the transcriptions themselves as the primary data, while the sound files are not consulted due to their unavailability to the
analyst. Although the unavailability of audio files can be explained with reference
to historical, practical and technological details (Wichmann 2007), this remains a
justifiable scientific objection for at least two reasons.
Firstly, like phonetics and phonology, many studies in conversation analysis
and discourse analysis need audio data because the interpretation depends on audible phenomena that could never be fully represented in the transcripts: there is far
more in the speech signal to be accounted for than could ever be captured in an orthographic transcription (Wichmann 2007: 76), and [n]ot even the most detailed
transcription can faithfully represent the raw data of spoken interaction (Mosegaard Hansen 1998: 207; cf. also section 4). Prosodic and segmental phenomena
play a more significant role than earlier anticipated in turn-taking, discourse marking, etc. (cf. Couper-Kuhlen 1996; Venditti and Hirschberg 2003; Wichmann 2007).
For example, Walker (2004) uses a variety of (mostly telephone-based) corpora to
explore how features such as pitch movements, glottal aspects, lingual and labial
closures are systematically associated with the signalling of transition relevance.
Secondly, it may be problematic for corpus users to rely fully on the judgements of the transcriber, since transcribers often differ with respect to their interpretations of spoken data. Ensuring inter-annotator agreement is a major challenge
in corpus compilation, and a study such as Raymond et al. (2002) has shown that
even experienced transcribers have different interpretations, particularly when annotating data at phonetic or prosodic levels. This underlines, of course, the analysts need for access to the raw spoken data when using corpora (Andersen 2001:
211ff).
Fortunately, new technology has made it much easier to access audible spoken
data in corpus linguistics. The state-of-the-art spoken corpora contain hypertext
598
Gisle Andersen
links between the transcripts and the audio and/or video data, cf. corpora such as
COLA, COLT, ICE-GB and NoTa (Johannessen and Hagen 2008; Andersen 2010).
There is thus a tendency in corpus building to move away from ambitious paralinguistic annotation schemes aimed at transcribing infinity (Cook 1990: 1) to implementing a more a rudimentary, but nevertheless searchable, prosodic transcription coupled with the actual sound files (Wichmann 2007). Importantly, video data
facilitates the analysis of gestures and facial expressions as part of conversational
behaviour, but the utilisation of video material in corpus-based pragmatics is still
in its infancy. The importance of using video data can be illustrated with reference
to minimal response signals (backchannels). Since active listenership may be realised verbally or non-verbally, typically by a gesture such as nodding, the articulated realisations cannot and should not be seen as separate from non-verbal realisations performing the same discourse function (Bjrge 2009). In other words, a
comprehensive study requires a holistic approach to language data, in which all aspects of utterances are investigated (cf. also Selting and Couper-Kuhlen 2001;
Couper-Kuhlen and Ford 2004). The fruitfulness of such an approach can be
shown with reference to a recent dissertation, Loehr (2004), that explores the functional parallelism between intonation and gestures. Specifically the study demonstrates the ability of gestures like intonation to signal completeness/incompleteness, information status, focus, emphasis, contrast, and even discourse
structure; in fact that hand movements reflect the on-going status of discourse,
while [i]ntonation signalled the discourse status much less subtly (Loehr 2004:
235). The exploration of the interaction between gesture and intonation is a
relatively recent academic pursuit, and studies have looked at temporal alignments
of the two modalities and the mapping between their respective hierarchical structures. Interestingly, there is strong evidence, provided initially by Kendon (1972,
1980) and McNeill (1985, 1992), that gesture and speech in general are two surface
facets of a single underlying thought being expressed (Loehr 2004: 1). Despite the
potential richness of corpus annotation, such a connection could not have been explored using merely transcription-based methods.3
4.
Corpora as basis for the study of topics in pragmatics
In the following, I address topics in pragmatics that have been devoted particular
attention from a corpus-based perspective, namely discourse markers (with a
special focus on Aijmer 2002), prosody (focusing on Wichmanns (1994, 2000,
2004, 2005) work), listener activities/backchannels (Bublitz 1988; Tottie 1991),
and corpus-based models of discourse structure (Sinclair and Coulthard 1975;
Stenstrm 1984, 1994; Sinclair 1996, 2004).
4.1.
599
Discourse markers
It is perhaps not surprising that discourse markers/particles or pragmatic

markers/particles are among the most focused topics in the literature, as this class
lends itself easily to form-based retrieval in corpora.4 Studies include items like
well, like, kind of, sort of, I mean etc. in English (Brinton 1996: 32), eh bien, donc
and allors in French (Mosegaard Hansen 1998), aber, doch, and ja in German (Abraham 1991), etc. With respect to terms and inventory, Fischer (2006) provides a
good overview of conceptual and terminological differences. The meanings of discourse markers are acknowledged to be excruciatingly hard to state (Wierzbicka
1991: 341). They make only a limited contribution to propositional content, but
contribute considerably at the pragmatic level, to the organisation of discourse (for
instance as markers of speaker transition or by marking boundaries in the discourse), to marking a speakers attitude towards the propositional content of the utterance (e.g. the Norwegian particle vel signalling reduced speaker commitment;
Fretheim 1981) and to various functions pertaining to the speaker-hearer relation,
such as you know indicating common ground between speaker and hearer (Jucker
and Smith 1998). Among the substantial contributions are Lenks (1998) study of
connectives and Andersens (2001) sociolinguistic study based on COLT and a
comparable subset of the BNC. The former focuses on anyway/anyhow, however,
still, incidentally, actually and what else in British and American English represented by LLC, CSAE and ICE-GB, whereas Andersen (2001) describes two discourse markers, like and innit, showing their pragmatic function, grammaticalisation and distribution (cf. Rhlemann, this volume). These two studies are
theoretically based in Sperber and Wilsons (1986/1995) relevance theory and analyse the pragmatic function of discourse markers as procedural constraints on
relevance, as does Jucker (1993) in his account of well. Comparative and cross-linguistic aspects of discourse markers are covered by Aijmer and Simon-Vandenbergen (2006) and Fischer (2000), analysing German ja, also, ne, oh and ach and
English yes, yeah, oh and well. Several of the studies in Jucker and Ziv (1998) and
Andersen and Fretheim (2000) are also based on corpora. Furthermore, Mosegaard-Hansen (1998) presents a detailed analysis of the six French markers bon,
ben, eh bien, puis, donc and alors. She bases her study on a collection of recorded
and transcribed corpora obtained from the archives of the UFR de linguistique
francaise of the Universit de Paris III (Mosegaard-Hansen 1998: 201). Among
the markers investigated are donc and alors, forms which both originate in temporal anaphoric expressions (deriving from the Latin forms dunc/illa hora, at that
time/at that hour, respectively). While donc is a monosemous marker of mutual
manifestness, in the sense of Sperber and Wilson (1995), the much more common
marker alors is multiply polysemous and its core use involves a shift in perspective, e.g. a new topic, marking a result or a conclusion, functions which are derived
metonymically from their diachronic sources.
600
Gisle Andersen
Aijmers (2002) contribution is illustrative of both the advantages of the corpus-based approach, as well as some of its problems. She describes a class of
items with fuzzy boundaries as they are used in LLC with occasional comparisons
with LOB and COLT. The inventory of forms includes now signalling emphasis or
transition, oh and ah and their functions as discourse particles signalling topic
shift, clarification or elaboration, just as a marker of speaker involvement, affect or
rhetorical use, the evidential particles (hedges) sort of and kind of, vagueness
markers like and all that, and that sort of thing, etc., and actually used with evidential functions. The value of Aijmers contribution lies first of all in its extensive
qualitative analysis of individual tokens in relevant contexts of use, as well as her
firm reliance on data, in what she considers a bottom-up approach. She develops
very specific criteria needed to distinguish discourse particles from other (often adverbial) uses of the items investigated, and focuses on core meanings and their derivative relation to the discourse functions of individual items, including their
grammaticalisation. A fundamental notion in Aijmers account is that of indexicality, that is to say, the potential of discourse particles to point to the speakers
epistemic attitude to the utterance and affective attitude to the hearer as well as to
the preceding and following discourse (Aijmer 2002: 55). This capacity can be likened to the indexicality of deictic expressions like anaphoric pronouns, in the
sense that they point to the presence of some entity in the immediate situation-athand (Ochs 1996: 414), including the speaker and hearer, spatial and temporal dimensions of the speech situation, epistemic stances like certainty/uncertainty, attitudes towards hearer relationship, social rank, group identity and attitudes towards
the speech act or some of its characteristics. Aijmer argues convincingly that discourse particles must be construed as polysemous items in what can be seen as
a modified meaning minimalist programme, having a single core meaning and
various related senses, as opposed to the proposed homonymy analysis in a meaning maximalist programme, in which particles are seen as having multiple meanings determined by the context of use, explicitly refuted by stman (1995) and
Mosegaard Hansen (1998). It is precisely their polysemy that can explain the grammaticalisation and the diachronic and synchronic relation between a discourse particle and its predecessor. For example, the temporal and topic transitional meanings of the discourse particle now are seen to co-occur as a result of a development
from adverb to discourse particle. Aijmer performs a meticulous qualitative analysis of each individual item, and the treatment of each particle includes descriptions
of a variety of features, including the pragmatic-functional range, collocational
analysis, prosody, grammaticalisation and relation to the words core meaning,
and, where relevant, comparisons with other particles, such as the comparison of
now with well with their respective adverbial functions.
A significant aspect of this contribution is the collocational analysis of each
particle. Specific collocational characteristics are viewed as a functional clue to
assist the hearer when a particle is ambiguous between different senses (Aijmer
601
2002: 28). To exemplify, the pragmatic value of oh is dependent on its co-occurrence with well, as the collocation oh well specifically indicates surprise and acceptance, at the same time belittling or trivialising a previous utterance (Aijmer
2002: 104). This can be contrasted with the collocation ah well, which has an upgrading effect, signalling the importance of the upcoming message. Collocations
also have a potential for distinguishing between marker and non-marker (i.e.
usually adverbial) use of an item, in that now in the contexts of well now, now then
and now look are invariably analysed as discourse markers. However, Aijmer does
not fully explore the possibility that collocations may also be markerhood indicators. It is for example notoriously difficult to draw the line between marker and
non-marker use of just (Tottie 1986), in other words the restrictive adverb with a
propositional meaning of exactly or only. Aijmer devotes relatively limited attention to the collocational patterns of just (but see Erman 1997) and it seems fair
to assume that the collocational analysis could have been fruitfully extended to include descriptions of how specific phrases are (perhaps exclusively) associated
with marker use, while others are not. An objection that could be made towards
empirical studies of discourse markers in general, is that collocational features are
largely underexplored. There are good grounds for claiming, as does Aijmer herself that [a]mong the characteristic properties of discourse particles which have
not received enough attention in the literature is their clustering (collocations
with other particles) (Aijmer 2002: 31; see also Vicher and Sankoff 1989). In
studies of discourse markers, the mention of collocational features is often coincidental rather than systematic. Aijmers work is illustrative in this respect, as her account of just amounts to a mere listing of the top most frequent collocational
patters (as in Aijmer 2002: 161), and the collocational analysis for actually is
ignored altogether. Moreover, the presumably common collocation oh really is
ignored in the section on collocations, albeit acknowledged as a unit with a separate status as a topicaliser in the discussion of pragmatic functions.
Aijmer gives no reason for her selection of the seven items studied. Her only
reference to criteria for inclusion is frequency in LLC, but we are not told whether
this frequency as discourse particles or overall frequency of the forms in question.
One might well ask why a particle like well is not accounted for in her description;
presumably this is because she finds this particle sufficiently covered in the abundant previous literature. Besides, the common use of like as a discourse marker is
only devoted sporadic attention, illustrating a difference between LLC and subsequent English corpora.
4.2.
Prosody
The availability of spoken corpora has prompted many studies on prosody, from
the early LLC-based studies using prosodic transcriptions as primary data, to
studies using instrumental analysis of speech data as a basis for research. In prag-
602
Gisle Andersen
matics, many corpus-based studies fail to explore prosody systematically as part of

the analysis. With respect to the literature on discourse markers outlined above,
this general shortcoming is surprising, since prosody may be crucial for the determination of markerhood status, and since grammaticalisation processes tend to be
accompanied by reduction in form. Aijmers (2002) study is exceptional in that it
utilises the prosodic annotations of LLC extensively, which adds to the credibility
of the analysis.
An important effort in the study of intonation is the Intonational Variation in
English (IViE) Corpus, containing speech data and intonation transcriptions from
nine urban dialects of British English. The data represent five different speaking
styles, including conversation, map task dialogues, retold stories, reading passages
and sentences. Fletcher et al. (2004) use this corpus in a comparative study of the
high-rising terminal (HRT) in four varieties of English. HRT refers to the use of the
high-rising intonation contour generally associated with yes/no questions in English in declarative utterances. With respect to the function of this feature, Guy
and Vonwiller suggest that HRT correlates with the semantic complexity of the
text and therefore the need for checking to see if the audience is understanding
what is being said (1989: 25). Previous studies of Australian and New Zealand
English have shown that HRT is associated with low prestige varieties, is socially
stigmatised, occurs most often in the telling of narratives and is used predominantly by young adolescent females as part of narratives (Allan 1990). Contrary to
this, Fletcher et al. (2004) show that it is the male speakers in IViE who use HRTs
the most. According to the authors, the advantage of using the IViE corpus for this
study is that it provides sufficient coverage of intonational phenomena across a
very large corpus comprising several, very different, varieties of British English
(Fletcher et al. 2004: 407).
Several contributions by Wichmann (1994, 2000, 2004, 2005) exemplify how a
corpus-based method provides a valuable complement to previous conversation
analytic work based on experimentally controlled perception studies (e.g. CouperKuhlen and Selting 1996). Much of her work is based on SEC, while more recent
studies explore the British component of the ICE corpus (ICE-GB). Although corpora which capture speech only in its natural habitat (Wichmann 2004: 1527)
often do not maintain the sound quality needed for instrumental analysis, Wichmanns work firmly demonstrates how an auditory analysis of the data can lead to
enlightening scientific results. Wichmann (2000) analyses how intonation is used
to signal the structure of spoken texts and its role in turn-taking and in signalling
cooperation and clarifying the relationship between utterances in conversational
discourse. For example, major shifts in topic are distinguishable from units with
close cohesive ties on prosodic grounds. Furthermore, Wichmann (2004) shows
the advantages of using a speech corpus in a study of please-requests. She describes the pitch contour and accentual status, focusing on how the realisation
varies according to the speech situation and the speech act function of utterances,
603
as well as the utterance position of please. Her study explores formulaic expressions in which please occur, notably yes please and please do, the isolate
please itself, which is used to reinforce a more explicit request (Wichmann 2004:
1539), as well as the exaggerated segmental realisation [phj li::z] (Wichmann
2004: 1540), where the epenthetic schwa leads to what is in effect a disyllabic realisation, usually expressing scorn or disapproval. The fact that her study is based on
a corpus representing speakers of varying backgrounds and different contexts of
use enables the author to relate her findings of prosodic patterns to the social distance and symmetrical/asymmetrical power relations between interlocutors. This
provides an interpretational perspective which supplements the existing accounts
of the discourse effects of terminal contours, e.g. of a falling contour as contributing a feeling of closure (Croft 1995) and a final rise as signalling openness
(Cruttenden 1997). At the same time she redefines the role of please itself used as a
face-threat mitigator in requests as a gesture of courtesy that contextualises the
accompanying request as occurring within a known set of rights and obligations
(Wichmann 2004: 1544). The importance of intonation for utterance interpretation
is further corroborated in Wichmanns (2005) study of emotional requests, which
also demonstrates the value of investigating low-frequency items. The intonation
of please help yourself with a high onset on please and a falling tone on self functions as a neutral invitation, while please help yourself, with a fall on please, is
much more pleading, functioning as an emphatic, emotional exhortation. Collectively, Wichmanns corpus-based work has contributed consideraby to our understanding of the relation between intonation and speaker attitude, showing that the
attitudinal information carried by intonation is not an inherent feature of specific
intonation contours but a result of context-based pragmatic inference.
Other corpus-based discourse studies focus on the segmental aspects of prosody. Two early contributions by Stenstrm (1990a, 1990b) explore prosodic segmentation and pauses in monologue and dialogue. Stenstrms (1990b) category of
pauses is taken to include so-called verbal fillers like hedges (sort of), softeners (you know), stallers (well) and initiators (anyway). Stenstrm points out that
pauses play important roles in reformulations and as turntakers, turnyielders and
turnholders. Based on the same corpus, Romero (1994) explores the role of pauses
as content-bearing elements, showing that filled and unfilled pauses contribute to
the realisation of Hallidays (1985) theme/rheme structure of utterances. The interpersonal function of the filled pause is to indicate a new move in the conversation,
which may or may not involve a speaker shift. In their textual function, filled
pauses are used as the starting point of the following message. Finally, Romero
(1994) argues that filled pauses also have an ideational function of signalling an element referred to elsewhere in the discourse, typically by anaphora. Though he
may be right in claiming that filled pauses carry meaning of a structural and interactional nature, this ideational function seems dubious. Though the filled pause
admittedly co-occurs with anaphoric referential expressions, it seems unjustified to
604
Gisle Andersen
claim that the pause has to be considered a meaning-carrying particle which conveys semantic information in terms of structure and content (Romero 1994: 502),
as this seems to suggest that pauses are capable of taking on referential meanings
and contributing to the propositional content of utterances.
4.3.
Listener activities and backchannels
Studies in corpus-based pragmatics have become increasingly concerned with the

role of listeners in conversation. Corpus data can be effectively used to demonstrate how interactants actively signal listenership, while allowing the other
speaker to continue to talk. Response signals convey comprehension and agreement or disagreement and are often also furnished with attitudinal meanings, signalling involvement of various types (positive/negative evaluation) and of various
strengths, ranging from weak doubt or scepticism to enthusiastic agreement. Listenership can be expressed by auditory or visual signals. By far, the existing literature has mostly focused on the auditory side, which includes backchannels such as
mm, mhm or uh-huh, attitudinally salient interjections such as wow or damn, discourse markers such as oh, yes/yeah and okay, lexical repetitions, etc. To a much
lesser degree, non-verbal signals such as nods, furrowed brows and gestures have
been studied (Norrick and Fischer 2009).
The term backchannel first appears in the pioneering work of Yngve (1970), referring to the existence of what I call the back channel, over which the person who
has the turn receives short messages such as yes and uh-huh without relinquishing
the turn (Yngve 1970: 568). The LLC has been instrumental in providing an empirical basis for the study of listenership in discourse. Orestrm (1983) provides an
early LLC-based account which describes backchannels in the wider context of
turn-taking which includes a rather narrow set of items (m, yes, yeah, mhm, no,
quite, aha, good) but excludes tokens preceded by the markers you know and you
see. Similarly, Bublitz (1988) uses a subset of the LLC (excluding meetings and interviews) in a study of recipient action which includes hearer activities and certain reactive or recipient speaker contributions (Bublitz 1988: 141). Acknowledging a generally prevailing tendency to agree with, assent to and support the
fellow-speakers attitudes and statements (Bublitz 1988: 153), Bublitz makes a
distinction between primary speakers, secondary speakers and hearers. Unlike a
secondary speaker, who makes a minor contribution to the topic at hand, the hearer
signals that he is taking note of what has been said and meant and uses particular hearer signals taken from a closed set (Bublitz 1988: 153). Bublitzs study is a
bottom-up approach where a careful qualitative analysis of backchannels precedes
generalisations about their basic functions, their high flexibility with respect to discourse placement, and their formal features (usually short and monosyllabic). The
closed paradigm of forms includes yes, yeah, yep, uhuh, h, m, mhm, I see and
really (with a falling tone). Bublitz excludes no from the category on the grounds
605
that it provides an expression of stating the speakers position as to the current

topic (Bublitz 1988:181), thus its function is more than simply taking note of the
previous speakers utterance.
Tottie (1991) provides a comparison between the British and American English
represented by LLC and CSAE. This contribution is important as it addresses
theoretical and methodological problems in connection with corpus-based listenership research. It focuses on differences between supportive and regulative functions of backchannels, that is, the dual function of signalling understanding/agreement and encouraging the speaker to continue. Tottie points out that [t]he term
backchannel has been extended to designate backchannel message (Tottie
1991: 256). In fact, in its current use, the term is multiply polysemous, as it is also
used to refer to forms that can realise the supportive and regulative functions of listenership signals, as well as actual instantiations (tokens) in real data (Andersen
2009). Tottie also problematises the fact that the two datasets may not necessarily
be comparable in all respects, for example since the LLC transcription records
tone in detail but does not account for lengthening, which is shown in the CSAE
transcription (Tottie 1991: 260). This illustrates that the information available to a
researcher does not always make corpora fully equivalent. A more substantial difference concerns the data themselves, not the transcription conventions. Tottie suggests that, in order to draw justifiable conclusions in a comparative corpus-based
study, it is necessary to control for a variety of non-linguistic factors in order to
avoid that emerging differences are artefacts of biased data. She explicitly discusses factors such as the speakers age, gender and affinity with each other, subject matter and degree of consensus, as well as whether or not the conversations are
surreptitiously recorded. Ideally, in comparative studies, the data should match on
all of these points, certainly if one is to make any sense of quantitative comparisons. Tottie acknowledges that her corpus data match in some of these respects, but
not all. Nevertheless, she is able to demonstrate convincingly the qualitative differences in intonation, syllabicity, articulation and length of the British and American variety of mm.
Tottie raises the issue of what distinguishes a backchannel from a turn, stating
that the most important principle is that backchannel status can be determined
only on the basis of the following utterance (Tottie 1991: 260). It is problematic
that Tottie links backchannelling so closely to observable speech production, without taking speaker/listener intentions into account. A backchannel such as mm may
be intended as mere supportive device without the listener intending to take the
floor. But it may well be that the primary speaker chooses not to continue his talk,
in which case the backchanneller cedes the opportunity and there is a change of
speaker. It is problematic to argue that this changes the functional characteristics of
the backchannel item to any great extent. Therefore, a sharp division between an
mm that does not involve change of speaker and one that does so does not seem justified. Totties study concludes with a valuable proposal for a future research pro-
606
Gisle Andersen
gramme on backchannelling, proposing the need to investigate whether subject

matter, sex, and number of participants determine the use of backchannelling and
how their prosodic realization is linked to discourse function (1991: 270). Although many of her questions still remain unanswered, subsequent corpus-based
research has provided important insights into this field of study. For example,
McCarthy (2002; cf. also 2003) studies response tokens used in everyday conversation, focusing on differences between the British CANCODE corpus and the
Cambridge-Cornell Corpus of Spoken North American English. Moreover, the
conversation analytical work of Gardner (1998) argues that research-based findings on receipt tokens should have bearings on how language is taught, concluding
that it would be worth considering exposing learners to these interactional exponents, and providing opportunities for practice and feedback (Gardner 1998:
220221). Heintz (2003) compares the backchannel behaviour of German and
American English speakers, demonstrating differences between Germans who are
proficient in English and those that are not. Concerned with the use of English as a
lingua franca (ELF), Bjrges (2009) recent investigation is valuable as it sees verbal and non-verbal aspects of backchannelling in context. She investigates video
recordings of ELF interaction involving students partaking in simulated business
negotiations, showing that backchannel behaviour varies according to conflict
level. Some of these results are corroborated by Andersen (2009, based on a similar
dataset), who considers backchannels in polylogic discourse, with a particular
focus on the interactional placement of backchannels, that is, the direction of backchannelling between two teams representing a seller and buyer. Finally, Kjellmer
(2009) considers the insertion point of backchannels, observing that, although
most often inserted at the end of a completed utterance, backchannel insertions
often occur when the utterance has not been completed. What this signifies is that
the information-content of an utterance can be conveyed to the listener before it
has been expressed in toto (Kjellmer 2009: 107).
4.4.
Models of discourse structure
Given the short history of corpus-based pragmatics, special attention should be

paid to scholarly contributions which are less concerned with micro-phenomena
but which focus on the theoretical basis of this research domain, addressing basic
issues such as the units of discourse and their sequential organisation, the mapping
of lexical, morphological and syntactical form with illocutionary function, participant relations, textual co-production, etc. Inspired by theoretical precursors like
Firth (1935) and Sinclair (1966), the seminal contribution by Sinclair and Coulthard (1975) was original in its exploration of naturally occurring speech in an academic atmosphere which massively discarded the notion that studying real data
might have any significance. Their choice of the classroom as the venue for
collecting samples of speech was motivated by their wish to control as many of
607
the potential variables as possible (Sinclair and Coulthard 1975: 6) and overcome
the many problems inherent in natural conversation and opt for a more simple
type of spoken discourse, one which has a much more overt structure, where one
participant has acknowledged responsibility for the direction of the discourse,
for deciding who shall speak when, and for introducing and ending topics (Sinclair and Coulthard 1975: 6). Based on a restricted corpus of transcribed speech
data, they develop an analytical apparatus which, despite subsequent revisions
(e.g. Sinclair and Brazil 1982; Sinclair and Coulthard 1992) and criticism (Stubbs
1984), still remains influential and has been inspirational for a wealth of research.
They propose a discourse model of exchange and transaction structure which
shows that spoken discourse can be systematically described according to its structure, just as syntax can be. The model, which is closely modelled on Hallidays
Categories of a Theory of Grammar (sic.; Sinclair and Coulthard 1975: 24; cf.
Halliday 1961), encompasses a hierarchy of discourse components, specified as a
ranking scale containing the four components transaction, exchange, move and act.
In its original version, the model also included a top-level of lesson which has in
later revisions of the theory been dismissed as a stylistic type specific to the usage
domain of classroom discourse. At each level, discourse units are divided into
classes which realise an element of a structure, and the internal structure of the unit
is described. Exchanges have a three-part structure containing the elements initiation, response, feedback (IRF-structure), with the corresponding classes of move:
opening, answering and follow-up (Sinclair and Coulthard 1975: 26). Despite later
criticism for example concerning the unnatural nature of classroom discourse (in
which teachers ask pseudo-questions) upon which the model is based, this contribution effectively involves no less than the inception of a research field, discourse analysis.
A decade later, Stenstrm (1984) develops this model further in a comprehensive analysis of question/response strategies. Her contribution is valuable in that it
uses LLC as a test bed for the discourse model in an analysis of authentic conversational data. She adds empirical support to the theory by accounting for how the
different levels defined in the model of analysis were manifested in the corpus
(Stenstrm 1984: 94). As its title suggests, the primary focus is on exchanges. Although there is a bias towards the analysis of responses, as opposed to the initiation
and feedback moves, Stenstrm provides a detailed account which represents a
much wider user domain than the original empirical basis for the model, selected
to represent the greatest possible variety of speech situations (Stenstrm 1984:
262), ranging from informal conversation to procedural dialogue (planned, public discourse; e.g. courtroom interaction). She extends the model in various ways
by introducing notions such as topic-bound transactions (anaphorically related to
the previous context) versus proposition-bound transactions (forward-oriented)
and by specifying optional and obligatory moves in exchange sequences. Her study
incorporates categories like hesitation, pauses, hedges, reformulations and the like,
608
Gisle Andersen
and shows how they are linguistically realised in a corpus which, for its time, was
both extensive and varied. The Stenstrmian model (Kirk 2002) is further developed into what the author herself refers to as a synthesis of much of the work that
has been done so far in the analysis of conversation (Stenstrm 1994: xii). Kirk
(2002) provides a good illustration of how the model can be applied in the analysis
of real corpus data, arguing that although there are many competing analytic models available (e.g. Edwards and Lampert 1993; Eggins 1996; Langford 1994; etc.),
none is as rigorous or as syntactically systematic as that of Anna-Brita Stenstrm
(Kirk 2002: 144). Despite Kirks praise, the model has raised criticism as regards
the conceptual nature of discourse units and their relation to each other, and to the
hierarchy of units, theoretical difficulties [which] are of course problems of the
subject in general, and are not specific to the book (Knowles 1995: 122). The relation of the turn (not part of Sinclair and Coulthards model) to the rest of the hierarchy is particularly problematic, as an exchange boundary may come in the
middle of a turn, and since a turn boundary may come in the middle of a move, for
instance if a speaker finishes off someone elses word in joint utterance production.
In his more recent work, Sinclair has introduced the notion of extended units
of meaning, first in the seminal paper Sinclair (1996), and more comprehensively
in Sinclair (2004). He presents the hypothesis that the notion of a linguistic item
can be extended so that units of meaning are expected to be largely phrasal
(Sinclair 1996: 82). Although the fixedness of word strings is most obvious in idiomatic/proverbial usage and phrasal verbs, Sinclairs model extends the concept of
phrasehood to words in general. Corpus evidence shows that each word in the language has its own unique lexico-grammatical profile and that different senses of
for instance a noun such as bank have their own distinctive sets of collocates. Sinclair argues that these observations force the linguist to reassess the notion of
meaning. His model includes structural categories with an increasing level of abstraction: collocation (a frequent co-occurrence of words, such as his example
naked eye), colligation (the co-occurrence of grammatical choices, e.g. choice of
preposition with/to in front of the naked eye), semantic preference (the collocation of a lexical item with items from a specific (more or less general) semantic
subset Bednarek 2008: 119) and semantic prosody (persistent collocation with
items with specific connotations; Louw 1993; cf. also Hunston 2009). Despite the
term and importantly for the current study Sinclair views semantic prosodies as
attitudinally significant and places them on the pragmatic side of the semantics/
pragmatics continuum (Sinclair 1996: 87). Indeed, speakers choice of a semantic
prosody is what links meaning to communicative purpose: Without it, the string
of words just means it is not put to use in a viable communication all subsequent choices within the lexical item relate back to the prosody (Sinclair 1996:
88). This can explain the pragmatic difference between semantically equivalent
sentences like I will make no promises and I will not make any promises. Hence,
Sinclairs analysis has an explanatory power to account for the creation of new
609
pragmatic expressions like make no mistake, which turns out strikingly more common than any other collocation of make no in a large corpus such as COCA. Sinclair (2004) develops the analytical model further by specifying the core (the invariable component of a lexical item) and the semantic prosody as the two
obligatory components of a lexical item, and the remaining three (collocation, colligation and semantic preference) as optional components, by relating the categories of co-selection to the paradigmatic and syntagmatic dimensions of language.
Furthermore, Stubbs (1996, 2001) is another linguist who argues that meaning
is a largely a phraseological phenomenon and whose corpus-based accounts also
contribute to the rigour and maturity of the field of corpus-based discourse analysis. Like the previous models, he draws on the British neo-Firthian tradition by
Halliday and Sinclair, but this account is more sociologically oriented than the
ones referred to above. It takes into account the status of speakers and listeners and
the importance of social institutions for the interpretation of the meaning of texts.
Stubbss linguistic analysis, largely based on collocation patterns, focuses on how
users lexis and grammatical choices (e.g. active vs. passive voice) contribute to
different meanings that can be interpreted in a social context.
The rapidly growing interest in corpus-based pragmatics is underlined by several collective volumes which for reasons of space have to be mentioned in passing. Adolphs (2008), del and Reppen (2008) and Biber et al. (2007) are significant contributions that are exclusively or predominantly corpus-based and which
address the main theoretical and methodological tenets of this field, while Baker
(2006) provides an enlightening evaluation of a variety of methodologies for corpus-based discourse analysis.
5.
Concluding remarks
The overview presented above has demonstrated the breadth and complexity of the
field of corpus-based pragmatics. Most studies consider overt and easily retrievable features such as discourse markers, connectives, hedges etc., and are formbased (cf. section 1). There are fewer studies where corpora are used to explore a
particular discourse function and its possible realisations, although such studies
appear to be on the increase. Relevant examples would be Deutschmanns (2006)
study which explores particular speech act function apologies and its various
realisations, and Stenstrm et al. (2002) who demonstrate how quotations are
marked by different structures, including BE + like, GO and SAY as reporting verbs,
voice modulation and zero-quotations.
There appears to be an ongoing increase in the different topics that are explored
by corpora. Schmid (2001) studies various patterns for expressing presuppositions.
Corpora are also used in the study of tropes such as metaphor (Sandford 2008) and
610
Gisle Andersen
hyperbole in recent contributions such as Norrick (2004) and McCarthy and Carter
(2004b), who argue that this approach can add substance to the already healthy
existing literature on rhetorical strategy, conventional and creative metaphor, irony
and over- and understatement. (2004b: 158). These contributions illustrate the
problem associated with one-to-one searching, as it assumes an a priori understanding of the forms that may realise a particular discourse function. This is especially the case in studies of creative language processes such as the use of tropes.
I have argued in favour of supplementing corpus-based pragmatics studies with
a corpus-driven approach. Corpus-based pragmatics studies require a hermeneutical and holistic approach to language data, because face-to-face interaction involves multiple layers of information. Video-based research methods are therefore
valuable. Modern multimodal corpora developed with state-of-the-art technology
provide a promising outlook for such studies in the future.
Notes
1. I am grateful to the editors and the anonymous reviewer for valuable comments on the
first draft of this paper.
2. This can for example be seen from the studies included in three recent special issues of
Journal of Pragmatics focusing on corpus linguistics (e.g. McCarthy and Carter 2004a),
of which a vast majority is based on English.
3. Another shortcoming of some corpora is that the data are only available as short snippets
of sound and not as longer stretches of spoken dialogue. For many research purposes, access to longer text passages is a prerequisite, e.g. studies of the structure and development
of narratives in conversation.
4. In the current account, I opt for the term discourse markers, which seems the most commonly used, especially in the most recent literature. For discussions on terminology,
cf. Brinton (1996); Andersen (2001), Fischer (2006) and Mosegaard Hansen (2006), who
argues that discourse particle is misleading on formal grounds.
References
Aarts, Jan and Willem Meijs (eds.)
1986
Corpus Linguistics II. New Studies in the Analysis and Exploitation of Computer Corpora. Amsterdam: Rodopi.
Abraham, Werner (ed.)
1991
Discourse Particles. Amsterdam: Benjamins.
del, Annelie and Randi Reppen (eds.)
2008
Corpora and Discourse: The Challenges of Different Settings. Amsterdam:
Benjamins.
Adolphs, Svenja
2008
Corpus and Context: Investigating Pragmatic Functions in Spoken Discourse.
611
Aijmer, Karin and Anne-Marie Simon-Vandenbergen (eds.)

2006
Pragmatic Markers in Contrast. Amsterdam: Elsevier.
Aijmer, Karin
1986
Why is actually so frequent in spoken English? In: Gunnel Tottie and Ingegerd
Bcklund (eds.), English in Speech and Writing: A Symposium, 119129.
Aijmer, Karin
1989
Aijmer, Karin
1996
Conversational Routines in English. London: Longman.
Aijmer, Karin
2002
English Discourse Particles: Evidence from a Corpus. Amsterdam: Benjamins.
Algeo, John
1988
British and American grammatical differences. International Journal of Lexicography 1: 131.
Allan, W. Scott
1990
The rise of New Zealand intonation. In: Allan Bell and Janet Holmes (eds.),
New Zealand Ways of Speaking English, 115128. Clevendon: Multilingual
Matters.
Altenberg, Bengt and Karin Aijmer
2000
The English-Swedish Parallel Corpus: a resource for contrastive research and
translation studies. In: Christian Mair and Marianne Hundt (eds.), Corpus Linguistics and Linguistic Theory, 1534. Amsterdam: Rodopi.
Altenberg, Bengt
1986
ICAME bibliography. ICAME News 10: 6279.
Altenberg, Bengt
1987
Prosodic patterns in spoken English. Studies in the Correlation between Prosody and Grammar for Text-To-Speech Conversion. Lund: Lund University
Press.
Altenberg, Bengt
1991
Amplifier collocations in spoken English. In: Stig Johansson and Anna-Brita
Stenstrm (eds.), English Computer Corpora, 127147. Berlin: Mouton de
Gruyter.
Altenberg, Bengt
1997
Exploring the Swedish component of the International Corpus of Learner English. In: Barbara Lewandowska-Tomaszcyk and Patrick James Melia (eds.),
PALC97: Practical Applications in Language Corpora, 119132. Ldz: Ldz
University Press.
Altenberg, Bengt
2006
The function of adverbial connectors in second initial position in English
and Swedish. In: Karin Aijmer and Anne-Marie Simon-Vandenbergen (eds.),
Pragmatic Markers in Contrast, 1138. Amsterdam: Elsevier.
Andersen, Gisle and Thorstein Fretheim
2000
612
Gisle Andersen
Andersen, Gisle
2001
Pragmatic Markers and Sociolinguistic Variation: a Relevance-Theoretic Approach to the Language of Adolescents. Amsterdam: Benjamins.
Andersen, Gisle
2005
Assessing algorithms for automatic extraction of anglicisms in Norwegian
texts. In: Proceedings from The Corpus Linguistics Conference Series, Vol. 1,
no. 1, Corpus Linguistics 2005.
Andersen, Gisle
2006
Engelske importord i norsk. Sprknytt 3/2006: 2732.
Andersen, Gisle
2009
Responses in academic and professional discourse. Paper presented at the
11th International Pragmatics Conference, Melbourne, 1217 July 2009.
Andersen, Gisle
2010
How to use corpus linguistics in sociolinguistics. To be published in Anne
OKeeffe and Michael McCarthy (eds.), Routledge Handbook of Corpus Linguistics, 547562. Londin: Routledge.
Aston, Guy and Lou Burnard
1998
The BNC Handbook. Exploring the British National Corpus with SARA. Edinburgh: Edinburgh University Press.
Baker, Paul
2006
Using Corpora in Discourse Analysis. London: Continuum.
Bald, Wolf-Dietrich
1990
An example of phonological reduction in English. In: S. Ramsaran (ed.),
Studies in the Pronunciation of English. A Commemorative Volume in Honour
of A.C. Gimson, 317322. London: Routledge.
Banjo, Ayo
1996
The sociolinguistics of English in Nigeria and the ICE project. In: Sidney
Greenbaum (ed.), Comparing English Worldwide: The International Corpus of
English, 239248. Oxford: Clarendon Press.
Bauer, Laurie and Janet Holmes
1996
Getting into a flap! /t/ in New Zealand English. World Englishes 15: 115124.
Beaugrande, Robert de
1999
Linguistics, sociolinguistics, and corpus lingistics: Ideal language versus real
language. Journal of Sociolinguistics, 3: 128139.
Bednarek, Monika
2008
and Lingustic Theory 4: 119139.
Biber, Douglas (ed.)
1994
Sociolinguistic Perspectives on Register. Oxford/New York: Oxford University Press.
1989
Styles of stance in English. Lexical and grammatical marking of evidentiality
and affect. Text 9: 93124.
Biber, Douglas, Ulla Connor and Thomas A. Upton
2007
Discourse on the Move: Using Corpus Analysis to Describe Discourse Structure. Amsterdam: Benjamins.
613
Biber, Douglas
1986
Spoken and written textual dimensions in English: Resolving the contradictory
findings. Language 62: 384414.
Biber, Douglas
1987
A textual comparison of British and American writing. American Speech 62:
99119.
Biber, Douglas
1988
Press.
Biber, Douglas
1993
An analytical framework for register studies. In: Douglas Biber and Edward
Finegan (eds.), Sociolinguistic Perspectives on Register, 3156. New York:
Biber, Douglas
1995
Dimensions of Register Variation. A Cross-Linguistic Comparison. Cambridge: Cambridge University Press.
Biber, Douglas
2009
A corpus-driven approach to formulaic language in English: Multi-word patterns in speech and writing. International Journal of Corpus Linguistics 14.3:
275311.
Bjrge, Anne-Kari
2009
Conflict or cooperation: the use of backchannelling in ELF negotiations. English for Specific Purposes 29: 191203.
Brazil, David
1982
Impromptuness and intonation. In: Nils Erik Enkvist (ed.), Impromptu Speech:
Brinton, Laurel J.
1996
Pragmatic Markers in English: Grammaticalization and Discourse Functions.
Brodda, Benny
1988
Tracing turns in the London-Lund Corpus with BetaText. Literary and Linguistic Computing 3: 94104.
Bublitz, Wolfram
1979
Tag questions, transformational grammar, and pragmatics. Papers and Studies
in Contrastive Linguistics 9: 521.
Bublitz, Wolfram
1988
Supportive Fellow-Speakers and Cooperative Conversations. Amsterdam:
Benjamins.
Chafe, Wallace L
1992
In: Jan Svartvik (ed.), Directions in Corpus Linguistics. Proceedings of Nobel
Symposium 82, 7997. Berlin: Mouton de Gruyter.
Chafe, Wallace L., John W. Du Bois and Sandra A. Thompson
1991
Towards a new corpus of Spoken American English. In: Karin Aijmer and
Bengt Altenberg (eds.), English Corpus Linguistics. Studies in Honour of Jan
Svartvik, 6482. London: Longman.
614
Gisle Andersen
Chafe, Wallace
1992
In: Jan Svartvik (ed.), Directions in Corpus Linguistics. Proceedings of
Nobel Symposium 82 Stockholm, 48 August 1991, 7997. Berlin: Mouton de
Gruyter.
Chomsky, Noam
1957
Syntactic Structures. The Hague: Mouton.
Coates, Jennifer
1987
Epistemic modality and spoken discourse. Transactions of the Philological Society 85(1): 110131.
Collins, Peter
2005
It-clefts and wh-clefts: Prosody and pragmatics. Journal of Pragmatics: 38:
17061720.
Cook, Guy
1990
Transcribing Infinity: Problems of Context Presentation. Journal of Pragmatics 14: 124.
Couper-Kuhlen, Elizabeth and Cecilia E. Ford (eds.)
2004
Sound Patterns in Interaction: Cross-linguistic Studies from Conversation.
Couper-Kuhlen, Elizabeth and Margret Selting (eds.)
1996
Prosody in Conversation. Cambridge: Cambridge University Press.
1996
The prosody of repetition: on quoting and mimickry. In: Elizabeth CouperKuhlen and Margret Selting (eds.), Prosody in Conversation: Interactional
Studies, 366405. Cambridge: Cambridge University Press.
Cresti, Emanuela and Massimo Moneglia (eds.)
2005
C-ORAL-ROM: Integrated Reference Corpora for Spoken Romance Languages. Amsterdam: Benjamins.
Croft, William
1995
Intonation units and grammatical structure. Linguistics 33: 839882.
Crowdy, Steve
1995
The BNC spoken corpus. In: Geoffrey Leech, Greg Myers and Jenny Thomas
(eds.), Spoken English on Computer. Transcription, Mark-up and Application.
Cruttenden, Alan
1997
Intonation. Cambridge: Cambridge University Press.
Crystal, David and Randolph Quirk
1964
Systems of Prosodic and Paralinguistic Features in English. The Hague: Mouton.
Crystal, David
1972
The intonation system of English. In: Dwight Bolinger (ed.), Intonation,
110136. Harmondsworth: Penguin.
Crystal, David
1975
The English Tone of Voice. Essays on Intonation, Prosody and Paralanguage.
London: Arnold.
Davidson, Benjamin D.
1977
Aspects of information structuring in modern spoken English. In: Wolf-Diedrich Bald and Robert Ilson (eds.), Studies in English Usage: The Resources of
615
Present-day English Corpus for Linguistic Analysis, 157170. Frankfurt am

Main: Peter Lang.
Davies, Mark
2009
The 385+ Million Word Corpus of Contemporary American English (1990
2008+): Design, Architecture, and Linguistic Insights. International Journal
of Corpus Linguistics 14: 15990.
Deutschmann, Mats
2006
Social variation in the use of apology formulae in the British National Corpus.
In: Antoinette Renouf and Andrew Kehoe (eds.), The Changing Face of Corpus Linguistics, 205222. Amsterdam/New York: Rodopi.
Downing, Angela
2006
The English pragmatic marker surely and its functional counterparts in Spanish. In: Karin Aijmer and Anne-Marie Simon-Vandenbergen (eds.), Pragmatic
Edwards, Jane A. and Martin D. Lampert
1993
Talking Data: Transcription and Coding in Discourse Research. Hillsdale:
Lawrence Erlbaum.
Eggins, Suzanne
1996
Analysing Casual Conversation. London: Continuum.
Enkvist, Nils Erik (ed.)
1982
Impromptu Speech: a Symposium. bo: bo Akademi.
Enkvist, Nils Erik
1973
Linguistic stylistics. Janua Linguanun Series Critica 5. The Hague: Mouton.
Erman, Britt
1986
Some pragmatic expressions in English conversation. In: Gunnel Tottie and
Ingegerd Bcklund (eds.), English in Speech and Writing: a Symposium, 131
Erman, Britt
1987
Pragmatic Expressions in English: a Study of You Know, You See, and I Mean
in Face-To-Face Conversation. Stockholm: Almqvist and Wiksell.
Erman, Britt
1997
Guys just such a dickhead; the context and function of just in teenage talk.
In: Ulla-Brite Kostsinas, Anna-Brita Stenstrm and Anna-Malin Karlsson
(eds.), Ungdomssprk i Norden, 8295. Stockholm: Institutionen fr nordiska
sprk.
Faber, Dorrit and Karen M. Lauridsen
1991
The compilation of a Danish-English-French corpus in contract law. In: Stig
Johansson and Anna-Brita Stenstrm (eds.), English Computer Corpora. Selected Papers and Research Guide, 235243. Berlin: Mouton de Gruyter.
1982
Phatic, metalingual and metacommunicative functions in discourse. Gambits
and repairs. In: Nils Erik Enkvist (ed.), Impromptu Speech: a Symposium,
Fang, Alex C.
1993
Building a corpus of computer science. In: Jan Aarts, Pieter de Haan and Nelleke Oostdijk (eds.), English Language Corpora: Design, Analysis and Exploitation. 7378. Amsterdam: Rodopi.
616
Gisle Andersen
Fill, Alwin
1992
Joking in English and German: A contrastive study. In: Christian Mair and
Manfred Marcus (eds.), New Departures in Contrastive Linguistics. Vol. 2,
2131. Innsbruck: Institut fr Anglistik, Universitt Innsbruck.
Firth, John R
1935
The technique of semantics. Reprinted in 1957 in Papers in Linguistics,
193451,723. Oxford: Oxford University Press.
Fischer, Kerstin
2000
From Cognitive Semantics to Lexical Pragmatics: The Functional Polysemy of
Discourse Particles. Berlin: Mouton de Gruyter.
Fischer, Kerstin
2006
Approaches to Discourse Particles. Amsterdam: Elsevier.
Fletcher, Janet, Esther Grabe and Paul Warren
2004
Intonational variation in four dialects of English: the high rising tune. In:,
Sun-Ah Jun (ed.) Prosodic Typology. The Phonology of Intonation and Phrasing, 390410. Oxford: Oxford University Press.
Foolen, Ad
2006
Polysemy Patterns in Contrast: The Case of Dutch toch and German doch. In:
Karin Aijmer and Anne-Marie Simon-Vandenbergen (eds.), Pragmatic Markers
in Contrast, 5972. Amsterdam: Elsevier.
Francis, W. Nelson
1967
The Brown University Standard Corpus of English: Some implications for
TESOL. In: Betty Wallace Robinett (ed.), On Teaching English to Speakers of
Other Languages, 13135. Washington, D.C.: TESOL.
Fretheim, Thorstein
1981
Ego-dempere og alter-dempere. Maal og minne 12: 86100.
Furiasso, Cristiano and Knut Hofland
2007
The retrieval of false anglicisms in newspaper texts. In: Roberta Facchinetti
(ed.), Corpus Linguistics 25 Years on, 347364. Amsterdam/New York: Rodopi.
Gardner, Rod
1998
Between speaking and listening: the vocalisation of understandings. Applied
Linguistics 19.2: 204224.
Garnham, Alan
1986
Review of R. Wardhaugh, How Conversation Works. Journal of Literary Semantics 15: 153154.
Garnham, Alan, Richard Shillcock, Gordon Brown, Andrew Mill and Anne Cutler
1981
Slips of the tongue in the London-Lund Corpus of spontaneous conversation.
Geluykens, Ronald
1986
Questioning intonation: An empirical study into the prosodic feature rising
intonation and its relevance for the production and recognition of questions.
Antwerp Papers in Linguistics 48.
Geluykens, Ronald
1987a
Intonation and speech act type: An experimental approach to rising intonation
in declaratives. Journal of Pragmatics 11: 483494.
617
Geluykens, Ronald
1987b
Tails (right-dislocations) as a repair mechanism in English conversation. In:
Jan Nuyts and Georges de Schutter (eds.), Getting Ones Words into Line: On
Word Order and Functional Grammar, 119129. Dordrecht: Fortis.
Geluykens, Ronald
1988
On the myth of rising intonation in polar questions. Journal of Pragmatics 12:
467485.
Granger, Sylviane
1997
On identifying the syntactic and discourse features of participle clauses in academic English: native and non-native writers compared. In: Jan Aarts, Inge
de Mnnink and Herman C. Wekker (eds.), Studies in English Language and
Teaching in Honour of Flor Aarts, 185198. Amsterdam/Atlanta: Rodopi.
Granger, Sylviane and Stephanie Tyson
1996
Connector usage in the English essay writing of native and non-native EFL
speakers of English. World Englishes 15: 1727.
Greenbaum, Sidney (ed.)
1996
Comparing English Worldwide: The International Corpus of English. Oxford:
Clarendon Press.
Greenbaum, Sidney
1991
ICE: the International Corpus of English. English Today 28: 37.
Greenbaum, Sidney
1992
A new corpus of English: ICE. In: Jan Svartvik (ed.), Directions in Corpus Linguistics. Proceedings of Nobel Symposium 82, 171179. Berlin: Mouton de
Gruyter.
Gustafsson, Marita
1982
Textual aspects of topicalization in a corpus of English. ICAME News 6:
4676.
Gustafsson, Marita
1983
Fronting of adverbials in four genres of English. In: Sven Jacobson (eds.),
Papers from the Second Scandinavian Symposium on Syntactic Variation,
Gustafsson, Marita
1985
Topicalizations revisited. In: Jorma Tommola and T. Vinanen (eds.), Working
Papers in English Studies 6, 4350. Publications of the Department of English,
Turku: University of Turku.
Guy, Gregory and Julia Vonwiller
1989
The meaning of an intonation in Australian English. Australian Journal of Linguistics 4: 117.
Haan, Ferdinand de
2002
Strong modality and negation in Russian. In: Randi Reppen, Susan M.
Fitzmaurice and Douglas Biber (eds.), Using Corpora to Explore Linguistic
Variation, 91110. Amsterdam: Benjamins.
Haegeman, Lillian
1984
Pragmatic conditionals in English. Folio Linguistica 13: 485502.
1961
Categories of the theory of grammar. Word 17: 24192.
618
Gisle Andersen

1985
An Introduction to Functional Grammar. London: Edward Arnold.
Hardy, Donald E. and Amy Leuchtmann
1996
Topic versus cohesion in the prediction of causal ordering in English conversation. Discourse Processes 21: 237254.
Haslerud, Vibeke and Anna-Brita Stenstrm
1995
The Bergen Corpus of London Teenager Language (COLT). In: Geoffrey
Leech, Greg Myers and Jenny Thomas (eds.), Spoken English on Computer.
Transcription, Mark-up and Application, 235242. London: Longman.
Hasselgrd, Hilde
2006
Not now on Non-Correspondence between the Cognate Adverbs now and
n. In: Karin Aijmer and Anne-Marie Simon-Vandenbergen (eds.), Pragmatic
Hedstrm, Karl Erik
1984
A study of repairs in speech. In: Stockholm Papers in English Language and
Literature 4, 69101. Stockholm: Department of English, Stockholm University.
Heinz, Bettina
2003
Backchannel responses as strategic responses in bilingual speakers conversations. Journal of Pragmatics 35: 11131142.
Hofland, Knut and Stig Johansson
1986
The Tagged LOB Corpus: KWIC Concordance. Microfiche. Bergen: Norwegian Computing Centre for the Humanities.
Holmes, Janet
1988
Doubt and certainty in ESL textbooks. Applied Linguistics 9: 2144.
Hundt, Marianne, Carolin Biewer and Nadja Nesselhauf
2007
Corpus Linguistics and the Web. Amsterdam: Rodopi.
Hunston, Susan
2009
Semantic prosody revisited. In: Rosamund Moon (ed.), Words, Grammar,
Text Revisiting the Work of John Sinclair, 85103. Amsterdam: Benjamins.
Johannessen, Janne B. and Kristin Hagen (eds.)
2008
Sprk i Oslo: Ny forskning omkring talesprk. Oslo: Novus.
Johansson, Stig and Knut Hofland
1994
Towards an English-Norwegian parallel corpus. In: Udo Fries, Gunnel Tottie
and Peter Schneider (eds.), Creating and Using English Language Corpora,
Johansson, Stig
2006
How well can well be translated? On the English discourse particle well and its
correspondences in Norwegian and German. In: Karin Aijmer and Anne-Marie
Simon-Vandenbergen (eds.), Pragmatic Markers in Contrast, 115138. Amsterdam: Elsevier.
Jucker, Andreas and Sara Smith
1998
And people just you know like Wow: discourse markers as negotiating strategies. In: Andreas Jucker and Yael Ziv (eds.), Discourse Markers: Descriptions and Theory, 171201. Amsterdam: Benjamins.
Jucker, Andreas and Yael Ziv (eds.)
1998
Discourse Markers: Descriptions and Theory. Amsterdam: Benjamins.
619
Jucker, Andreas
1993
The discourse marker well: a relevance-theoretical account. Journal of Pragmatics 19.5: 43552.
Kendon, Adam
1972
Some relationships between body motion and speech: An analysis of an
example. In: Aron Siegman and Benjamin Pope (eds.), Studies in Dyadic Communication, 177216. New York: Pergamon Press.
Kendon, Adam
1980
Gesticulation and speech: Two aspects of the process of utterance. In: Mary
Ritchie Key (ed.), The Relationship of Verbal and Nonverbal Communication,
Kennedy, Graeme
1998
An Introduction to Corpus Linguistics. London: Addison Wesley Longman.
Kilgarriff, Adam and Grefenstette, Gregory
2003
Introduction to the Special Issue on Web as Corpus. Computational Linguistics
29: 115.
Kirk, John
2002
Applications of the Stenstrm model of discourse structure. In: Angela Hasselgren and Leiv Egil Breivik (eds.), From the COLTs Mouth and Others,
Kjellmer, Gran
2009
Where do we backchannel? On the use of mm, mhm, uh huh and such like. International Journal of Corpus Linguistics 14.1: 81112.
Knowles, Gerry and Lita Lawrence
1987
Automatic intonation assignment. In: Roger Garside, Geoffrey Leech and
Geoffrey Sampson (eds.), The Computational Analysis of English. A Corpusbased Approach, 139148. London: Longman.
Knowles, Gerry, Briony Williams and Lita Taylor (eds.)
1996
A Corpus of Formal British English Speech. The Lancaster/IBM Spoken-English Corpus. London: Longman.
Knowles, Gerry, Lita Taylor and Briony Williams
1992
A Corpus of Formal British English Speech. London: Longman.
Knowles, Gerry
1995
Review of Stenstrm 1994. ICAME Journal 19: 120123.
Knowles, Gerry
1986
Tone grouping by numbers. In: Lancaster Papers in Linguistics 44. Lancaster:
Department of Linguistics and Modern English Language, University of Lancaster.
Kussmaul, Paul
1978
In fact, actually, anyway : Indikatoren von Sprechakten im informellen gesprochenen Englisch. Die Neueren Sprachen 27: 357369.
Kyt, Merja and Matti Rissanen
1988
The Helsinki Corpus of English Texts: Classifying and coding the diachronic
part. In: Merja Kyt, Ossi Ihalainen and Matti Rissanen (eds.), Corpus Linguistics Hard and Soft, 169179. Amsterdam: Rodopi.
Langford, David
1994
Analysing Talk. London: Macmillan.
620
Gisle Andersen
Leech, Geoffre
1992
Corpora and theories of linguistic performance. In: Jan Svartvik (ed.), Directions in Corpus Linguistics. Proceedings of Nobel Symposium 82, 105122.
1995
Spoken English on Computer. London: Longman.
Lenk, Uta
1998
Marking Discourse Coherence: Functions of Discourse Markers in Spoken
English. Tbingen: Narr.
Lindquist, Hans and Christian Mair (eds.)
2004
Corpus Approaches to Grammaticalization in English. Amsterdam: John Benjamins.
Loehr, Daniel P.
2004
Gesture and Intonation. PhD. Dissertation, Georgetown University, Washington, DC.
Loman, Bengt
1982
The segmentation problem in the study of impromptu speech. In: Nils Erik
Enkvist (ed.), Impromptu Speech: a Symposium,105130. bo: bo Akademi.
Louw, Bill
1993
Text and Technology. In Honour of John Sinclair, 157176. Amsterdam: Benjamins.
Mauranen, Anna
2004
Theyre a Little Bit Different Observations on Hedges in Academic Talk.
In: Karin Aijmer and Anna-Brita Stenstrm (eds.), Discourse Patterns in
Spoken and Written Corpora, 173197. Amsterdam: Benjamins.
Mauro, Tullio de, Federico Mancini, Massimo Vedovelli and Miriam Voghera
1993
Lessico di Frequenza dellItaliano Parlato. Milan: Etas.
McCarthy, Michael and Ronald Carte
2004a
Introduction. Journal of Pragmatics 36: 147148.
McCarthy, Michael and Ronald Carter
2004b
Theres millions of them: hyperbole in everyday conversation. Journal of
McCarthy, Michael
2002
Good listenership made plain: British and American non-minimal response
tokens in everyday conversation. In: Randi Reppen, Susan M. Fitzmaurice and
Douglas Biber (eds.), Using Corpora to Explore Linguistic Variation, 7390.
McCarthy, Michael
2003
Talking Back: Small Interactional Response Tokens in Everyday Conversation. Research on Language and Social Interaction 36.1: 3363.
McEnery, Tony and Andrew Wilson
2001
Corpus Linguistics. 2nd ed. Edinburgh: Edinburgh University Press.
McEnery, Tony and Richard Xiao
2004
The Lancaster Corpus of Mandarin Chinese: A Corpus for Monolingual and
Contrastive Language Study. Paper presented at LREC 2004. May 2004, Lisbon.
621
McEnery, Tony, Richard Xiao and Yukio Tonio

2006
Corpus-based Language Studies: An Advanced Resource Book. New York:
Routledge.
McNeill, David
1985
So you think gestures are nonverbal? Psychological Review 92.3: 350371.
McNeill, David
1992
Hand and Mind: What Gestures Reveal about Thought. Chicago: University of
Chicago Press.
Mey, Jacob L.
2004
FoI Corpus Linguistics Part III Introduction. Journal of Pragmatics 36: Editorial. 1519.
Milroy, Leslie
1987
Observing and Analysing Natural Language. Oxford: Blackwell.
Ming, Tao and Liang Chen
2010
A discourse-pragmatic study of the word order variation in Chinese relative
clauses. Journal of Pragmatics 42: 168189.
1998
The Function of Discourse Particles. Amsterdam: Benjamins.
2006
A dynamic polysemy approach to the lexical semantics of discourse markers.
In: Kerstin Fischer (ed.), Approaches to Discourse Particles, 2141. Oxford:
Elsevier.
Nsslin, Siv
1984
The English Tag Question: A Study of Sentences Containing Tags of the Type
Isnt It?, Is It?. Stockholm: University of Stockholm.
Nevalainen, Terttu
1987a
A quantitative survey of nuclear tone variation in English. In: Proceedings of
the Eleventh International Congress of Phonetic Sciences. Vol 3, 180183.
Tallin: Institute of Language and Literature, Academy of Sciences of the Estonian S.S.R.
Nevalainen, Terttu
1987b
Adverbial focusing and intonation. Lingua 73: 141165.
Nevalainen, Terttu
1992
Intonation and discourse type. Text 12: 397427.
Norrick, Neal and Kerstin Fischer
2009
Listener activities and responses: Panel Introduction at the 11th International
Pragmatics Conference, Melbourne, 1217 July 2009.
Norrick, Neal
2004
Hyperbole, extreme case formulation. Journal of Pragmatics 36.9: 1727
1739.
Ochs, Elinor
1996
Linguistic resources for socializing humanity. In: John J. Gumperz and
Stephen Levinson (eds.), Rethinking Linguistic Relativity, 407438. Cambridge: Cambridge University Press.
Orestrm, Bengt
1977
Why /dhi/ Book? SSE Report, Survey of Spoken English, Lund University.
622
Gisle Andersen
Orestrm, Bengt
1982
When is it my turn to speak?. In: Nils Erik Enkvist (ed.), Impromptu Speech:
Orestrm, Bengt
1983
Turn-taking in English Conversation. Lund: Liber/Gleerup.
stman, Jan-Ola
1982
The symbiotic relationship between pragmatic particles and impromptu
speech. In: Nils Erik Enkvist (ed.), Impromptu Speech: a Symposium, 147
stman, Jan-Ola
1995
Pragmatic particles twenty years after. In: Brita Wrvik, Sanna-Kaisa, Hiltunen, and Risto Hiltunen (eds.), Organization in Discourse: Proceedings from
the Turku Conference, 95108. Turku: University of Turku.
Partington, Alan
1998
Patterns and Meanings: Using Corpora for English Language Research and
Teaching. Amsterdam: Benjamins.
Raymond, William D., Mark Pitt, Keith Johnson, Elizabeth Hume, Matthew Makashay,
Robin Dautricourt and Craig Hilts
2002
An Analysis of Transcription Consistency in Spontaneous Speech from the
Buckeye Corpus. ICSLP-02, Denver.
Renouf, Antoinette
1986
The exploitation of a computerised corpus of English text. In: Michele Rivas
(ed.), Actes du VIIIeme Colloque G.E.R.A.S, 123136. Paris: Universite de
Paris-Dauphine.
Renouf, Antoinette
2007
Corpus development 25 years on: from super-corpus to cyber-corpus. In: Roberta Facchinetti (ed.), Corpus Linguistics 25 Years on, 2750. Amsterdam/
New York: Rodopi.
Renouf, Antoinette and Andrew Kehoe (eds.)
2006
The Changing Face of Corpus Linguistics. Amsterdam: Rodopi.
Rissanen, Matti
1986
Variation and the study of English historical syntax. In: David Sankoff (ed.),
Diversity and Diachrony. 97109. Amsterdam: Benjamins.
Romero, Jesus
1994
Ahm, ehm you call it theme? A thematic approach to spoken English. Journal of Pragmatics 22: 495509.
Sajavaara, Kari
1982
Contrastive analysis and impromptu speech. In: Nils Erik Enkvist (ed.), Impromptu Speech: a Symposium, 195206. bo: bo Akademi.
Sandford, Daniel
2008
Discourse and metaphor: A corpus-driven inquiry. Corpus Linguistics and
Lingustic Theory 4: 209234.
Schelfhout, Carla, Petro-Arno Coppen, Nelleke Ostdijk and Frans Van Der Slik
2005
Interjections in Dutch: a corpus-based approach, available at
http://www.neerlandistiek.nl/publish/articles/000100/article.pdf.
Schmid, Hans-Jrg
2001
Presupposition can be a bluff: How abstract nouns can be used as presupposition triggers. Journal of Pragmatics. 33.10: 15291552.
623
Schneider, Stefan
2007
Reduced Parenthetical Clauses as Mitigators. Amsterdam: Benjamins.
2001
Studies in Interactional Linguistics. Amsterdam: John Benjamins.
Shastri, S.V.
1988
The Kolhapur Corpus of Indian English and work done on its basis so far.
ICAME Journal 12: 1526.
Sinclair, John M. (ed.)
1987
Looking up. An Account of the COBUILD Project in Lexical Computing. London: Collins ELT.
Sinclair, John M.
1966
Indescribable English. Inaugural lecture, University of Birmingham, 1966
(unpublished).
Sinclair, John M.
1991
Corpus, Concordance, Collocation. Oxford: Oxford University Press.
Sinclair, John M.
1996
The search for units of meaning. Textus 9: 75106.
Sinclair, John M.
2004
Trust the Text: Language, Corpus and Discourse. London: Routledge.
1975
Towards an Analysis of Discourse: the English used by Teachers and Pupils.
1992
Towards an analysis of discourse. In: Malcolm Coulthard (ed.), Advances in
Spoken Discourse Analysis, 134. London: Routledge.
Sinclair, John M. and David Brazil
1982
Teacher Talk. Oxford: Oxford University Press.
1986
1995
Relevance: Communication and Cognition. 2nd ed. Oxford: Blackwell.
Stenstrm, Anna-Brita and Annette Myre Jrgensen
2009
Youngspeak in a Multilingual Perspective. Amsterdam: Benjamins.
Stenstrm, Anna-Brita, Gisle Andersen and Ingrid Kristine Hasund
2002
Trends in Teenage Talk: Corpus Compilation, Analysis and Findings. Amsterdam: Benjamins.
Stenstrm, Anna-Brita, Gisle Andersen, Kristine Hasund, Kristine Monstad and Hanne Aas
1998
Users Manual to Accompany The Bergen Corpus of London Teenage Language (COLT). Bergen: Department of English at the University of Bergen.
1982
Feedback. In: Nils Erik Enkvist (ed.), Impromptu Speech: a Symposium,
1984
Questions and Responses in English Conversation. Malmo: Gleerup/Liber.
1986a
A study of pauses as demarcators in discourse and syntax. In: Jan Aarts and
Willem Meijs (eds.), Corpus Linguistics II. New Studies in the Analysis and
Exploitation of Computer Corpora, 203218. Amsterdam: Rodopi.
624
Gisle Andersen
1986b
What does really really do? Strategies in speech and writing. In: Gunnel Tottie
and Ingegerd Bcklund (eds.), English in Speech and Writing: A Symposium,
1988
Adverbial commas and prosodic segmentation. In: Merj Kyt, Ossi Ihalainen and Matti Rissanen (eds.), Corpus Linguistics Hard and Soft, 221236.
Amsterdam: Rodopi.
1990a
Adverbial commas and prosodic segmentation. In: Jan Svartvik (ed.), The London-Lund Corpus of Spoken English: Description and Research, 253266.
(Lund Studies in English 82.) Lund: Lund University Press.
1990b
Pauses in monologue and dialogue. In: Jan Svartvik (ed.), The London-Lund
Corpus of Spoken English: Description and Research. 211252. (Lund Studies
in English 82.) Lund: Lund University Press.
1991
Expletives in the London-Lund Corpus. In: Karin Aijmer and Bengt Altenberg (eds.), English Corpus Linguistics. Studies in Honour of Jan Svartvik,
1994
An Introduction to Spoken Interaction. London: Longman.
1995
Taboos in teenage talk. In: Gunnell Melchers and Beatrice Warren (eds.),
Studies in Anglistics, 7179. Stockholm: Almqvist and Wiksell.
2006
Taboo words in teenage talk: London and Madrid girls conversations compared. In: Clare Mar-Molinero and Miranda Stewart (eds.), Language Variation and Change, 115138. Amsterdam: Benjamins.
Stubbs, Michael
1984
Review of Sinclair and Brazil 1982. Applied Linguistics 5: 7174.
Stubbs, Michael
1996
Text and Corpus Analysis. London: Blackwell.
Stubbs, Michael
2001
Words and Phrases: Corpus Studies of Lexical Semantics. Oxford: Blackwell.
Svartvik, Jan (ed.)
1992
Directions in Corpus Linguistics. Proceedings of Nobel Symposium 82 Stockholm, 48 August 1991. Berlin: Mouton de Gruyter.
Svartvik, Jan and Randolph Quirk (eds.)
1980
A Corpus of English Conversation. Lund: Gleerup/Liber.
Svartvik, Jan
1980
Well in conversation. In: Sydney Greenbaum, Geoff Leech and Jan Svartvik
(eds.), Studies in English Linguistics, 167177. London: Longman.
Svartvik, Jan
1982
The segmentation of impromptu speech. In: Nils Erik Enkvist (ed.), Impromptu
Speech: a Symposium, 131145. bo: bo Akademi.
625
Svartvik, Jan
1987
Graphic English prosody. In: Ishrat Lindblad and Magnus Ljung (eds.), Proceedings from the Third Nordic Conference for English Studies, 335346.
Svartvik, Jan
1988
A computerized corpus of spoken English. In: Vinca Liedloff (ed.), Standardization in Computerized Lexicography, 3352. Saarbrcken: Institut der Gesellschaft zur Frderung der angewandten Informationsforschung e.V. an der
Universitt des Saarlandes.
1999
youth. Journal of Sociolinguistics 3: 147172.
Taylor, Lita and Gerry Knowles
1988
Progress report on the Spoken English Corpus. In: Merja Kyt, Ossi Ihalainen and Matti Rissanen (eds.), Corpus Linguistics Hard and Soft, 237244.
Amsterdam: Rodopi.
Thavenius, Cecilia
1982
Exophora in English conversation: a study of third person pronominal reference. In: Nils Erik Enkvist (ed.), Impromptu Speech: a Symposium, 291305.
bo: bo Akademi.
Thomas, Jenny and Andrew Wilson
1996
Methodologies for studying a corpus of doctor-patient interaction. In: Jenny
Thomas and Mick Short (eds.), Using Corpora for Language Research, 92
109. London Longman.
2001
Tottie, Gunnel
1986
The importance of being adverbial. Adverbials of focusing and contingency in
spoken and written English. In: Gunnel Tottie and Ingegerd Bcklund (eds.),
English in Speech and Writing: A Symposium, 93118. Stockholm: Almqvist
and Wiksell.
Tottie, Gunnel and Carita Paradis
1982
From function to structure: Some pragmatic determinants of syntactic frequencies in impromptu speech. In: Nils Erik Enkvist (ed.), Impromptu Speech: a
Symposium, 307317. bo: bo Akademi.
Tottie, Gunnel
1991
Conversational style in British and American English: The case of backchannels. In: Karin Aijmer and Bengt Altenberg (eds.), English Corpus Linguistics.
Studies in Honour of Jan Svartvik, 254271. London: Longman.
Traugott, Elizabeth Closs and Bernd Heine (eds.)
1991
Van den Heuvel, Theo
1988
TOSCA: An aid for building syntactic databases. Literary and Linguistic Computing 3: 147151.
Venditti, Jennifer J. and Julia Hirschberg
2003
Intonation and discourse processing. In: Maria Josep Sol, Daniel Recasens
and Joaqun Romero (eds.), Proceedings of the 15th International Congress of
Phonetic Sciences, 107114. Barcelona, Spain.
626
Gisle Andersen
Verschueren, Jef
2009
Introduction: The pragmatic perspective. In: Jef Verschueren and Jan-Ola stman (eds.), Key Notions for Pragmatics, 127. Amsterdam: Benjamins.
Vicher, Anne and David Sankoff
1989
The emergent syntax of pre-sentential turn openings. Journal of Pragmatics
13: 8197.
Virtanen, Tuija
2009
Discourse linguistics meets corpus linguistics: theoretical and methodological
issues in the troubled relationship. In: Antoinette Renouf and Andrew Kehoe
(eds.), Corpus Linguistics, Refinements and Reassessments, 4966. Amsterdam: Rodopi.
Walker, Gareth
2004
The phonetic design of turn endings, beginnings, and continuations in conversation. PhD thesis, Department of Language and Linguistic Science, University of York.
Walsh, Steve, Anne OKeeffe and Michael McCarthy
2008
post-colonialism, multi-culturalism, structuralism, feminism, post-modernism and so on and so forth: A comparative analysis of vague category
markers in academic discourse. In: Annelie del and Randi Reppen (eds.),
Corpora and Discourse: The Challenges of Different Settings, 930. Amsterdam: Benjamins.
Waltereit, Richard
2006
The rise of discourse markers in Italian: a specific type of language change. In:
Kerstin Fischer (ed.), Approaches to Discourse Particles, 6176. Amsterdam:
Elsevier.
Wardhaugh, Ronald
1985
How Conversation Works. Oxford: Basil Blackwell, London: Andre Deutsch.
Westney, Paul
1986
How to be more or less certain in English: Scalarity in epistemic modality. International Review of Applied Linguistics 24: 311320.
Wichmann, Anne
1994
F0 troughs and prosodic phrasing: an investigation into the linguistic information contained in a speakers base-line when reading. Hermes 13:
205212. Aarhus School of Business.
Wichmann, Anne
1996
Prosodic style: a corpus-based approach. In: Gerry Knowles, Anne Wichmann
and Peter Alderson (eds.), Working with Speech. Perspectives on Research into
the Lancaster/IBM Spoken English Corpus, 168188. London/New York:
Longman.
Wichmann, Anne
2000
Intonation in Text and Discourse. London: Longman.
Wichmann, Anne
2004
The intonation of Please-requests: a corpus-based study. Journal of Pragmatics 36: 15211549.
Wichmann, Anne
2005
Please: from courtesy to appeal: the role of intonation in the expression of attitudinal meaning. English Language and Linguistics 9.2: 229253.
627
Wichmann, Anne
2007
Corpora and spoken discourse. In: Roberta Facchinetti (ed.), Corpus Linguistics 25 Years on, 7388. Amsterdam/New York: Rodopi.
Wierzbicka, Anna
1991
Cross-cultural Pragmatics: The Semantics of Human Interaction. Berlin:
Mouton de Gruyter.
Yngve, Victor
1970
On getting a word in edgewise. In: Papers from the 6th Regional Meeting, Chicago Linguistic Society, 567577. Chicago: Chicago Linguistic Society.
22.

Christoph Rhlemann
Introduction
Due to the massive dependence of pragmatic phenomena on context, corpora, as a
unfit for use in pragmatic research. Nonetheless, corpus linguistic analyzes, both
into key pragmatic phenomena. The aim of this paper is to illustrate key quantitative corpus studies into phenomena of pragmatic interest. The paper is divided
into six sections. The opening section addresses the question of how context-sensitive corpora are. Section 2 presents a case study into semantic prosody, an attitudinal phenomenon at the semantics/pragmatics interface. Section 3 presents a
relevance-theoretic study of the pragmatic marker like. Section 4 is concerned with
quantitative studies on reference. Section 5 introduces corpus research into speech
acts. The concluding section looks to the future, outlining recent attempts at building multimodal corpora.
1.
Pragmatics and quantitative corpus linguistics: a troubled

relationship?
Pragmatics is concerned with meaning in context. Because speakers can mean

more than they say, pragmatics is the art of the analysis of the unsaid (Mey 1991:
245). The relationship between pragmatics, thus understood, and corpus linguistics
is seen by some as a troubled one. The reason is simple: corpora record text, not
meaning, and they record context only crudely, particularly in spoken corpora.
Spoken corpora are based on transcriptions made from audio recordings. Since
tape recorders cannot filter out non-speech noises, thus selectively listening to
speech as humans can many, and often large, stretches of corpus transcripts
may be inaudible or unintelligible. Further, the information provided in spoken
corpora about the contexts in which the spoken texts were produced is fairly minimal; we get to know who (sex, age, class, etc.) was talking to whom when. Moreover, other types of context are recorded only abstractly: we learn in what kinds of
setting the talk occurred, or what type of interaction it was. Yet another type of abstract context, which is considered so crucial that in most large corpora the data
are categorized according to it, is type of situation, or register, such as academic
writing, fiction, and conversation. The types of context largely missing from
630
Christoph Rhlemann
spoken corpora include the almost infinite wealth of concrete situational, nonverbal, and social context that conversationalists in their specific contexts of situation
are connected to (cf. Cook 1990). In sum, we may have impoverished textual evidence, and only rudimentary or abstract contextual evidence. Therefore, corpora
have long been seen by some researchers as unfit for use in pragmatic research.
Indeed, some pragmatic features inevitably escape corpus linguistic analysis.
This is because in part-of-speech (POS)-tagged corpora, only those phenomena
can be studied fully whose lexical form(s) and pragmatic function(s) display a
straightforward one-to-one relationship. This relationship is found, for example, in
the words sorry and pardon, which are regularly included in apologies (cf. Jucker
2009). The form-function match is already weaker in compliments which need not
necessarily be realized using typical conventionalized patterns (cf. Jucker et al.
2008). Where there is a complete form-function mismatch, as in cases of conversational implicature, a quantitative corpus study will be useless: what listeners take
to be implicated in an utterance cannot be retrieved exhaustively from a corpus but
can only be inferred (post hoc) with varying degrees of confidence.
What a corpus can do even in those cases where the form-function mismatch of
a phenomenon prevents exhaustive searches, is to provide the analyst with illustrative examples that are not only attested and, in this sense, authentic but also embedded in their co-texts, thus giving some evidence of the context in which they
were used. Such corpus illustrations can usefully complement, or even replace, the
invented and often completely decontextualized examples that have formed the
basis of much pragmatic enquiry.
Another approach to studying pragmatics corpus-linguistically is to use pragmatically-annotated corpora. For example, a small subcorpus of the Michigan Corpus of Academic Spoken English (MICASE) is tagged for some speech acts (cf.
Maynard and Leicher 2007) and in the Corpus of Verbal Response Mode (VRM)
Annotated Utterances, all utterances are coded twice: once for their literal meaning
and once for their pragmatic meaning, using a principled taxonomy of speech acts
(cf. Stiles 1992). However, the number of such pragmatically-annotated corpora is
still small and they are coded for selected aspects of pragmatic interest only.
Given these limitations, it may be surprising that quantitative corpus analyses
of pragmatic phenomena have grown into a large body of literature and produced a
wealth of new insights. In the following sections I will present some corpus linguistic studies into pragmatic units. The approach will be selective: only a few key
studies can be presented. References to related studies, however, will be provided.
Unless otherwise indicated, the illustrative examples will be taken from the British
National Corpus (BNC), a 100-million-word corpus of contemporary British English (cf. Hoffmann et al. 2008).
2.
631
The semantics/pragmatics interface: semantic prosody
One of the clearest strengths of corpus linguistics is the analysis of what Sinclair
(1991) refers to as the idiom principle, that is, broadly, how language patterns at
the level of phraseology. A very large body of corpus linguistic literature points to
idiom phenomena such as collocation, collostruction, colligation and so forth as
important building blocks of phraseological patterning. Another, crucial, phraseological phenomenon is semantic prosody, which I will focus on in this section.
Semantic prosodies probably best illustrate Channells (2000) claim that some
pragmatic phenomena can only be revealed in studies of large corpora. This is because semantic prosodies are normally hidden not only from introspection but also
from observation of small numbers of examples. I will first briefly outline essential
characteristics of semantic prosody and then illustrate the phenomenon in a case
study of BREAK out.
Semantic prosody is closely related to semantic preference, which is defined
by Stubbs as the relation, not between individual words, but between a lemma or
word-form and a set of semantically related word forms (2002: 65). Stubbs cites
the example of the adjective large, which often collocates with nouns denoting
quantity and size such as number, scale, part, amounts, etc. Semantic prosody,
on the other hand, can, as a rule of thumb (Partington 2004: 149), be seen as a subcategory of semantic preference. The distinguishing feature of semantic prosody is
that the type of semantic preference is related to polarity: the collocates of a word
are either typically positive or, more often, negative. Because of the inherently
evaluative nature of semantic prosody, alternative terminologies have been proposed, such as pragmatic meaning (Channell 2000), emotive prosody (Bublitz
2003) or evaluative prosody (Morley and Partington 2009). Further, because of the
evaluation they convey, semantic prosodies are said to be on the pragmatic side of
the pragmatics/semantics continuum (Sinclair 2004: 34).1 Bublitz (1996) and Morley and Partington (2009) also note the cohesive role of semantic prosody, which the
latter describe as the mechanism which shows how one elemental type of meaning evaluative meaning is frequently shared across units of discourse and, by ensuring consistency of evaluation or evaluative harmony, plays a vital role in keeping
the discourse together, in its cohesion (2009: 139). An illustrative example from
the Corpus of Contemporary American English (COCA) (cf. Davies 2009) is (1):
(1) Hello, everyone. I am Dr. Sanjay Gupta. Your health, there is nothing more important. # And now that winter weather has set in, there are certain health and
safety issues tat [sic] you need to be more concerned about, including how to
avoid succumbing to sickness. (COCA: CNN_YourHealth 2002)
This example involves the phrasal verb SET in, of which Sinclair (1991) observed
that its subjects refer to unpleasant states of affairs (Sinclair 1991: 74). Sinclair
cites as main vocabulary for SET in words like rot, decay, malaise, despair, ill-will,
632
Christoph Rhlemann
etc. In (1), the subject is winter weather, a rather neutral phrase. However, winter
weather is negatively evaluated. This becomes clear when we consider the context:
health and safety issues, need to be concerned about, and succumbing to sickness
are clear indices of the presence of negative evaluation. The evaluative harmony, to
use Partingtons term, in the excerpt is thus ensured by the accumulation of
negative items. See also Louws (1993: 173) observation that in many cases semantic prosodies hunt in packs and potentiate and bolster one another, an observation Bublitz (2003: 387) refers to as the need to establish a common emotive
ground by accumulating equi-polar means.
Excerpt (1) is illustrative of yet another crucial aspect of semantic prosody: it
can serve as an effort-saving device lighten[ing] the burden on the listener and
free[ing] the speaker of the tedium of labouring a point (Morley and Partington
2009: 144). This is, in Louws (1993: 157) words, because semantic prosody is defined as an aura of meaning with which a form is imbued by its collocates. In other
words, because SET in typically co-selects negative subjects, the negativity of its
collocates colours the meaning of SET in in such a way that it is perceived as
negative even if it co-occurs with a subject that is not clearly negative, such as
winter weather. Louws notion of imbuing has met with criticism. Whitsitt (2005),
for example, argues that there is no evidence for assuming that we can see the results of a diachronic process of imbuing (2005: 296; emphasis in original; see also
Bednarek 2008; Morley and Partington 2009). Whitsitt cites as evidence words
such as alleviate and heal, which habitually co-occur with clearly unpleasant words
and yet certainly [do] not come to have an unpleasant meaning because of that
company (2005: 297). On the other hand, the fact that speakers/writers can achieve
certain rhetorical effects such as irony by deliberately deviating from typical collocational patternings, as in outbreak of sanity, shows that the underlying semantic
prosody clearly persists, even and especially, in collocational clashes (Morley and
Partington 2009: 150). How can this evaluative persistence be explained?
An answer that has recently been given is the theory of priming (Hoey 2005).
As a word, such as SET in, is learnt through encounters with it in speech and writing, it is loaded with the cumulative effects of those encounters such that it is part of
our knowledge of the word that it co-occurs with other words (Hoey 2003), such
as, in the case of SET in, heavily negative items like rot, decay and so on. That is, in
(1), the speaker need not spell out that winter weather is taken as bad, for example
by adding a negative adjective such as cold, severe, or harsh. S/he can trust the addressee to know the primings of SET in, that is, a set of instructions on how to use
it, how it normally interacts with other items. Among these primings is a words
semantic prosody, an instruction, which tells the reader when you find me in a
text, read the surrounding discourse in a favourable/infavourable light, unless
theres something else around which tells you not to (Partington forthcoming).
The evaluation conveyed by semantic prosodies may be obvious or hidden to
the naked eye. Morley and Partington (2009: 151) suggest that it is best con-
633
sidered as a cline, with items such as murder and venerable on the overt-evaluation end expressing fairly clearly unfavourable and favourable evaluation, items
such as peddle and fraught with somewhere in the middle, and items such as SET in
and not + BUDGE (see below) occupying the covert-evaluation end where the
evaluation can only be identified using corpus linguistic methods.
The linguistic analysis and description of semantic prosody faces several challenges. One problem is that while it may be easy to achieve a consensus that words
such as rot, decay etc. describe negative states of affairs, the detection of negative
or positive semantic prosody may be much less straightforward in other cases.
Whether a word is used and interpreted positively or negatively depends in part on
the wider context in which it is used and in part on the speaker who uses it. Stubbs
(1996) cites the example of intellectuals, whose collocates e.g., activist, student,
leftwing, liberal may have a negative ring in some circles but be positively
evaluated in others (cf. Bednarek 2008: 122). Semantic prosody therefore cannot
be objectively derived from corpus data and requires a lot of inference on the part
of the analyst (Bednarek 2008: 132).2
A second problem is the evaluative inconsistency of semantic prosodies. Prosodies can be switched off or even reversed. Morley and Partington (2009: 149)
discuss the example of not + BUDGE, which usually carries a negative prosody.
However, under certain circumstances, for example when the subject is first person, it can be used favourably. An example of positive evaluation in the use of not
+ BUDGE is given in Hunston (2007):
(2) The Prime Minister rejected resounding calls for the resignation of the government, I will not budge, he said
Further, the distribution of semantic prosody and word form may be asymmetrical
(Bublitz 1996). That is, it may be misleading to claim that verb X or noun Y has
positive or negative prosody because in many cases different forms of one and the
same word have different evaluative tendencies. Consider the phrasal verb BREAK
out: when used transitively, as in Just as linotype operators at the Sun were breaking
out the champagne to celebrate the arrival of 1,000 pay packets (CHU 308), its
prosody is clearly positive (cf. Louw 2000); when used intransitively, as in Fires keep
breaking out (A18 1206), its prosody is decidedly negative (cf. the discussion below).
Another problem is that semantic prosodies can be subject to variation across
registers. OHalloran (2007: 15) demonstrates that, for example, the past tense
form erupted has largely positive associations in the sports report register, but
largely negative ones in the hard news register. That is, the non-register specific
concept of semantic prosody may, in some cases, have to be replaced by the concept of register prosody, which indicates that some prosodies have probabilistic
relationships to register (OHalloran 2007: 4; see also Bublitz 1996).
Finally, semantic prosody cuts across the traditional lexis/grammar dichotomy.
The GET-passive has been shown to have a negative prosody in that it typically at-
634
Christoph Rhlemann
tracts past participles sharing an adversative core meaning (Carter and McCarthy
1999; Stubbs 2002; Rhlemann 2007b). Typical examples are GET stuck, GET
caught, and GET killed. Further, semantic prosody may also be observed at a level
below the word. Rhlemann (2006) demonstrates that the prefix dys-, as in dyslexia, dysfunction, dysplasia, etc., as well as the prefix dis-, when used as a productive morpheme, as in disease, disappear, disabled, etc. form words that have
negative prosody.
In the remainder of this section a case study into the pattern (inanimate) N +
(intransitive) BREAK out will bring into relief some of the above said.
A search for the verbal lemma BREAK immediately followed by out finds
1,126 occurrences in the BNC. (3) shows a random excerpt from the concordances.
As is suggested by the inanimate nouns (underlined in (3)) preceding the string,
negative things such as hostilities, war, mayhem etc. break out.
(3) if hostilities ever broke out. Weve
When war broke out he returned to
complete mayhem broke out. Hats were flung
a fight broke out at the Zuwaya
resembling panic broke out among the ladies
Is this association with negative items systematic? Table 1 presents the 15 most
frequent nouns preceding BREAK out in the BNC; right hand column gives the
Z-scores, which measure the strength of attraction between collocates.
Table 1.
Top 15 nouns collocating with BREAK out
No.
Collocate
Freq.
Z-score
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
war
fire
fighting
fight
row
world
violence
blaze
riot
argument
hostility
scuffle
rioting
way
sweat
259
89
46
39
36
32
20
17
17
17
16
14
13
12
11
167.8
82.8
113.8
75.3
48.5
12.6
29.4
72.4
44.7
16.3
43.5
132.3
82.1
1.1
37.2
635
As can be seen from Table 1, BREAK out most commonly collocates with inanimate nouns that share physical violence and, to a lesser degree, fire meanings. War,
the most extreme manifestation of violence, is not only by far the most frequent
collocate but also has the highest Z-score; note also that the 259 occurrences of war
+ BREAK out account for more than a fifth of all occurrences of BREAK out. War is
thus a very strong collocate of BREAK out. Other nouns in the violence group include fighting (rank 3), fight (4), row (5) etc.; world, ranked 6th, occurs in combination with war forming the compound world war. The fire group includes the
nouns fire (2) and blaze (8). Inspection of concordance lines suggests that it is invariably destructive, not warming or cosy, fire that breaks out. Since destructive
fire is an inevitable concomitant of war, we might subsume the fire meanings under
the heading of the violence meanings, defining the negative prosody of BREAK out
thus: BREAK out co-selects inanimate nouns that typically express violence (including destructive fire) meanings. BREAK out illustrates that the concepts of semantic prosody and semantic preference are not mutually exclusive in that the
phrasal verb is a semantic prosody with a clear semantic preference.
Interestingly, there are six occurrences of peace + BREAK out in the BNC. Inspection of the concordance lines suggests that covert evaluation is the driving
force behind these uses. Consider (4) (from the Economist, 1991):
(4) Where will the dollar head when peace breaks out? (ABH 2961)
In (4), the prospect of peace (at the time of the Gulf war) is taken as a threat to the
dollar, which has risen before the war but, it is suggested, might fall when the war
is over. That peace may have bad consequences for the dollar is not made explicit.
Rather, the negative aura of meaning surrounding BREAK out is tacitly projected
onto peace.
In sum, prosodies are a useful resource for the diagnosis of overt and covert
speaker evaluation. Further, as Louw argues, there is the same well-calculated collocational deviance in instances of irony (Louw 1993). The prospect, then, is that
semantic prosodies may help to computationally uncover irony, a pragmatic phenomenon that has so far escaped quantitative empirical study. Moreover, semantic
prosodies may assist in the study of persuasion: Propaganda, advertising and promotional copy will now be gradable against the semantic prosodies of the whole
language (Louw 1993: 173). Exploiting the diagnostic potential of semantic prosodies for these kinds of speaker meaning has only just begun and more insights
may be expected from this promising avenue of research.
3.
Pragmatic markers: like in a relevance-theoretic perspective
Pragmatic markers are words or phrases that do not add so much to the propositional content of utterances as they metalingually flag how discourse relates to
636
Christoph Rhlemann
other discourse. Pragmatic markers thus play a crucial role in facilitating processes
of pragmatic inference.
Because they are lexically relatively fixed and thus relatively easily retrievable
from a corpus, corpus research into pragmatic markers has been extremely productive.3 This section presents Andersen (2001), a corpus study on a marker that is
particular in that it attends to several discourse functions and, not surprisingly, has
attracted a wealth of recent research: like. Andersens study, which is based on the
Bergen Corpus of London Teenage Language (COLT), is notable because the approach to like is informed by relevance theory (Sperber and Wilson 1995). This
theory holds that human communication is geared towards the maximisation of relevance such that from the wealth of information and stimuli in our cognitive environments we tend to select that information and those stimuli that are most relevant to us in any given communicative situation.
Andersen identifies five broad functions of like used as a pragmatic marker.
They are illustrated in (5)-(9). In (5), like carries out an approximation function indicating that the numerical information is approximate rather than precise; in (6),
like is in an exemplification function. The common denominator of uses of like classified as approximation and exemplification is that they involve non-identical resemblance between the encoded and the communicated concepts (Andersen 2001:
237). The third function Andersen identifies is metalinguistic use, which concerns
the speakers relation to the proposition in terms of its formal linguistic characteristics (Andersen 2001: 243); in example (7), like marks a particular word choice
(reminder) as potentially inexact or inappropriate. Another function like can carry
out is as interpretive use, a term Andersen prefers over the more commonly used
term quotative on the grounds that like may be used to preface not only presentations of speech, as in (8), but also gestures and facial expressions that can be seen
as metarepresentations of speaker attitude (Andersen 2001: 254; see also Buchstaller 2008). Finally, Andersen stresses the role of like as a hesitational/linking device, that is, its function in accompanying false starts, self repairs and cut-off utterances and in providing a discourse link between syntactically distinct units of
discourse (Andersen 2001: 259). In (9), for example, like acts as a pause filler granting the speaker planning time (note the co-occurrence with the hesitation form er):
(5) I mean Ive been in two shops now theres fifty pound difference like, you
know (KB2 2401)
(6) think its the way he looks, like, if you know what I mean, you know like []
pull his face and like, look over glasses (KB1 768)
(7) PS007 >: Oh yes, I was assuming that erm a I do, I didnt really set it
out as a formal agenda just as a
PS002 >: Just as a note
PS007 >:a reminder, like. (KB0 1337)
637
(8) Yeah thats what I, why, thats what I said to Susanna and she was like dont be
ridiculous! (KP5 2163)
(9) PS03T >: well you should wipe the outside of the frame didnt you?
PS03S >: er, yes, like, you like just flick round it (KBB 6472)
The functional distribution of like in the COLT is shown in Table 2:
Table 2.
Functional distribution of like (cf. Andersen 2001: 266)
Function
Approximation
Exemplification
Metalinguistic focus
Interpretive use (quotative)
Hesitational/discourse link
21
19
19
7
35
As is shown in Table 2, the hesitational/linking functions in the corpus are most

salient, followed by approximation, exemplification, and metalinguistic use. The
quotative function, by contrast, is relatively infrequent. This may be contrary to expectations, given the higher rates for quotative like in Tagliamonte and Hudson
(1999). The discrepancy between Andersen and Tagliamonte and Hudson, however, may be due to the fact that quotative like is a recent development that has
been spreading to many regional varieties of English (cf. Buchstaller 2008) and
that Tagliamonte and Hudsons data were collected in 1996 whereas the COLT data
stem from the early 1990s.
Do all of the different functions of discourse marker like have a common root?
Andersen (2001) interprets discourse marker like as a looseness marker, that is,
as a signal that the relation between an utterance and its underlying thought is
not a one-to-one relationship, but a relation of non-identical resemblance (2001:
230). As such, it is a highly co-constructive item inviting recipients to collaborate
in the negotiation of meaning.
4.
Reference
Reference undoubtedly touches on one of the most fundamental questions concerning language: how can we speak and, in speaking, communicate something
meaningful about things, people, and states of affairs? The answers linguists and,
more importantly, philosophers have suggested are by far too varied and complex
to go into in sufficient detail in this short section (see Schwarz-Friesel and Consten
this volume). A few remarks must suffice.
638
Christoph Rhlemann
It is often claimed that referring expressions refer to their referents, such that,
for example, the expression the present U.S. American president refers to
Barack Obama. On this view, the expression denotes an object in the real world;
see Russells (1905) influential paper, entitled On denoting. Denotation, in
Russell, is taken to be a relation between an expression, considered in abstraction, and the thing that is the expressions referent or denotatum (Lycan 2008:
19). Russell distinguishes between meaning and denotation. An expression such
as the present U.S. American prime minister would have to be assigned meaning because it could be perfectly understood, but, since the US political system
does not provide for a prime minister, it would fail to denote anything. In Russells view, the sentence The present U.S. American prime minister is wise
would be considered to be false. A number of objections have been brought
against Russells theory of definite descriptions, chiefly by Strawsons (1950) article aptly entitled On referring. Strawson thought of referring not as an abstract relation between an expression and an object but argued that referring is
not something an expression does; it is something that someone can use an expression to do (1950: 326). The variable that he brings into the equation is the
context of an utterance [] and by context I mean, at least, the time, the place,
the situation, the identity of the speaker, the subjects which form the immediate
focus of interest, and the personal histories of both the speaker and those he is
addressing (1950: 336). Whether the present U.S. American president is used
to refer to Mr Obama or his predecessor or his successor or any other U.S.
American president crucially depends on the circumstances of its use. Also, the
sentence The present U.S. American prime minister is wise would not be seen
as lacking truth-value; since it builds on a false presupposition, the question
whether the statement is true or false simply does not arise. On this view, which
links reference intimately to context and, hence, inference, reference is deeply
pragmatic.
The following two subsections section present corpus-based research into two
types of referring expressions: definite noun phrases and deixis.
4.1.
Reference through definite noun phrases
One type of material that corpus linguists standardly work with are frequency lists,
that is, lists in which the words in a corpus are ranked in order of their frequency in
that corpus.
Table 3 displays what are, according to Kilgarriff (1998), the 20 most frequent
items in the conversational subcorpus and the written subcorpus of the BNC.
Table 3.
639
20 most frequent items in the conversational subcorpus (C) and the written
subcorpus (W) of the BNC
C
Rank
Frequency
Item
Tag
Frequency
Item
Tag
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
167,640
135,217
128,165
115,247
92,239
90,886
77,611
68,846
62,382
58,810
48,322
47,391
43,977
42,241
41,654
38,515
35,156
34,901
34,837
34,477
i
you
it
the
s
and
nt
a
that
yeah
he
to
they
do
oh
what
is
of
was
in
pnp
pnp
pnp
at0
vbz
cjc
xx0
at0
dt0
itj
pnp
to0
pnp
vdb
itj
dtq
vbz
prf
vbd
prp
5,776,384
2,789,403
2,421,302
1,939,617
1,695,860
1,468,146
1,892,937
1,845,350
1,839,964
1,834,957
1,768,898
1,606,027
1,605,749
1,603,178
1,590,305
1,580,267
1,561,041
1,490,673
1,435,574
1,426,207
the
of
and
a
in
to
is
to
was
it
for
with
he
be
on
i
that
by
at
you
at0
prf
cjc
at0
prp
to0
vbz
prp
vbd
pnp
prp
prp
pnp
vbi
prp
pnp
cjt
prp
prp
pnp
Table 3 reveals clear differences in the ways that language users refer in conversation (C) and writing (W) respectively. Among the 20 most frequent items in C,
there are (i) more personal pronouns (the pronoun they is not included in the top 20
in W) and (ii) far less NP-related items (prepositions and the definite and the indefinite article) than in W: there are ten NP-related items in W but only four in C.
Table 3 further shows that the definite article the is by far the most common word
in writing (roughly twice as common as the next one, the preposition of). Indeed,
the is not only the most frequent word in writing but also the most frequent word in
the whole of the BNC (cf. Kilgarriff 1998) and in most other general corpora, such
as the Cambridge International Corpus (CIC) (cf. McCarthy 1998) or the Bank
of English (BoE) (cf. Sinclair 1999). With its roughly 6 million occurrences in
the 90-million-word written subcorpus of the BNC, the accounts for more than six
percent of all word tokens in that subcorpus. It is ironic that, a good 100 years on,
640
Christoph Rhlemann
Russells verdict that to the philosophical mathematician [the] is a word of great

importance (1919: 167) should be empirically confirmed by corpus frequency
counts.
This evidence from the BNC suggests two major interpretations: (i) the most
common referring technique in writing is the use of both the definite NP and, less
importantly, the indefinite NP, while (ii) in conversation, reference is expressed
most commonly by means of pronouns. While this latter finding will be discussed
in the next subsection (5.2), the remainder of this subsection takes a closer look at
the use of the definite article both in conversation and writing.
The overriding function of the definite article is to specify that the referent of
the noun phrase is assumed to be known to the speaker and the addressee (Biber et
al. 1999: 263). That is, broadly speaking, the functions as a marker of given information. It does so in a variety of ways. In a study carried out on the Longman
Spoken and Written English (LSWE) Corpus, Biber et al. (1999) identify the following reference patterns of definite NPs marked by the.4 They are illustrated by
relevant examples:
(10) Anaphoric: A MAN died and a girl was badly injured when fire ripped
through a house yesterday. The girl, who had been clinging to a
third-floor window ledge, fell just as firemen were about to grab
her. (CBF: 2921)
(the girl is understood as the girl that was badly injured when fire ; reference is backward-looking)
(11) Indirect anaphoric:
A woman died yesterday after being knocked down by a shoplifter fleeing with a 2.58 descaler. Mrs Lillian Amelia Smith,
81, sustained a fractured skull at the store at Newham, east London, on Friday. Police are treating the incident as murder.
(AKH: 260)
(the incident can be identified via inference as the events described in the
preceding report)
(12) Cataphoric: The next main point is about the complexity of the system.
(HHW: 3638)
(the complexity in question is identifiable via the of-phrase following it; reference is forward-looking)
(13) Situational: I think there might be parcel for you at the door is it? (FPU: 358)
(based on frame knowledge the reference of door is understood as the
front door where parcels are normally delivered)
(14) Generic:
Just for once, a Frenchie has conceded that the Brits do something better. (A0C 1386)
641
(the denotes the whole class of Brits without specific reference to particular Brits)
(15) Idiom:
But, quite frankly, whats the point? (A5Y: 230)
(the is an integral part of the idiom whats the point?; it has no referring or
denoting function; see also Searle 1969: 72)
Intriguing results emerge from the analysis of the extent of use of these patterns in
different registers by Biber et al. (1999). The registers considered include three
written registers (fiction, news, and academic writing) and conversation. The results are summarized in Table 4:
Table 4.
Percentage use of reference patterns in four registers (C: conversation;

F: fiction; N: news reportage; A: academic writing) (cf. Biber et al. 1999: 266)
Situational
Anaphoric
Indirect anaphoric
Cataphoric
Generic
Idiom
Uncertain
55
25
5
5
<2.5
<2.5
10
10
30
10
15
<2.5
<2.5
15
10
30
15
30
5
<2.5
5
10
25
15
40
5
<2.5
5
The findings presented in Table 4 allow for a number of observations. Generally,

there are marked cross-register differences in the use of the reference patterns.
Specifically, anaphoric reference, which may intuitively be seen as the major reference type of definite NPs, accounts for less than a third in all four registers. Cataphoric reference represents, respectively, 30 percent in news reportage, 40 percent in academic writing, but only five percent in conversation. Situational
reference, conversely, accounts for 55 percent in conversation while it is found in
only ten percent in the written registers.
The analysis in Biber et al. (1999) also enables us to understand more deeply
the relative frequencies of the definite article in the word frequency list discussed
above. We saw that the is by far the most frequent word in the written part of the
BNC but less highly ranked in the conversational part of the BNC. Applying the
Biber et al. analysis to these frequency-based rankings, we can now see that definite NP reference is not only relatively less important in conversation, where it is
second to reference through personal pronouns, than in writing, where it is by far
the most important reference type. We can also see that the difference in the use
of the definite NP between conversation and writing is in fact much more dramatic because reference through definite NP fulfills different functions in the two
modes: while it is largely (indirect) anaphoric and cataphoric and, hence, endo-
642
Christoph Rhlemann
phoric in writing, it is overwhelmingly situational and, hence, exophoric in conversation.

Another type of exophoric reference is through the use of deixis. Its use in
discourse presentation will be the focus of the next section. That section will also
facilitate a more detailed understanding of the overrepresentation of personal pronouns in conversation we observed in Table 3 above.
4.2.
Deixis and discourse presentation5
Deictic reference is a prime example of exophoric reference because establishing

referents of deictic words necessarily requires extra-linguistic context. For
example, the references of the person-deictic word I and the time-deictic word tonight are completely dependent on the speech situation in which the two words are
uttered. Deictic words presuppose a deictic centre (origo) relative to which they
are computed. The deictic centre is in most cases associated with the current
speaker. It can also be seen as the viewpoint from which the speaker refers to the
dimensions of person, time, and place (Lyons 1977: 638). This viewpoint constrains the use of deictic words such that person, time, and place deictics normally
conform to this, one, viewpoint, thus forming a deictic system (Levinson 1983:
68).
One area of language use in which speakers can use deictic words that do not
conform to their own deictic system is discourse presentation, alternatively referred to as speech reporting and constructed dialogue. This will be explained
with reference to corpus research carried out by McIntyre et al. (2004), a study
based on the Lancaster Speech, Writing and Thought Presentation Spoken Corpus,
a small, balanced corpus of contemporary spoken English drawn from the conversational subcorpus of the BNC and oral history archives from Lancaster University
containing elicited interviews. The authors propose categories for the presentation
of not only speech but also thought and writing. For space limitations, only the categories as well as the studys initial results for speech presentation will here be reported on.
McIntyre et al.s (2004) model of speech presentation provides for six major
categories. With reference to the examples listed below, the categories include
the two direct categories Direct, as in (16), and Free Direct, as in (17). Both
modes represent speech in the form in which it is directly manifest to a listener
(Leech and Short 1981: 345) in an anterior situation. A distinction between the
two is that Direct requires a preceding reporting clause, whereas Free Direct is not
accompanied by a reporting clause. The categories further include Indirect, as in
(18), and Free Indirect, as in (19). Like Free Direct, Free Indirect has no accompanying reporting clause. In (Free) Indirect mode the propositional content of the
original speech is specified, but no claim is made to present the words and structures originally used to utter that proposition (McIntyre et al. 2004: 61). Further,
643
the categories include Representation of Speech Act, as in (20). This category

presents the illocutionary force of an utterance or text (part) with an optional
noun or prepositional phrase indicating the topic (McIntyre et al. 2004: 61) but
does not claim to represent the propositional content or the original wording of
that content. The final category is Representation of Voice, as in (21), which
captures minimal references to speech with no indication of the illocutionary
force, let alone the propositional content or form of the utterance (part) (McIntyre
et al. 2004: 62).
(16) Direct (D):
Yeah she said erm you have it my love dont you

worry but I said no Ill give you some no no I dont
want no money for it at all she said, you take it my
love.
(KB6 1355)
(17) Free Direct (FD):
[Speaker is reporting how someone asked him/her for

change for a fiver].
I said no! [] only. So well can you lend me a
pound? I said no!
(KD5 7945)
(18) Indirect (I):
And I thought she might have come today cos she said
shed bring up the money but she hasnt, not yet.
(KB6 474)
(19) Free Indirect (FI):
Father said can my girls come? No they couldnt come

(McIntyre et al. 2004: 60)
(20) Representation of
Speech Act (RSA):
The only, the only er thing I complained about with

you is, is the butter.
(KBC 6240)
(21) Representation
of Voice (RV):
the children didnt even know he was talking to them.

(KB0 375)
Which of these modes is the most frequent in everyday speech? The answer
McIntyre et al.s analysis suggests is straightforward:
644
Christoph Rhlemann
Table 5.
Proportions of speech presentation categories (McIntyre et al. 2004)
Category
Frequency
Rank
D
FD
I
FI
RSA
RV
38 %
4%
12 %
2%
27 %
17 %
1
5
4
6
2
3
As is shown in Table 5, Direct is by far the most common speech presentation

mode in McIntyre et al.s spoken corpus, accounting for more than a third of all
presentations. The second most frequent mode is RSA, accounting for more than a
quarter, followed by RV, accounting for less than a fifth. Free Direct is relatively
underrepresented in the corpus (ranking fifth), a fact which may be due to the nonconversational interview sections in the corpus. In spontaneous conversation, Free
Direct may be more common (cf. Stenstrm et al. 2002: 110 ff.).
In order to make sense of the dominance of the direct mode in speech it is important to consider this mode in terms of reference and deixis. A fundamental difference between the direct modes and all other modes lies in the speaker perspective (Coulmas 1986: 2). In Direct and Free Direct, the perspective is that of the
presentee: all deictic features are appropriate to the speaker in the anterior situation. By using words which are ostensibly marked as the words of a non-present
speaker, the presenter assumes the role of the presentee: he/she enacts the presentee and his/her speech. By contrast, in the non-direct categories (Free) Indirect,
RSA, and RV, the perspective is that of the presenting speaker: all deictic features
are appropriate to the speaker in the posterior, discourse presenting, situation (cf.
McIntyre 2004: 60).
The switch into the presentees perspective and deictic system that characterizes (Free) Direct has important implications on the referential plane. Each time
the perspective and hence the deictic system changes, so do the referents of the
same deictics. Prime examples of such referential changes are the pronouns I and
you. The two pronouns are core person deictics referring to people immediately
present in the speech situation. (As such, they are distinguished from third-person
pronouns such as he or she which typically refer anaphorically or cataphorically to
entities in the text.) Reconsider example (16), reprinted here as (22). The speaker is
presenting three utterances: two by a speaker who is not present in the posterior,
presenting, situation (utterances 1 and 3) and one she made herself in that situation
(utterance 2). For illustration purposes let us call the presenter speaker A and the
presentee speaker B. The letters A and B in the excerpt indicate co-referential
deictics (presented speech is in bold and deictics are underlined):
(22) 1
2
3
645
Yeah she said erm you(A) have it my love dont you(A) worry
but I(A) said no I(A)ll give you(B) some
no no I(B) dont want no money for it at all she said, you(A) take it my
love.
In the excerpt, a number of switches in deictic reference can be observed. In utterance 1, speaker A uses you twice, the person deictic appropriate to speaker Bs
deictic system to refer to an addressee, in that case, speaker A. Then, in utterance
2, speaker A uses I twice to refer to herself (speaker A), and you to refer to
speaker B thus deploying the deictics appropriate to her own deictic system. Finally, in utterance 3, speaker A switches back into speaker Bs deictic system using
I, whose referent is speaker B, and you, whose referent she is herself (speaker A).
That is, because in Direct presentation, speakers assume different perspectives
and the corresponding deictic systems, reference in Direct mode jumps back and
forth between referents: deictic-system oscillation leads to reference oscillation.
Further, we need to be aware that such discourse presentational changes in perspective are not isolated events occurring here and there but pervasive not only
in longer conversations but often within one and the same utterance, particularly
in narrative.
The study by McIntyre et al. is thus significant on two counts. First, it suggests
one answer (out of several) to the question why I and you are among the most frequent words in conversation, as seen in the previous section.6 I and you are so common in conversation because the most frequent discourse presentation mode is Direct: when presenting discourse, conversationalists mostly use the mode in which
deictic references such as I and you used by speakers in anterior situations are
copied and pasted from that anterior situation into a posterior presenting situation.
Second, McIntyre et al. underscore a view of deixis as a far more flexible system than is often thought. The study provides empirical evidence to suggest that
presenters are by no means confined to presenting discourse and the references
therein from their point of view, with all deictic references conforming to their
origo and the corresponding deictic system. Presenters are free, and make ample
use of that freedom, to shuttle between various origos, deictic systems and perspectives thus creating a continuous oscillation of reference. To judge by the commonness of direct speech presentations, presenters perform this referential oscillation
with great ease. What about the listeners? How can they resolve the ambiguities in
reference that the constant oscillation is likely to bring with it? Clearly, different
voices may be marked off by intonational means, speaker change may be signalled by use of reporting clauses such as he said, I says, etc., utterance openers
such as oh and well may flag the start of a direct speech presentation. But in many
cases no such additional processing instructions are being used. Still, listeners
seem to process switches in perspective and deictic system equally easily as pres-
646
Christoph Rhlemann
enters do. Precisely what enables listeners to resolve reference, which presenters
handle with such flexibility, is still underresearched. Therefore, investigating reference resolution in discourse presentation might add valuable insights to pragmatic theories of inference.
5.
Speech acts
As noted in the introductory section, the relation between quantitative corpus

analysis and speech act analysis is not a one-to-one match because corpora record
surface forms while speech acts are defined on the basis of their function, not their
form (Jucker 2009: 7). However, speech acts are often realized using illocutionary force indicating devices (Searle 1969: 30) or largely routinized forms. Such
features have proven useful anchors for corpus searches for speech acts. The
number of corpus studies into speech acts is to date still small. Two such studies
will be presented in this section, one on the speech act of complimenting and one
on the indirect speech act of suggesting.7
5.1.
Compliments
Compliments have received a fair amount of attention in pragmatic research

(for an overview see Jucker 2009). In a pioneering study, Manes and Wolfson
(1981), using a notebook method (cf. Jucker 2009), claim specific frequencies
for the compliment patterns they found in their data. The study that will be
briefly reviewed here, Jucker et al. (2008), aims to assess the accuracy of Manes
and Wolfsons findings with the help of the BNC. The study is also intended to
highlight methodological problems involved in approaching speech acts via corpora.
The two studies report strikingly different frequencies of compliments/patterns. Using search strings which correspond to the compliment patterns established by Manes and Wolfson, hand-searching subsets of matches and extrapolating their frequencies to the entire sets, Jucker et al. conclude that there are
approximately 343 compliments in the 100-million-word BNC; Manes and Wolfson, by contrast, collected 686 compliment sequences. Further, Manes and Wolfson established nine compliment patterns; the frequencies Jucker et al. found for
these patterns diverge considerably from the frequencies reported by Manes and
Wolfson. The frequencies are shown in Table 6:

Table 6.
647
Frequencies in Manes and Wolfson (1981) and Jucker et al. (2008)
Pattern
Manes & Wolfson
Jucker et al.
1
2
3
4 and 6
5
7
8
9
53.6 %
16.1 %
14.9 %
5.7 %
2.7 %
1.6 %
1.6 %
76.4 %
3.2 %
2.3 %
5.0 %
6.4 %
3.5 %
3.2 %
1.0 %
The differences are most marked with regard to patterns 13. While pattern 1 accounts for slightly more than half of all compliments in Manes and Wolfsons data,
this pattern accounts for more than two thirds in the BNC. Even more striking, in
relative terms, are the differences for patterns 2 and 3: the frequencies reported in
Manes and Wolfson are more than five times as high as in Jucker et al.8 The three
patterns are illustrated in (23)-(25) (optional elements in brackets):
(23) NP+copula+(intensifier)+adjective:
You look so beautiful, he whispered.
(FSF 1954)
(24) I+(intensifier)+verb of liking+NP
You know I really like you!
(KE1 1859)
(25) Pronoun+ copula+(intensifier)+(indefinite article)+adjective+(NP)
Another two pounds a week for that? He looked anxiously at her. Oh,
thats very generous, thank you. That would be marvelous.
(CCM 1399)
The significance of the Jucker et al. corpus study lies not in an assumed superiority
of the corpus method over the notebook (or any other non-corpus linguistic)
method and thus in the claim that the Jucker et al. frequencies were more reliable
than Manes and Wolfsons. By contrast, Jucker et al. demonstrate and discuss in
great detail the merits and, more importantly, shortcomings of the corpus method
as an alternative approach to speech acts. As the authors point out, the surface patterns used are crude (Jucker et al. 2008: 290). This is because the hits either include too many irrelevant matches (a problem of precision) or they fail to retrieve
all relevant examples (a problem of recall). Indeed, precision and recall were
very low, ranging between less than 1 and 20 percent. For example, poor recall may
be the reason why a much lower number of compliments was found in the BNC: al-
648
Christoph Rhlemann
though highly differentiated search strings were used, not all compliments could be
retrieved because compliments, particularly in spoken sections, may include some
minor deviation from the search string, such as pauses or repairs.
Indeed, as the authors point out, [a]lmost every query method fails to have
complete precision and recall (Jucker et al. 2008: 276). To solve this problem,
large pragmatically-annotated corpora would be needed. These, however, are still
in very short supply. Given that the notebook (and any other non-corpus-based)
method have their drawbacks too (see Jucker 2009 for a balanced discussion), the
corpus approach, even in its present imperfect form, is nonetheless a welcome addition to the variety of methods used to study speech acts.
5.2.
Suggestions
Indirect speech acts are cases in which one illocutionary act is performed indirectly by way of performing another (Searle 1975: 60), as in Can you pass the
salt? A much debated question in pragmatics is how listeners resolve the ambiguity between, in this case, question and request. It is often assumed that the adequate interpretation of indirect speech acts requires a complex chain of inferences
by which a recipient first decodes the literal force (in this case, that a question has
been posed to him/her about his/her ability to pass the salt), then realizes that
something is up with this question (for example, that it is not in accordance with
the Cooperative Principle because it fails to be relevant) and only then infers that
the question is not to be taken literally but rather indirectly, as a request by the
speaker to be passed the salt. Corpus studies, by contrast, emphasize the role less of
inferences but more of collocational patterning. This section will highlight one
such study, namely Adolphss (2008), which is based on the Cambridge and Nottingham Corpus of Discourse in English (CANCODE). The focus in this study is
on speech act expressions introducing suggestions; the expressions investigated
include, among others, why dont you.
Why dont you can be used both in direct and indirect speech acts: when used to
introduce a genuine question, a direct speech act is performed; when used to make
a suggestion, a question form is used to put forward a proposal for consideration by
the addressee (in Searles taxonomy of speech acts, suggestions would count as
directives):
(26) PS52K>: I dont believe that.
PS52C>: Why dont you believe it? Its a survey
(KP6 1737)
(27) PS1C1>: but Ive got nobody to go with!
PS1JA>: Oh! Why dont you come with us?
(KDW 2752)
649
The majority of uses of why dont you, Adolphs observes, perform a suggestion.
She elaborates a functional profile of why dont you, consisting of its collocations,
the discourse factors bearing upon its use and its contextual distribution. Thus, she
observes that suggestion-why dont you, henceforth S-WDY, is often preceded by a
form of SAY introducing speech presentation, as in (28). No such association with
speech presentation is reported for question-why dont you, henceforth Q-WDY.
(28) oh I think its because I was saying why dont you come up like this week
(KDM: 7460)
As to right-hand collocates, at N+1 (that is, in the slot immediately after why
dont you) S-WDY is regularly followed by a group of transitive verbs including
ask, get, tell, and use. The set of verbs that Q-WDY collocates with is distinctly
different: they include like, want, and have to. Another key collocate is also found
at N+1: the marker just which, in association with S-WDY, serves to down-tone the
imposition implied in suggestions (remember that suggestions are a type of directives). Consider:
(29) Why dont you just sit down somewhere?
(KBH 4400)
Further, S-WDY and Q-WDY are distinguished by the type of response they typically trigger: while Q-WDY require a more detailed answer (Adolphs 2008: 62),
the responses to S-WDY range from minimal acknowledgment tokens to agreement, or evaluations of the suggestion (ibid.). Finally, Adolphs found that
S-WDY was most frequent in the intimate speaker relationship category in the
CANCODE, that is, it occurs most frequently in interactions between people
whose social closeness is maximal (partners, family, very close friends). In this
context type, speakers are most off-guard and the imposition implied in S-WDY
as a directive is felt as less threatening.
Thus, we see that the uses of Q-WDY and S-WDY are, indeed, idiomatic in
the sense that a large number of distinctly different co-textual and contextual factors bear upon their use and it appears plausible to assume that these idiom factors are salient enough for communicative partners to disambiguate the two speech
acts.
This is not to say that all indirect speech acts are idioms and do not require inferential processes for their interpretation. What Adolphss case study shows is
merely that some indirect speech acts may be less indirect than rather idiomatic,
and it is as yet by no means clear whether the share of such idiomatic speech acts in
all possible types of (indirect) speech act is large or small. No doubt, it is still early
days for corpus linguistic research into speech acts. However, the beginnings are
promising and the prospects are that corpora can make important contributions to
speech act theory.
650
6.
Christoph Rhlemann
Looking to the future
Current corpora facilitate fascinating observations of how words are actually used.
However, they fail to represent communication beyond the word. A particular
challenge for current research is therefore to integrate corpus linguistic methods
and theories of multimodal linguistic research (Carter and Adolphs 2008: 276). At
present, corpora targeted on aspects of multimodal communication are both small
in size and number. Their central purpose is to facilitate explorations of how meaning is made through multimodal patterns, that is, patterns of interaction between
verbal and nonverbal choices (Carter and Adolphs 2008: 281).
Research faces major challenges before this goal is reached. Given that nonverbal meaning seems to make up a very large chunk of overall meaning
(Birdwhistell 1970: 1578 estimates that probably no more than 30 to 35 percent
of the social meaning of a conversation or an interaction is carried by the words) it
will not be surprising that the nonverbal semiotic systems may be as highly differentiated as the verbal system. For example, Rim and Schiaratura (1991: 248) present a taxonomy of speech-related hand gestures which includes six broad variables: speech markers, ideographs, iconic gestures, pantomimic gestures, deictic
gestures, and symbolic gestures. Ekman and OSullivan (1991: 176) discuss evidence for the cross-cultural recognition of at least six emotions expressed via facial
actions: happiness, anger, fear, sadness, surprise, and disgust. Variables of prosody
include rhythm, volume, tempo, voice quality, and intonation with its manifold
subvariables. Obviously, the task of developing a coding scheme to capture such a
wealth of individual variables corpus-linguistically is daunting. It becomes even
more daunting considering that the variables interact rather than act independently
of one another. Therefore, an important goal is the development of tools that provide an integrated approach to the representation of the data (Carter and Adolphs
2008: 283; emphasis in original). Another complicating factor is that verbal and
nonverbal choices are hard to align since within any sequence a substantial
number of utterances and gestures made by speaker and hearer overlap (Carter
and Adolphs 2008: 284). That is, unlike speaking turns which are taken orderly in
the sense that normally one speaker speaks at a time, nonverbal turns are much
less restricted: while a speaker is speaking (and acting nonverbally) the listener(s)
may produce, in response to the speakers unfolding utterance, nonverbal signals
and actions themselves.
Given these challenges, it is small wonder that current multimodal corpus analyses are decidedly selective, focussing on narrow multimodal phenomena rather
than trying to study multimodal patterning in its (at present overwhelming) complexity. One such work in progress is Baldry and Thibault (2006), who take a systemic-functional approach to analyze gaze in a corpus of TV car advertisements.
Another work in progress is Carter and Adolphss (2008) Headtalk project. This
project, which is based on a small corpus of several hours of video-taped MA and
651
PhD supervision sessions at Nottingham University, is intended to explore the patterning of multimodal backchannels, focussing specifically on head nods as a type
of nonverbal backchannel in co-occurrence with verbal backchannels.
To conclude, multimodal corpus linguistics is very much in its infancy
(Baldry and Thibault 2006: 181). However, since this strand of research offers intriguing prospects for an enhanced description of how speakers mean more than
they say, the construction, annotation, and exploitation of multimodal corpora may
in future become a major site of corpus pragmatic research.
Notes
1. Corpus linguistic explorations into evaluative language are numerous. They include,
among many others; Aijmer (1989) on tails, that is, postponed items succeeding the core
of the clause, as in Pathetic behaviour that is, innit?, which primarily seem to fulfill an affective-stance function; Biber and Finegan (1988) on adverbial stance types across text
clusters; and Norrick (2009) on interjections in narrative.
2. An initial attempt at quantifying good and bad prosodies without relying on the
researchers subjective, evaluative judgments is Dilts and Newman (2006), who use a
method based on experimentally measured judgments of goodness and badness obtained prior to, and independently of, corpus-based studies (2006: 240; emphases in
original).
3. A few selected corpus papers on discourse markers include: Aijmer (1987) on the mental
processes signaled by oh and ah; Stenstrm (1998) on cos as a continuation (rather than a
cause) marker; Lenk (1998) on the coherence inducing function of discourse markers. For
a recent collection of papers on discourse markers see Jucker and Ziv (1998).
4. Definite NPs can be marked not only by the (by far the most frequent definite determiner)
but also by possessive determiners (e.g., his, her), and demonstrative determiners (e.g.,
this, that) (Biber et al. 1999:26910). In Biber et al.s study, definite NPs were identified
using an interactive programme which looked for NPs co-occurring with the definite article only (Biber, personal communication).
5. For deixis and indexicality cf. Hanks this volume.
6. The reasons why I and you are so highly common in spontaneous speech are undoubtedly
manifold: they include planning-induced repetition, as in But I cert I I I I I ju I it it just
sounds [] (KB7: 3681), use of I and you in high-frequent discourse markers such as I
know, I see, you know etc. and the fact that, in conversation, speaker and addressee are in
immediate contact, and the interaction typically focuses on matters of immediate concern (Biber et al. 1999: 333) (for a more detailed discussion see Rhlemann 2007a).
7. Further corpus studies into speech acts include Aijmer (1996) on routinized speech act
expressions based on the London-Lund Corpus (LLC) and Kohnen (2000), a pilot study
into explicit performatives based on the Lancaster Oslo/Bergen (LOB) Corpus, the London-Lund Corpus (LLC), and the (historical) Helsinki Corpus (HC); for speech acts in
general see Collavin this volume.
8. Note that the differences in patterns 57 are more important, in relative terms, than the
differences in 1, 4 and 6; however, since the percentage values for patterns 57 are low,
the differences cannot be assigned much significance.
652
Christoph Rhlemann
References
Adolphs, Svenja
2008
Corpus and context. Investigating pragmatic functions in spoken discourse.
Aijmer, Karin
1987
Oh and Ah in English conversation. In: Meijs, Willem (ed.), Corpus Linguistics
and Beyond. 6168. Amsterdam/Atlanta, GA: Rodopi.
Aijmer, Karin
1989
Aijmer, Karin
1996
Conversational Routines in English. London: Longman
Andersen, Gisle
2001
Pragmatic Markers and Sociolinguistic Variation. A Relevance-Theoretic Approach to the Language of Adolescents. Amsterdam: Benjamins.
2006
System and Corpus: Exploring Connection, 16483. London/Oakville: Equinox.
Bednarek, Monika
2008
and Linguistic Theory 4(2): 119139.
1988
Adverbial stance types in English. Discourse Processes 11: 134.
Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad and Edward Finegan
1999
Longman Grammar of Spoken and Written English. Harlow: Pearson.
Birdwhistell, Ray L.
1970
Kinesics and Context: Essays on Body Motion Communication. Philadelphia:
Bublitz, Wolfram
1996
Semantic prosody and cohesive company: somewhat predictable. Leuvense
Bijdragen 85: 132.
Bublitz, Wolfram
2003
Buchstaller, Isabelle
2008
The localization of global linguistic variants. English World-Wide 29(1): 1544.
Carter, Ronald and Svenja Adolphs
2008
Linking the verbal and the visual: new directions for corpus linguistic. Language and Computers 64: 275291
Carter, Ronald and Michael J. McCarthy
1999
The English get-passive in spoken discourse: description and implications for
an interpersonal grammar. English Language and Linguistics 3(1): 4158.
Channell, Joanna
2000
Corpus-based analysis of evaluative lexis. In: Susan Hunston and Geoff
Thompson (eds.), Evaluation in Text: Authorial Stance and the Construction of
Discourse, 3855. Oxford: Oxford University Press.
653
Cook, Guy
1990
Transcribing infinity: Problems of context presentation. Journal of Pragmatics
14: 124.
Coulmas, Florian (ed.)
1986
Direct and Indirect Speech. Berlin: Mouton de Gruyter.
Davies, Mark
2009
The 385+ million word Corpus of Contemporary American English (1990
2008+). Design, architecture, and linguistic insights. International Journal of
Dilts, Philip and John Newman
2006
A note on quantifying good and bad prosodies. Corpus Linguistics and
Linguistic Theory 2(2): 233242.
Ekman, Paul and Maureen OSullivan
1991
Facial expression: Methods, means, and moves. In: Robert S. Feldman and
Bernard Rim, Fundamentals of Nonverbal Behaviour, 163199. Cambridge:
Hoey, Michael
2003
Lexical priming and the properties of text. Available at: http://www.monabaker.
com/tsresources/LexicalPrimingandthePropertiesofText.htm (last visited October 2009)
Hoey, Michael
2005
Lexical priming. A new theory of words and language. London: Routledge.
Hoffmann, Sebastian, Stefan Evert, Nicholas Smith, David Lee and Ylva Berglund Prytz
2008
Corpus Linguistics with the BNCweb a practical guide. Frankfurt am Main:
Peter Lang
Hunston, Susan
2007
Semantic prosody revisited. International Journal of Corpus Linguistics 12
(2): 249268.
Jucker, Andreas H.
2009
Speech act research between armchair, field and laboratory. The case of compliments. Journal of Pragmatics doi:10.1016/j.pragma.2009.02.004
Jucker, Andreas H. and Yael Ziv (eds.)
1998)
Discourse Markers. Descriptions and Theory. Amsterdam: Benjamins.
Jucker, Andreas H., Gerold Schneider, Irma Taavitsainen and Barb Breustedt
2008
Fishing for compliments. Precision and recall in corpus-linguistic compliment
research. In: Andreas H. Jucker and Irma Taavitsainen (eds.), Speech Acts in
the History of English, 316341. Amsterdam: Benjamins.
Kilgarriff, Adam
1998)
BNC database and word frequency lists. Available at:
http://www.kilgarriff.co.uk/bnc-readme.html (last visited October 2009).
Kohnen, Thomas
2000
Corpora and speech acts: The study of performatives. In: Christian Mair and
Marianne Hundt (eds.), Corpus Linguistics and Linguistic Theory. Papers from
the Twentieth International Conference on English Language Research on
Computerized Corpora (ICAME 20). Freiburg im Breisgau 1999, 17786.
Amsterdam: Rodopi.
654
Christoph Rhlemann
Lampert, Andrew, Robert Dale and Ccile Paris

2006
Classifying Speech acts using Verbal Response Modes. Proceedings of 2006
Australian Language Technology Workshop (ALTW 2006), 3441.
Leech, Geoffrey and Michael H. Short
1981
Style in Fiction. London: Longman
Lenk, Uta
1998
Marking discourse coherence. Functions of discourse markers in spoken English. Tbingen: Narr.
1983
Louw, Bill
1993
Text and Technology, 15792. Amsterdam: Benjamins.
Louw, Bill
2000
Contextual prosodic theory: Bringing semantic prosodies to life. In: Chris
Heffer and Helen Saunston (eds.), Words in context. Discourse Analysis Monograph 18 [CD Rom]. Birmingham: University of Birmingham.
Lycan, William G.
2008
Philosophy of language. A contemporary introduction. New York: Routledge.
Lyons, John
1977
Semantics. Volumes I and II. Cambridge: Cambridge University Press.
Manes, Joan and Nessa Wolfson
1981
The compliment formula. In: Florian Coulmas (ed.), Conversational Routine.
Explorations in Standardized Communication Situations and Prepatterned
Speech, 115132. The Hague: Mouton.
Maynard, Carson and Sheryl Leicher
2007
Pragmatic annotation of an academic spoken corpus for pedagogical purposes.
In: Eileen Fitzpatrick (ed.), Corpus Linguistics beyond the Word. Corpus Research from Phrase to Discourse, 107115. Amsterdam: Rodopi.
McCarthy, Michael J.
1998
Spoken Language and Applied Linguistics. Cambridge: Cambridge University
Press.
McIntyre, Dan, Carol Bellard-Thomson, John Heywood, Tony McEnery, Elena Semino and
Mick Short
2004
Investigating the presentation of speech, writing and thought in spoken British
English: A corpus-based approach. ICAME Journal 28: 4976
Mey, Jacob L.
1991
Pragmatic gardens and their magic. Poetics 20: 233245
Morley, John and Alan Partington
2009
A few Frequently Asked Questions about semantic or evaluative prosody.
International Journal of Corpus Linguistics 14(2): 139158.
Norrick, Neal
2009
Using large corpora of conversation to investigate narrative. The case of interjections in conversational storytelling performance. International Journal of
655
OHalloran, Kieran
2007
Critical discourse analysis and the corpus-informed interpretation of metaphor
at register level. Applied Linguistics 28(1): 124.
Partington, Alan
2004
Utterly content in each others company. Semantic prosody and semantic
preference. International Journal of Corpus Linguistics 9(1): 131156.
Partington, Alan
forthc.
Semantic prosody, evaluation, priming and phrasal irony. In: Karin Aijmer and
Christoph Rhlemann, Corpus Pragmatics. Exploring Speaker Meaning Computerized Corpora.
Rim, Bernard and Loris Schiaratura
1991
Gesture and speech. In: Robert S. Feldman and Bernard Rim. Fundamentals
of Nonverbal Behaviour, 239281. Cambridge: Cambridge University Press.
Rhlemann, Christoph
2006
Coming to terms with conversational grammar: Dislocation and dysfluency. International Journal of Corpus Linguistics 11(4): 385409.
Rhlemann, Christoph
2007a
Conversation in Context. A Corpus-driven Approach. London: Continuum.
Rhlemann, Christoph
2007b
Lexical grammar: the GET-passive as a case in point. ICAME Journal 31:
111127.
Russell, Bertrand
1905
On denoting. Mind 14: 479493. Reprinted in Mind (2005) 114: 873887.
Russell, Bertrand
1919
Introduction to Mathematical Philosophy. London: Routledge.
Searle, John R.
1969
Speech acts. An essay in the philosophy of language. New York: Cambridge
University Press.
Searle, John R.
1975
Indirect Speech Acts. In: Peter Cole and Jerry L. Morgan (eds.), Syntax and
Semantics III, 5982. New York: Academic Press.
Sinclair, John McH.
1991
Corpus, concordance, collocation. Oxford: Oxford University Press.
Sinclair, John McH.
1999
A way with common words. In: Hilde Hasselgard and Signe Oksefjell (eds.),
Out of corpora: Studies in honour of Stig Johansson, 157179. Amsterdam:
Rodopi.
Sinclair, John McH
2004
Trust the text. London: Routledge.
1995
Relevance. communication and cognition. 2nd ed. Oxford: Blackwell.
1998
From Sentence to Discourse: Cos (because) in Teenage Talk. In: Andreas H.
Jucker and Yael Ziv (eds.), Discourse Markers. Descriptions and Theory.
Stenstrm, Anna-Brita, Gisle Andersen and Ingrid K. Hasund
2002
Trends in teenage talk. Amsterdam: Benjamins.
656
Christoph Rhlemann
Stiles, William B.
1992
Describing talk. A taxonomy of verbal response modes. Newbury Park/CA:
Sage Publications.
Strawson, Peter F.
1950
Stubbs, Michael
1996
Text and corpus analysis. Oxford: Blackwell.
Stubbs, Michael
2002
Words and phrases. Corpus studies of lexical semantics. Oxford: Blackwell.
1999
youth. Journal of Sociolinguistics 3(2): 147172.
Whitsitt, Sam
2005
A critique of the concept of semantic prosody. International Journal of Corpus
Linguistics 10(3): 283305.
23.
The transcription of face-to-face interaction1

1.
Introduction
The transcription of face-to-face interaction presents formidable challenges for researchers who study pragmatics. Those venturing into this domain are confronted
by a wide variety of transcription systems that have been devised by researchers
from a diversity of disciplines over several decades. Although these systems are
not mutually exclusive, they often possess large differences in scope, emphasis,
and nomenclature, as well as in the symbols used to transcribe these dimensions.
It is beyond the scope of this chapter to reconcile these systems, and it is far
from clear that such a reconciliation would be desirable. Instead, we will provide
an overview of a number of these systems, and make suggestions about their suitability for various transcription needs. In addition, we will address transcription
issues with regard to speech in diverse populations (e.g., children, aphasics, and
cognitively impaired individuals). Finally, we will consider the challenges of transcribing the non-acoustic (i.e., facial and gestural) signals that are of special interest to pragmatics researchers. It is our hope that this overview and discussion will
provide some guidance to those who wish to practice the art of transcription.
2.
Some preliminaries
2.1.
Purpose of transcription
Researchers who study pragmatics create transcriptions in order to test their theories about discourse. In other words, no one creates transcriptions of face-to-face
interactions as an end in themselves; rather, transcription is always performed as a
means to an end. For example, President Richard Nixons Oval Office tapes (New
York Times 1974) were transcribed for the purpose of determining what the president knew about the Watergate cover-up. (The frequent use of the term expletive
deleted by the transcriber provided the public with some unintended pragmatic insight into the speech of the head of the Executive Branch.) Other researchers have
gone on to use Nixons recordings for their own research purposes (e.g., Novick,
Walton and Ward 1996), but the original transcriptions were not created with any
other purpose in mind.
The fidelity of transcription is a matter of some importance in the fields of
medical and legal transcription. However, in these applied settings, the focus is pri-
658
marily on what was said or intended as opposed to how it was uttered. Although
there may be important commonalities between transcription in applied and research domains, they will not be reviewed here.
2.2.
Transcription versus coding
Many issues in pragmatics may be profitably explored without recourse to transcription. Because transcription is extremely time intensive (see section 2.5), it
may in fact amount to overkill for many purposes. For instance, if one simply
wanted to compare the number of specific discourse markers (e.g., Schiffrin 1987)
in two stretches of discourse, it might in fact be easier to tally them directly from an
audio- or videotape of the interaction. This approach might also be more accurate,
because it would avoid many of the complications that are inevitable when an interaction is reduced to a transcription. Such coding should be possible for many
phenomena that are conceptually well-defined and macroscopic.
In other cases, a researcher may be interested in classifying utterances according to a particular taxonomy, such as illocutionary speech acts (e.g., Searle 1975).
In such cases, it would probably be easiest to work from a relatively broad transcription of the interaction. The focus of this review will be the broad range between counts and coding schemes, which require the use of a particular transcription system that produces a faithful record of an interaction.
2.3.
Issues of terminology
Phonetic transcriptions are frequently characterized as broad or narrow, in

which narrow transcription documents the allophonic variation in speech sounds.
International standards exist for transcribing phonetic information at both the
broad and narrow levels (Pullum and Ladusaw 1996), although not surprisingly,
reliability is higher for broad transcriptions than narrow ones (Shriberg and Lof
1991). In a similar way, the coding of non-phonetic dimensions, such as pauses or
gestures, can be characterized as broad or narrow. A paradox of transcription is
that, as a transcript becomes narrower (and theoretically, more faithful to the discourse it represents) it becomes more difficult to read and interpret. In other
words, there may be a tradeoff between the fidelity of a transcript and its intelligibility.
Consider, for example, a few turns from the relatively narrow transcription provided by Kyratzis (2001) in her analysis of the interactions of preschool friends:
1. Speaker1: if someone comes, then we **hi::de, really//
2. Speaker 2: ==yeah/
3.
were *shy:: wizards// (Kyratzis 2001: 363)
659
The symbols provide important information about intonation, lengthened segments, and latching, although most readers will need to refer to the authors list of
transcription conventions to fully decode these features.
It may also be helpful at this point to define the specific dimensions that are of
interest to transcribers of discourse. OConnell and Kowal (1995a, 1999, Kowal
and OConnell 2004) suggest that transcribed behaviors can be categorized as belonging to one of four classes of features. Verbal features refer to the words themselves (i.e., what was said), whereas prosodic features correspond to the ways in
which the words were spoken (e.g., pitch, duration, and loudness). Paralinguistic
features (such as laughter, breathing, sighing, or crying) may accompany the
spoken words, or they may occur independently. Finally, extralinguistic features
are behaviors defined as nonvocal and nonverbal (OConnell and Kowal 1999:
109), but which are nevertheless germane to the discourse being described (e.g., facial expression, gaze, and gesture). The purpose of a transcription system is to subdivide these features into a number of dimensions, which can then be labeled with a
particular code or set of codes. Transcription systems differ in their coverage of
these four features, the number of dimensions employed within each feature, and in
the specific codes that are used.
2.4.
Transcription layout
Edwards (1993b) has reviewed how transcription systems differ in the spatial arrangement of information as well as the type and level of description. Vertical arrangement, in which speakers turns are arranged sequentially (as in a script) may
be the most common, although other formats may be more informative. Representing the interaction between interlocutors in columns may be helpful in some cases,
and a partiture format, as in Ehlichs (1993) HIAT system (see section 4.6), is effective for capturing interactions with lots of simultaneous speech.
There are several different ways in which a transcriber can choose to arrange
prosodic, paralinguistic, and extralinguistic features within a transcript. One
choice, referred to by Edwards (1993b) as running text, places such information
following the words (e.g., the code laughter immediately after an utterance containing laughter). This method preserves the temporal contiguity of the information in the transcript. However, many transcription systems use an interspersed format for recording prosodic information (Edwards 1993b). Changes in
pitch, for example, can be directly mapped onto the syllables themselves, or indicated by specific codes. A third approach is to use a segment-plus-specification
(SPS) format, in which one tier provides the verbal dimension, and other tiers or
rows below the first tier provide syntactic, semantic, or pragmatic codes (Edwards
1993b). A fourth choice is referred to as utterance-plus-clarification. In this
format, utterances are broken apart and nonverbal or contextual information (e.g.,
gesture, gaze, or behavior of the speaker) appears below each speakers turn. As
660
Edwards (1993b) notes, even something as mundane as the arrangement of speakers turns can have important implications for how a transcript is analyzed and interpreted.
2.5.
Procedural issues
One issue all researchers in pragmatics can agree on is that transcription is very
labor intensive. MacWhinney (2001), for example, has estimated that it can take
over ten hours to transcribe one hour of discourse. Consequently, it makes sense to
transcribe as few dimensions as possible for the purposes at hand. For example, if
one wanted to study the use of discourse markers such as uh and um (Clark and
Fox Tree 2002), it might make little sense to transcribe head movements of the listener. However, such choices are crucial, for as Ochs (1979) and many others have
argued, a transcription system is not theory neutral: it already reflects the beliefs
and biases of the researcher. By choosing not to transcribe a particular dimension,
the researcher has implicitly decided that the dimension plays no role in the phenomenon in question. Considering the infinity of pragmatically salient dimensions
that the researcher could transcribe, the choice of the dimensions themselves is already something of a compromise. To put it another way, transcription is always
subjective and interpretive to some degree, and transcription inevitably leads to
data reduction. These issues are of fundamental importance, but are beyond the
scope of this chapter to do them justice.2
2.6.
The role of context
Unless it is occurring under unnatural laboratory conditions, face-to-face interaction is deeply embedded within a social context, involving dimensions that may
not be readily apparent to the transcriber. The relationship that exists between the
conversational participants may be crucial to the interpretation of the interaction
and should be specified when possible. However, assessing the relevant parameters, such as the amount of shared common ground (Clark 1996), may be quite
difficult. Most transcribers have not paid much attention to specifying context, although some of these issues have been explored by Cook (1995), Norris (2004),
and Bucholtz (2007b). Sensitivity to these issues can make for a more informative
transcript: acoustically, there is no difference in transcribing a brief absence of
speech as a pause or as shocked silence, but clearly the latter provides more information (perhaps at the expense of complete objectivity).
2.7.
Technological considerations
Technological developments have greatly impacted the way in which researchers

record conversational interaction. Most researchers now employ digitally recorded
661
video, which makes it possible to record acoustic, facial, and gestural information
(a helpful review of these issues may be found in Bavelas, Kenwood, and Phillips
2002).
An important technological issue with regard to transcription is the conversion
of transcription systems into mark-up languages that can be interpreted by computer programs. Such conversion is necessary to allow efficient data storage and
retrieval, but as with other aspects of transcription, a number of tradeoffs are involved (see Leech, Myers, and Thomas 1995, for a variety of such examples).
A variety of software packages exist for use in the creation of a transcript. These
programs can lessen some of the tedium involved in repeatedly playing a troublesome section of speech, or temporally aligning verbal and extralinguistic features.
Examples include VoiceWalker, described as a discourse transcription utility (Du
Bois 2006: 1), Anvil, an annotation program (Kipp 2003), and Transana (Woods
2007), an analysis package available at http://www.transana.org. Other widely used
and free software packages include ELAN, http://www.lat-mpi.eu/tools/elan/ and
Praat, http://www.praat.org. Programs that can assist in the coding of gestural components of language also exist (Neidle, Sclaroff and Athitsos 2001).
3.
Design principles
A number of theorists and researchers have made proposals concerning what an

ideal transcription system should encompass. These ideas will be reviewed chronologically in order to show their development over time.
Du Bois (1991) proposed five maxims for designing transcription systems and
23 design principles that follow from these maxims. The maxims are category definition (e.g., use categories that are explicit, necessary, and sufficient); accessibility
(e.g., use familiar and easily learned notations); robustness (e.g., avoid invisible or
fragile contrasts); economy (e.g., avoid verbose notations, use space meaningfully); and adaptability (e.g., allow for seamless integration of user-defined codes
and transcription categories).
Edwards (1993b) argued that the creators of transcripts should keep in mind issues of category design, readability, and computational tractability. By category
design, she means that the dimensions must be systematically discriminable,
exhaustive, and contrastive. Readability refers to placing related events in close
proximity, the visual separation of unlike elements, time-space iconicity, logical
priority, mnemonic marking, and efficiency and compactness. Computational tractability refers to systematicity and predictability in encoding dimensions. Failure
to consider computational tractability can lead to the underselection or overselection of instances from a transcription. For example, a search of a transcript for instances of going to would not necessarily identify instances of gonna, while a
search for the verb bear might also snare the ursine variety. Edwards (1995) later
662
expanded on these concerns, and also discussed issues of validity and reliability
when computerized language archives are used.
OConnell and Kowal (1994, 1995a, 1999, 2008) point out that the ultimate
purpose of a transcription (i.e., the purpose of the researchers and the readership of
a transcript) must be kept in mind. Second, these authors argue against inserting
transcription symbols within words to indicate prosodic features, because doing so
impairs the legibility of the transcript. A third point concerns the consistency of
notation: each symbol should encode only one dimension, and conversely, each dimension should be encoded by only one symbol. In addition, conventional typographic elements such as ampersands and ellipses should not be used as transcription symbols, because their primary purpose is already deeply ingrained in the
minds of readers. Finally, the authors argue that measures of continuous variable
dimensions, such as amplitude and time, must be made with accurate equipment,
and not subjectively.
Dressler and Kreuz (2000) encouraged the developers of transcription systems
to keep in mind seven design principles: (1) specificity (the tradeoffs of broad versus narrow transcription); (2) universality (not linking the conventions to a particular language, such as s for softly); (3) consensus (using symbols as others have
used them in the past); (4) transparency (using intuitive symbols, such as a rising
line, /, to indicate rising intonation); (5) parsimony (the use of a small number of
codes); (6) conventionality (using codes that can be easily located on a keyboard);
and (7) extensibility (the system should be open-ended to allow new dimensions to
be transcribed).
Mller and Damico (2002), approaching these issues from the perspective of
clinical linguistics and phonetics, proposed six guiding principles that share much
in common with the points already raised. However, for dealing with the complexities of disordered speech, they stress the importance of flexibility to ensure authenticity and individuality (Mller and Damico 2002: 312).
4.
Review of transcription systems
Many researchers choose to reinvent the wheel and create their own notational systems when transcribing their data. Although this may be justifiable in some cases,
it is rarely necessary. There are, in fact, many disadvantages to this approach. An
ad hoc system is likely to be less comprehensive and may be employed inconsistently. In addition, if examples are included in published research, the use of a new
system requires some mastery on the part of editors, reviewers, and readers. Finally, a transcription key must be provided, which wastes the resource of journal
space (Dressler and Kreuz 2000).
The list of transcription systems provided here is not intended to be comprehensive; as noted above, many systems have been developed and used only once.
663
In addition, some well-known systems are not particularly comprehensive, whereas others have been employed in relatively small geographic regions or for only
one language. Therefore, the goal has been to describe briefly the systems that are
(a) widely employed, (b) in current use or historically important, and (c) reasonably comprehensive. Most of these systems have been developed to transcribe
prosodic and paralinguistic features; the transcription of extralinguistic features,
such as gesture and eye gaze, will be reviewed in section 5. Finally, each transcription system is listed by name (if it has one), or by reference to researchers and publication(s) that describe the system in detail.
4.1.
Jeffersonian Transcription System
Gail Jeffersons transcription system (Sacks, Schegloff, and Jefferson 1974; Schenkein 1978; Atkinson and Heritage 1984; Jefferson 2002, 2004) has been widely employed and refined over a thirty-year period, and has become a de facto standard in
the field of conversation analysis (often referred to as simply CA). The journal
Research on Language and Social Interaction, for example, uses the Jeffersonian
notation as its default transcription system.
In its 1978 formulation (Schenkein 1978), the Jeffersonian system provided
about 18 codes for tracking seven different categories of conversational phenomena: simultaneous utterances, overlapping utterances, contiguous utterances, intervals within and between utterances, characteristics of speech delivery, transcriptionist doubt, and other transcript symbols. In a later formulation (Atkinson and
Heritage 1984), about eight codes, such as shifts in intonation and quieter talk,
were added, as well as the extralinguistic dimensions of gaze direction and applause. More recent formulations (e.g., Jefferson 2002) have introduced a handful
of other codes to mark slower speech or suppressed laughter. A helpful discussion
of the issues involved in employing this system may be found in Psathas and Anderson (1990).
4.2.
Ochs
Following the lead of the seminal paper of Sacks et al. (1974), Elinor Ochs (1979)
proposed a transcription system for verbal and nonverbal features. For verbal features, she proposed the coding of eight dimensions: utterance boundary, latching,
pause length, overlap, self-interruption, intonation or prosodic quality, audible
breathing, and metatranscription. Four additional dimensions were proposed for
nonverbal features: changes in gross motor activity, eye gaze, gestures, and body
orientation. In total, the system uses about 35 codes (for this and other systems, an
exact number is difficult to report, since some codes can be used iteratively in
Ochs system, for example, lengthened syllables are indicated with a colon, and
additional colons can be used to indicate additional beats in time).
664
4.3.
London-Lund Corpus transcription
The London-Lund Corpus of Spoken English (or LLC) (Svartvik and Quirk 1980,
Svartvik 1990) is of considerable historical importance. The project began in London in 1959 and Lund, Sweden in 1975, and represents one of the first attempts to
gather a comprehensive corpus of spoken English. It was derived from the Survey
of English Usage (or SEU), a corpus which consists of a million words. Half of this
total was drawn from spoken English.
The LLC uses a transcription system that principally denotes prosodic features.
It stems from the British School of intonation analysis (see Kingdon 1958,
OConnor and Arnold 1961). The nucleus, or main stressed syllable that has a
clearly perceptible movement of pitch, is divided into seven dimensions (e.g., fall,
level, fall-rise), and there are four codes for the booster, or range of pitch (e.g.,
higher than preceding syllable, very high). Stress is denoted as normal or heavy.
Pauses of varying durations, as well as simultaneous talk and laughter are coded as
well. Although this system is not as comprehensive as some others, it has been
widely employed, and may be sufficient for researchers whose primary interest is
prosody.
4.4.
Tannen
Deborah Tannens (1984/2005) work is well known in sociolinguistics, and her

system has frequently been employed by later researchers. Her system codes for
pauses, stress, pitch, intonation, vowel lengthening, and overlapping speech. Amplitude is described using six codes drawn from musical notation (e.g., piano, fortissimo), and appears under the transcription line. Brackets are used to demarcate
paralinguistic or extralinguistic information (e.g., [laughter]). The system uses
about 30 codes altogether.
4.5.
Discourse Transcription (Du Bois)
The Discourse Transcription system proposed by John Du Bois (Du Bois 1991; Du
Bois et al. 1992, 1993), and often referred to simply as DT, has been adopted by
many researchers and takes into account Du Bois design principles outlined
above. The system includes codes for pauses of various kinds, overlapping sequences, vocal quality, and utterance boundaries. In addition, it includes several
codes for prosody (primarily accent and pitch), transcribers comments, and even
smile quality. The Discourse Transcription system uses about 40 codes.
4.6.
665
HIAT (Ehlich)
HIAT, or Halbinterpretative Arbeitstranskriptionen (Heuristic Interpretative Auditory Transcription, Ehlich 1993) represents a movement away from standard orthography, which may lead to a loss of important information. A system of literary
transcription is proposed instead. This system makes use of symbols from the International Phonetic Alphabet, and uses vertical space on the page to represent
simultaneous events, much like a musical score. The coding of intonation is represented in a similar way. The system excels at tracking multiparty conversations and
overlapping speech, although such transcriptions require a great deal of space.
Computer programs designed to facilitate transcription into the HIAT system are
available.
4.7.
Gumperz and Berenz
Gumperz and Berenz (1993) approach transcription from a sociolinguistic perspective, emphasizing the situated interpretations of the conversational participants. As they put it, our main goal is to reveal the functioning of communicative
signs in the turn-by-turn interpretation of talk, not to record everything that can be
heard or to provide exact measures of duration and pitch (Gumperz and Berenz
1993: 119). The system uses about 22 codes, and may be attractive to researchers
who desire a broader approach to transcription.
4.8.
GAT (Selting et al.)
The Gesprchsanalytisches Transkriptionssystem (GAT) (Selting, Auer, Barden,

Bergmann, Couper-Kuhlen, Gnthner, Meier, Quasthoff, Schlobinski and Uhmann
1998) was created to help reduce the hitherto often unmotivated variation in transcripts (Selting et al. 1998: 91), and includes 14 dimensions and about 50 codes.
Like Tannens, this system provides many gradations for amplitude, using nomenclature derived from musical notation (e.g., piano, forte, crescendo, and diminuendo).
4.9.
CHILDES and CHAT (MacWhinney)
The Child Language Database Exchange System (CHILDES) was begun in 1981
in an attempt to gather together transcripts of child language (MacWhinney 2000).
The project, established by Brian MacWhinney, has grown and evolved over time
and now includes adult interactions in the TalkBank Project. Developments in
computer technology have greatly enhanced the utility of this resource for language researchers. The transcripts themselves are freely available on the Internet
(http://talkbank.org/). In addition, tools for coding and analyzing these corpora
666
have been devised. Codes for the Human Analysis of Transcripts, or CHAT, is the
transcription system and coding format, and CLAN (for Computerized Language
Analysis) is the software tool developed to create and analyze CHILDES transcripts. The current version of CHAT (MacWhinney 2000) provides researchers
with extensive sets of codes for use in transcription, and even accommodates other
notational schemes, such as the Jeffersonian Transcription System described in
section 4.1. However, the sophistication of this system may also be its principal
weakness because researchers may need to devote a considerable amount of time
and effort to mastering its intricacies.
4.10.
Dressler and Kreuz
Dressler and Kreuz (2000) reviewed 24 papers employing transcription that

appeared in one journal (Discourse Processes) over a five-year period. They
found that 21 codes could accommodate the majority of the researchers dimensions, and grouped these dimensions into five classes: intonation, temporal features, intensity, breathing, and transcribers comments. Although this system employs some higher-level categories, such as backchannel communication and
paralinguistic behavior, no attempt was made to include finer distinctions.
4.11.
Powers
Powers (2005), an anthropologist, produced a transcription handbook to be used by

ethnographers. Not surprisingly, therefore, the focus is somewhat different than for
the other systems described here. Specifically, Powers system codes for a smaller
number of dimensions (about 18), and does not include notations for intonation or
breathing. On the other hand, this system explicitly accommodates a number of dimensions of paralinguistic and extralinguistic features, such as weeping, reported
speech, and irony.
5.
Transcription of extralinguistic features
Communication in face-to-face dialogue is not limited to spoken words. Visible actions such as facial expressions and gestures can serve to reinforce words and decrease ambiguity in interpretation.
Bavelas and Chovil (2000) argue that visible actions are only important when
they are part of a communication; for example, scratching ones knee to emphasize
a conversation about a rash one had two days ago would be a communicative gesture, but the same action during a conversation about the price of milk would not.
The authors call these communicative actions visible acts of meaning (Bavelas
and Chovil 2000: 165), and include among them facial displays such as eyebrow
667
raises, hand gestures such as circular motions to depict a circle, and communicative body movements such as shrugging ones shoulders.
Bavelas and Chovil (2000) outline four characteristics that define a visible act
of meaning: (1) the action must occur in face-to-face dialogue and be reduced
when the receiver of the action cannot see the action, (2) the action must stand as a
symbol for something that is not physically present at the moment, (3) the meaning
of the action must be expressed either in words or by a demonstration that the receiver uses the information, and (4) the action must be integrated with the spoken
dialogue. The research questions at hand dictate how these visible actions are transcribed. Facial expressions can be transcribed either as physical actions or as
meaning-based actions (Bavelas, Kenwood, and Phillips 2002).
Several researchers have used Ekman and Friesens (1978) Facial Action
Coding System (FACS), a transcription system based on physical actions that utilizes 44 action units such as head turn right and lip stretcher, several of which
are coded to varying degrees of intensity.3 Chovil (1989) developed a meaningbased system to contrast with Ekman and Friesens (1978) physical transcription
system. Her system uses descriptions of the facial expression as a whole, such as
sadness and skepticism. Bavelas, Kenwood, and Phillips (2002) argue that this
meaning-based approach may not only be more useful for discourse research, but
also less time-consuming to researchers, and indeed Chovil (1989) demonstrated a
higher interrater reliability than FACS. Some might argue, however, that a more
subjective system reduces validity.
These two extremes musculature analysis and subjective ratings may not be
helpful for a variety of researchers. For those looking for a middle ground, Louwerse et al. (2007) devised an attractive alternative. Louwerse et al. (2007) used a
subset of Ekman, Friesen, and Hagers (2002) Facial Action Coding Scheme standard, coding just 20 facial movements that were of interest for their research questions. Other researchers may wish to employ this system or a different subset from
Ekman et al. (2002), based on their own particular research interests.
5.1.
Coding of gesture
The transcription of gestures presents additional challenges to the discourse researcher, because gestures occur simultaneously with talk and some means of mapping the two in time must be considered (see Goldin-Meadow 2003). The following is a review of a subset of gestural coding systems.
5.1.1.
Ochs
Elinor Ochs (1979) system, described in 4.2, includes five codes for gestures like
pointing, holding up, and offering.
668
5.1.2.
Schegloff
In Emanuel Schegloffs (1984) analysis of deictic gestures, he proposed indexing

hand and limb movements on a line above the transcription of the words being uttered. The system utilizes eight codes, denoting, for example, the onset of movement, maximum extension, and pointing, as well as temporal elements.
5.1.3.
Bull
Peter Bull (1987, 1989) proposed a Body Movement Scoring System, in which
body contact and object contact are described in terms of (1) the body part making
the motion, (2) the type of motion, and (3) the body part or object with which contact is made. One attractive aspect of this system is that its practitioners have been
able to achieve high interrater reliability (Bull and Connelly 1985).
5.1.4.
Ehlich
The HIAT system (Ehlich 1993), described in 4.6 above, includes 25 codes for referring to parts of the head, hands, arms, legs, and body.
5.1.5.
CoGesT (Gut et al.)
CoGesT, or Conversational Gesture Transcription system (Gut, Looks, Thies, Trippel and Gibbon 2002), is an attempt to create a notational system based on distinctions between categories of gestural form and function. The system makes distinctions on a variety of dimensions, such as form, phase, location, and directionality.
Specific examples include hand shapes, repetitions, and speed (Gut et al., 2002).
5.1.6.
McNeill
Susan Duncan has developed a coding manual that has been employed by David
McNeill and his collaborators (McNeill 2005). She suggests making eight passes
through the interaction to be analyzed, and in addition to acoustic and prosodic dimensions described earlier, adds the categories of handedness, hand orientation,
hand position, and phases (i.e., points in the gesture process).
5.2.
Eye gaze
The eye gaze of interlocutors can signal important pragmatic features during an interaction. Speakers, for example, tend to establish eye contact with their partners at
the end of a turn (Levelt 1993). A number of the systems described in section 4
contain at least some codes for eye gaze. An example would be the system pro-
669
posed by Ochs (1979), which provides six codes for looking up, down, left, right,
and towards and away from the camera. These codes can be paired with the person
or object being looked to. Damico and Simmons-Mackie (2002) have proposed a
system in which a layer of gaze and gesture information can be mapped onto a base
layer of broader transcription. This proposal is attractive because it allows extralinguistic features to be represented separately from the prosodic and paralinguistic features of discourse.
5.3.
Body posture and orientation
Speakers and listeners rarely remain static during an interaction, although deciding
which body movements are pragmatically salient may be difficult to determine.
Most of the systems described earlier could accommodate such meaningful movements as part of the transcribers comments. Ochs (1979) suggests using a U-shape
to indicate the direction of a speakers pelvis.
6.
Child language transcription
Interpreting the language of adults is difficult enough, yet child language researchers must deal with all these issues and more. A good overview is provided by
Bloom (1993), who proposes a model system for the computer-aided transcription
of the speech of children. She highlights two issues in particular: the biases and
distortions that may be introduced by the observer, and the massive amount of data
reduction, from the recording of the interaction to the transcription process itself.
The conversion of child language into forms that can be accessed electronically
has also been an issue in the transcription literature. Edwards (1992) proposed four
principles for the use of such archived data, which are similar to the design principles discussed in section 3. However, one of her suggestions, the consistent
coding of the data, has been somewhat contentious. Edwards (1993a) noted that the
use of novel variations such as falld and falled, might cause one or the other to
be overlooked in an electronic search for such instances. She argues that this is an
important issue because many forms used by children are rather rare. Her concerns
were expressed with regard to the early forms of CHAT, described above (a discussion of these issues may be found in MacWhinney and Snow 1992).
7.
Signed language transcription
The transcription of sign languages, such as American Sign Language, presents its
own set of issues and complexities. Hoiting and Slobin (2002), inspired in part by
the CHILDES system (see section 4.10), have proposed a scheme to address these
670
issues. This scheme is referred to as the Berkeley Transcript System (BTS), and
BTS transcripts are available via CHILDES (MacWhinney 2000).
8.
Transcribing cognitively impaired individuals
The difficulties involved in transcribing the interactions of adults and children may
pale in comparison to reproducing the productions of those with cognitive impairments. Ball and Rahilly (2002) make some suggestions for transcribing the prosodic features of disordered speech, and propose a scheme that is similar to the
HIAT system reviewed in section 4.6 (Ehlich 1993). TalkBank (see 4.9) also includes a section called Aphasia Bank for the dissemination of transcriptions of aphasic speech.
Haravon, Obler, and Sarno (1994) present a system for analyzing the discourse
of those with brain injury. They suggest that their approach has utility for studying
the productions of aphasics and those suffering from Alzheimers disease. Their
approach is notable in that it explicitly takes into consideration pragmatic issues
(in addition to morphology and syntax).
Mller and Guendouzi (2002) propose a multilayered approach in their system
for transcribing the discourse of Alzheimers patients. Specifically, they recommend employing a baseline or orthographic layer, a layer addressing prosody and
voice issues, and a discourse layer. The codes they use are similar to those in other
transcription systems described above, but the multiple layers provide more clarity
and allow the reader a better chance of making sense of the disordered speech. This
approach is taken even further by Mller and Damico (2002), who propose six
layers: in addition to the levels already described, they add gaze and gesture, speech
(phonetic transcription), and clinical analysis (analysis of specific behaviors).
9.
Critiques of transcription systems
Clearly, the most important attribute of a transcription system is the capability to

recreate an interaction with a high degree of fidelity. The degree to which this fidelity is achieved will depend on many factors that exist outside of the system
being employed. The experience level and the care taken by the transcriber are crucial, since even small errors can completely change the perceived meaning of an
utterance (Easton, McComish, and Greenberg 2000). In addition, once an interaction has been transcribed according to one system, it may be difficult to transfer it
into a different system (Allwood, Henrichsen, Grnqvist, Ahlsn, and Gunnarsson
2005).
In a series of empirical papers, Daniel OConnell and Sabine Kowal have explored a number of issues related to the validity, generalizability, and objectivity of
671
transcription systems in current use. Their pessimistic conclusion is that transcription itself is a limited and defective device (OConnell and Kowal 2008: 93,
emphasis in the original). Although this gloomy assessment may seem overstated,
it is a conclusion that they have come to as the result of their research, which is
summarized below as a cautionary tale for the enterprising transcriber.
As a starting point, OConnell and Kowal (1999) addressed the issue of standardization in transcription notation. In a review of three widely used transcription
systems, they found that a majority of the dimensions were used to transcribe prosodic features, whereas codes for extralinguistic features made up between zero and
22 % of the total for each system. Their conclusion, however, is that standardization is not practical, or even warranted, given the diversity of behaviors that researchers are interested in.
If standardization of transcription systems is not a realistic goal, then surely at
least the reproducibility of transcripts is achievable. However, OConnell and
Kowal (2000) found that reproductions of transcripts in textbooks had, on average,
an error rate of one change per 6.6 syllables. They attribute this high error rate to
the density and relative unfamiliarity of transcription systems, which overload the
scholars and typesetters who reproduce the examples.
The idea of conceptual overload was further explored in a study by Romero,
OConnell, and Kowal (2002). They asked undergraduate participants to reproduce
a 21-syllable question asked by a news reporter. Participants were assigned to a
variety of conditions in which they were provided with only the audio recording,
with an ordinary transcription (verbal features only), or with a transcript that had
been generated using one of three widely employed transcription systems in which
dimensions of prosodic features were explicitly coded. The participants task was
to reproduce the news reporters prosody as closely as possible. When the participants productions were compared to the original, it was found that only one of the
three transcriptions yielded reproductions were better than for those participants
who heard the original recording. In general, the participants found the prosodic
codes difficult to interpret.
Finally, OConnell and Kowal (1995b), in their review of five of the transcription systems mentioned above, conclude that all of these notational schemes violate, to some degree, the seven design principles proposed in OConnell and Kowal
(1994; see section 3).
It is also worth noting that the type of discourse can present considerable problems for transcribers. Lindsay and OConnell (1995) have shown that the fragmentary nature of spontaneous speech filled with incomplete sentences, hesitations,
and overlapping speech can be particularly troublesome to transcribe because of
its complexity (see also Bucholtz 2007b).
Given the tedium of transcribing long stretches of video or audiotape, it should
come as no surprise that such tasks are frequently assigned to graduate or even
undergraduate students with little background in theories of discourse or training
672
in transcription. Some of the issues surrounding the use of such transcribers, such
as issues of training, have been described by Tilley (2003).
The fidelity of a given transcript to a particular notation system can be assessed
by comparing the work of two (or more) transcribers who have independently
applied the system to the same stretch of discourse. The measurements can range
from simple measures of agreement to more sophisticated approaches, such as
Cohens kappa, which controls for chance performance (Cohen 1960). A tradeoff
exists between the number of dimensions employed by a particular transcription
system and a measure of interrater reliability.4 It is worth noting that some researchers have been critical of the quest to achieve high reliability, because putative errors may in fact provide important information (Pye, Wilcox, and Siren
1988).
10.
Conclusions
The range of issues and choices that confront the would-be discourse transcriber
may seem overwhelming. In reality, however, any research project involves a variety of choices and trade-offs, and viewed from this perspective, the selection of a
transcription system is no different from the choice of a statistical test. In both
cases, the ultimate goal is to illuminate the underlying systematicity that exists
within the data, and there may be a variety of legitimate ways to achieve this end.
Furthermore, even though transcription can be very labor intensive, it is possible to
find the process enjoyable (Bird 2005). It is our hope that the information we have
presented can provide guidance for those who wish to explore these issues in
greater depth.
Notes
1. Partial support for the preparation of this chapter was provided by a Center of Excellence
grant from the state of Tennessee to the Department of Psychology at the University of
Memphis, and by a grant from the Human and Social Dynamics program of the National
Science Foundation (BCS 0826825). Max Louwerse, Gina Caucci, and Wolfram Bublitz
provided helpful comments on earlier drafts of this chapter. Comments and questions
concerning this chapter may be directed to Roger Kreuz (rkreuz@memphis.edu).
2. Please refer to Bucholtz (2000, 2007a, 2007b); Green, Franquiz, and Dixon (1997); Jaffe
(2007); Mishler (1991); Preston (1982); Roberts (1997); and Tilley (2003) for further discussion and analysis.
3. For an evaluation of the system, see Sayette, Cohen, Wertz, Perrott and Parrot (2001).
4. For a more extended discussion, see Roberts and Robinson (2004) and Stelma and Cameron (2007).
673
References
Allwood, Jens, Peter J. Henrichsen, Leif Grnqvist, Elisabeth Ahlsn and Magnus Gunnarsson
2005
Transliteration between spoken language corpora. Nordic Journal of Linguistics 28: 536.
Atkinson, J. Maxwell and John Heritage
1984
Transcription notation. In: J. Maxwell Atkinson and John Heritage (eds.),
Structures of Social Action: Studies in Conversation Analysis, ix-xvi. Cambridge: Cambridge University Press.
Ball, Martin J. and Joan Rahilly
2002
Transcribing disordered speech: The segmental and prosodic layers. Clinical
Linguistics and Phonetics 16: 329344.
Bavelas, Janet B. and Nicole Chovil
2000 Visible acts of meaning: An integrated message model of language in face-to-face
dialogue. Journal of Language and Social Psychology 19: 163194.
Bavelas, Janet B., Christine Kenwood and Bruce Phillips
2002
Discourse analysis. In: Mark L. Knapp and John A. Daly (eds.), Handbook of
Interpersonal Communication. 3rd ed., 102129. Thousand Oaks, CA: Sage
Publications.
Bird, Cindy M.
2005
How I stopped dreading and learned to love transcription. Qualitative Inquiry
11: 226248.
Bloom, Lois
1993
Transcription and coding for child language research: The parts are more than
the whole. In: Jane A. Edwards and Martin D. Lampert (eds.), Talking Data:
Transcription and Coding in Discourse Research, 149166. Hillsdale, NJ:
Lawrence Erlbaum.
Bucholtz, Mary
2000
The politics of transcription. Journal of Pragmatics 32: 14391465.
Bucholtz, Mary
2007a
Variation in transcription. Discourse Studies 9: 784808.
Bucholtz, Mary
2007b
Reply: Variability in transcribers. Discourse Studies 9: 837842.
Bull, Peter E.
1987
Posture and gesture. Oxford: Pergamon.
Bull, Peter E.
1989
Psychological approaches to transcription. In: Derek B. Roger and Peter E.
Bull (eds.), Conversation: An Interdisciplinary Approach, 150165. Bristol:
Bull, Peter E. and Gerry Connelly
1985
Body movement and emphasis in speech. Journal of Nonverbal Behavior 9:
169187.
Chovil, Nicole
1989
Communicative functions of facial displays in conversation. Unpublished doctoral dissertation, Department of Psychology, University of Victoria, Victoria,
British Columbia, Canada.
674
Cohen, Jacob
1960
A coefficient of agreement for nominal scales. Educational and Psychological
Measurement 20: 3746.
Clark, Herbert H.
1996
Clark, Herbert H. and Jean E. Fox Tree
2002
Using uh and um in spontaneous speaking. Cognition 84: 73111.
Cook, Guy
1995
Theoretical issues: Transcribing the untranscribable. In: Geoffrey A. Leech,
Greg Myers and Jenny Thomas (eds.), Spoken English on Computer: Transcription, Mark-up and Application, 3553. New York: Longman.
Damico, Jack S. and Nina N. Simmons-Mackie
2002
The base layer and the gaze/gesture layer of transcription. Clinical Linguistics
& Phonetics 16: 317327.
Dressler, Richard A. and Roger J. Kreuz
2000
Transcribing oral discourse: A survey and a model system. Discourse Processes 29: 2536.
Du Bois, John W.
1991
Transcription design principles for spoken discourse research. Pragmatics 1:
71106.
Du Bois, John W.
2006
VoiceWalker: A discourse transcription utility. University of California Regents.
Du Bois, John W., Susanna Cumming, Stephan Schuetze-Coburn and Danae Paolino
1992
Discourse transcription. Santa Barbara Papers in Linguistics 4: 1225.
Du Bois, John W., Stephan Schuetze-Coburn, Susanna Cumming and Danae Paolino
1993
Outline of discourse transcription. In: Jane A. Edwards and Martin D. Lampert
(eds.), Talking Data: Transcription and Coding in Discourse Research, 4589.
Hillsdale, NJ: Lawrence Erlbaum.
Easton, Kristen L., Judith Fry McComish and Rivka Greenberg
2000
Avoiding common pitfalls in qualitative data collection and transcription.
Qualitative Health Research 10: 703707.
Edwards, Jane A.
1992
Computerized methods in child language research: four principles for the use
of archived data. Journal of Child Language 19: 435458.
Edwards, Jane A.
1993a
Perfecting research techniques in an imperfect world: Response to MacWhinney and Snow. Journal of Child Language, 20: 209216.
Edwards, Jane A.
1993b
Principles and contrasting systems of discourse transcription. In: Jane A. Edwards and Martin D. Lampert (eds.), Talking Data: Transcription and Coding
in Discourse Research, 331. Hillsdale, NJ: Lawrence Erlbaum.
Edwards, Jane A.
1995
Principles and alternative systems in the transcription, coding, and mark-up of
spoken discourse. In: Geoffrey A. Leech, Greg Myers and Jenny Thomas
(eds.), Spoken English on Computer: Transcription, Mark-up and Application,
1934. New York: Longman.
675
Ehlich, Konrad
1993
HIAT: A transcription system for discourse data. In: Jane A. Edwards and Martin D. Lampert (eds.), Talking Data: Transcription and Coding in Discourse
Research 8, 123148. Hillsdale, NJ: Lawrence Erlbaum.
Ekman, Paul and Wallace V. Friesen
1978
Facial Action Coding System. Palo Alto: Consulting Psychologist Press.
Ekman, Paul, Wallace V. Friesen and Joseph C. Hager
2002
Facial Action Coding System. CD-ROM. Salt Lake City, UT: A Human Face.
Goldin-Meadow, Susan
2003
Hearing Gesture: How our Hands Help us Think. Cambridge, MA: Harvard
University Press.
Green, Judith, Maria Franquiz and Carol Dixon
1997
The myth of the objective transcript: Transcribing as a situated act. TESOL
Quarterly 31: 172176.
Gumperz, John J. and Norine Berenz
1993
Transcribing conversational exchanges. In: Jane A. Edwards and Martin D.
Lampert (eds.), Talking Data: Transcription and Coding in Discourse Research, 91121. Hillsdale, NJ: Lawrence Erlbaum.
Gut, Ulrike, Karin Looks, Alexandra Thies, Thorsten Trippel and Dafydd Gibbon
2002
CoGesT: Conversational gesture transcription system version 1.0. Universitt
Bielefeld: Technical report 1.
Haravon, Anita, Loraine K. Obler and Martha T. Sarno
1994
A method for microanalysis of discourse in brain-damaged patients. In: Ronald
L. Bloom, Loraine K. Obler, Susan De Santi and Jonathan S. Ehrlich (eds.),
Discourse Analysis and Applications: Studies in Adult Clinical Populations,
4780. Hillsdale, NJ: Lawrence Erlbaum.
Hoiting, Nina and Dan I. Slobin
2002
Transcription as a tool for understanding: The Berkeley Transcription System
for sign language research (BTS). In: Gary Morgan and Bencie Wool (eds.),
Directions in Sign Language Acquisition, 5575. Amsterdam: Benjamins.
Jaffe, Alexandra
2007
Variability in transcription and the complexities of representation, authority
and voice. Discourse Studies 9: 831836.
Jefferson, Gail
2002
Is no an acknowledgement token? Comparing American and British uses of
(+)/(-) tokens. Journal of Pragmatics 34: 13451383.
Jefferson, Gail
2004
Glossary of transcript symbols with an introduction. In: Gene H. Lerner (ed.),
Conversation Analysis: Studies from the First Generation, 1331. Amsterdam:
Benjamins.
Kingdon, Roger
1958
The Groundwork of English Stress. London: Longman.
Kipp, Michael
2003
Anvil 4.0: Annotation of video and spoken language. Saarbrcken: University
of the Saarland.
Kowal, Sabine and Daniel C. OConnell
2004
The transcription of conversations. In: Uwe Flick, Ernst von Kardoff, Ines
Steinke and Bryan Jenner (eds.), A Companion to Qualitative Research: Para-
676
digms, Theories, Methods, Practice and Context, 248252. Thousand Oaks,

CA: Sage Publications.
Kyratzis, Amy
2001
Emotion talk in preschool same-sex friendship groups: Fluidity over time and
context. Early Education & Development 12: 359392.
1995
Spoken English on Computer: Transcription, Mark-up and Application. New
York: Longman.
Levelt, Willem J. M.
1993
Speaking: From Articulation to Intention. Cambridge, MA: MIT Press.
Lindsay, Jean and Daniel C. OConnell
1995
How do transcribers deal with audio recordings of spoken discourse? Journal
of Psycholinguistic Research 24: 101115.
Louwerse, Max M., Nick Benesh, Mohammed E. Hoque, Patrick Jeuniaux, Gwyneth Lewis,
Jie Wu and Megan Zirnstein
2007
Multimodal communication in face-to-face computer-mediated conversations.
Proceedings of the 28th Annual Conference of the Cognitive Science Society,
12351240. Mahwah, NJ: Lawrence Erlbaum.
MacWhinney, Brian
n.d.
TalkBank. Retrieved December 2008 from http://talkbank.org/.
MacWhinney, Brian
2000
The CHILDES Project: Tools for Analyzing Talk, Third Edition. Volume I:
Transcription Format and Programs. Mahwah, NJ: Lawrence Erlbaum.
MacWhinney, Brian
2001
From CHILDES to TalkBank. In: Margareta Almgren, Andoni Barrea, MaraJose Ezeizaberrena, Itziar Idiazabal, and Brian MacWhinney (eds.), Research
on Child Language Acquisition, 1734. Somerville, MA: Cascadilla.
MacWhinney, Brian and Catherine Snow
1992
The wheat and the chaff: Or four confusions regarding CHILDES. Journal of
McNeill, David
2005
Appendix. In: David McNeill, Gesture & Thought, 259288. Chicago: University of Chicago Press.
Mishler, Elliot G.
1991
Representing discourse: The rhetoric of transcription. Journal of Narrative
and Life History 1: 255280.
Mller, Nicole and Jack S. Damico
2002
A transcription toolkit: Theoretical and clinical considerations. Clinical Linguistics and Phonetics 16: 299316.
Mller, Nicole and Jacqueline A. Guendouzi
2002
Transcribing discourse: Interactions with Alzheimers disease. Clinical Linguistics and Phonetics 16: 345359.
Neidle, Carol, Stan Sclaroff and Vassilis Athitsos
2001
SignStream: A tool for linguistic and computer vision research on visual-gestural language data. Behavior Research Methods, Instruments, & Computers
33: 311320.
677
New York Times

1974
The White House Transcripts: Submission of Recorded Presidential Conversations to the Committee on the Judiciary of the House of Representatives by
President Richard Nixon. New York: Viking.
Norris, Sigrid
2004
Analyzing Multimodal Interaction: A Methodological Framework. New York:
Routledge.
Novick, David G., Lisa Walton and Karen Ward
1996
Contribution graphs in multiparty conversations. In: Proceedings of the International Symposium on Spoken Dialogue, 5356. (ISSD-96.) Philadelphia, PA.
Ochs, Elinor
1979
Transcription as theory. In: Elinor Ochs and Bambi B. Schieffelin (eds.), Developmental Pragmatics, 4372. New York: Academic Press.
1994
Some current transcription systems for spoken discourse: A critical analysis.
1995a
Basic principles of transcription. In: Jonathan A. Smith, Rom Harre, and Luk
Van Langenhove (eds.), Rethinking Methods in Psychology, 93105. Thousand
Oaks, CA: Sage.
1995b
Transcription systems for spoken discourse. In Jef Verschueren, Jan-Ola stman, and Jan Blommaert (eds.), Handbook of Pragmatics, 646656. Amsterdam: Benjamins.
OConnell, Daniel C., and Sabine Kowal
1999
Transcription and the issue of standardization. Journal of Psycholinguistic Research 28: 103120.
OConnell, Daniel and Sabine Kowal
2000
Are transcripts reproducible? Pragmatics 10: 247269.
2008
Communicating with One Another: Toward a Psychology of Spontaneous
Spoken Discourse. New York: Springer.
OConnor, Joseph D. and Gordon F. Arnold
1961
Intonation of Colloquial English. London: Longman.
Powers, Willow R.
2005
Transcription Techniques for the Spoken Word. Oxford: AltaMira Press.
Preston, Dennis R.
1982
Ritin fowklower daun rong: Folklorists failures in phonology. Journal of
American Folklore 95: 304326.
Psathas, George and Timothy Anderson
1990
The practices of transcription in conversation analysis. Semiotica 78: 7599.
Pullum, Geoffrey K. and William A. Ladusaw
1996
Phonetic Symbol Guide, 2nd ed. Chicago: University of Chicago Press.
Pye, Clifton A., Kim A. Wilcox, and Kathleen A. Siren
1988
Refining transcription: The significance of transcriber errors. Journal of
678
Roberts, Celia
1997
Transcribing talk: Issues of representation. TESOL Quarterly, 31: 167172.
Roberts, Felicia and Jeffrey D. Robinson
2004
Interobserver agreement on first-stage conversation analytic transcription.
Human Communication Research 30: 376410.
Romero, Catherine, Daniel C. OConnell and Sabine Kowal
2002
Notation systems for transcription: An empirical investigation. Journal of Psycholinguistic Research 31: 619631.
1974
Sayette, Michael A., Jeffrey F. Cohen, Joan M. Wertz, Michael A. Perrott and Dominic J.
Parrot
2001
A psychometric evaluation of the Facial Action Coding System for assessing
spontaneous facial expression. Journal of Nonverbal Behavior 25: 167185.
1984
On some gestures relation to talk. In: J. Maxwell Atkinson and John Heritage
Schenkein, Jim
1978
Explanation of transcript notation. In: Jim Schenkein (ed.), Studies in the Organization of Conversational Interaction, xi-xvi. New York: Academic Press.
Schiffrin, Deborah
1987
Searle, John
1975
Semantics. Volume 3: Speech Acts, 5982. New York: Academic Press.
Selting, Margret, Peter Auer, Birgit Barden, Jrg Bergmann, Elizabeth Couper-Kuhlen,
Susanne Gnthner, Christoph Meier, Uta Quasthoff, Peter Schlobinski and
Susanne Uhmann
1998
Gesprchsanalytisches Transkriptionssytem (GAT). Linguistische Berichte
173: 91122.
Shriberg, Lawrence D. and Gregory L. Lof
1991
Reliability studies in broad and narrow phonetic transcription. Clinical Linguistics and Phonetics 5: 225279.
Stelma, Juurd H. and Lynne J. Cameron
2007
Intonation units in spoken interaction: Developing transcription skills. Text &
Talk 27: 361393.
Svartvik, Jan (ed.).
1990
The London Corpus of Spoken English: Description and Research. Lund: Lund
University Press.
Svartvik, Jan and Randolph Quirk
1980
A Corpus of English Conversation. Lund: Lund University Press.
Tannen, Deborah
1984/2005 Conversational Style: Analyzing Talk Among Friends. Oxford: Oxford University Press.
679
Tilley, Susan A.
2003
Challenging research practices: Turning a critical lens on the work of transcription. Qualitative Inquiry 9: 750773.
Woods, David K.
2007
Transana. Retrieved December 2008 from http://www.transana.org.
About the authors
Gisle Andersen is Associate Professor in the Department of Professional and

Intercultural Communication (Norwegian School of Economics and Business Administration). He is the author of Pragmatic Markers and Sociolinguistic Variation a Relevance-theoretic Approach to the Language of Adolescents (Benjamins, 2001) and has co-authored Trends in Teenage Talk Corpus Compilation,
Analysis and Findings (Benjamins, 2002), Pragmatic Markers and Propositional
Attitude (Benjamins, 2000) and is currently co-editing Pragmatics of Society,
which is the 5th volume in the current handbook series. His research mainly
focuses on different aspects of spoken interaction, and his work deals especially
with the use of corpora for studies in pragmatics and sociolinguistics, covering
such topics as discourse markers, vague language and listenership. He has also
worked on written communication, lexicography and terminology, and the influence
of English on the Norwegian language. He has been deeply involved in various
corpus compilation projects, including COLT (The Bergen Corpus of London Teenage Language) and the Norwegian Newspaper Corpus, and has coordinated and
participated in projects within language technology and language resources. He is
participating in various projects funded by the European Commission and the Norwegian Research Council. He is also a board member of the ICAME organisation
(International Computer Archive of Modern and Medieval English).
Monika Bednarek is Lecturer in Linguistics at the University of Sydney, Australia. Her research interests include news and television discourse, corpus-based discourse analysis, ecolinguistics, and the linguistic expression of opinion and emotion. She is the author of Evaluation in Media Discourse (Continuum, 2006),
Emotion Talk Across Corpora (Palgrave Macmillan, 2008) and The Language of
Fictional Television: Drama and Identity (Continuum, 2010). Other publications
on these and related research areas include journal articles in Discourse & Communication, Functions of Language, Text and Talk and the Journal of Pragmatics.
Richard Breheny is Reader in Linguistics in the Division of Psychology and Language Sciences, University College London, UK. His research is in semantics,
pragmatics, and experimental pragmatics, including online processing and pragmatic development. Recent publications include Tian, Y., Breheny, R. & H.J. Ferguson Why we simulate negated information: a dynamic pragmatic account,
Quarterly Journal of Experimental Psychology 63, 2010; Breheny, R. A new look
at the semantics and pragmatics of numerically quantified noun phrases, Journal
682
About the authors
of Semantics 25, 2008; Breheny, R., Katsos, N. & J. Williams Are scalar implicatures generated on-line by default?, Cognition 100, 2006; Breheny, R. Communication and folk psychology, Mind and Language 21, 2006.
Wolfram Bublitz is Full Professor of English Linguistics at the University of
Augsburg, Germany. He received his doctorate from the University of Hamburg
and his habilitation from the University of Trier, Germany. He has also been Full
Professor at the University of Braunschweig, Exchange Professor at Northern Illinois University and Visiting Professor at the Universities of Pittsburgh and Hamburg. His research interests concern all areas of text analysis, (constructivist) pragmatics and recently computer-mediated communication, with an emphasis on
modal and discourse particles, modality and negation, discourse topic and topical
acts, repetition and other forms of parallelism, coherence and cohesion, collocation
and semantic prosody, politeness and evaluation, context and common ground,
systems of knowledge and memory, hypertext and quoting. His monographs and
edited books include Englische Pragmatik (2nd ed. Schmidt, 2009), Metapragmatics in Use (ed. with A. Hbler, Benjamins, 2007), Coherence in Spoken and
Written Discourse: How to Create It and How to Describe It (ed. with U. Lenk and
E. Ventola, Benjamins, 1999).
Piotr Cap is Professor of Linguistics in the Institute of English, University of
dz, Poland. His research interests are in pragmatics, (critical) discourse analysis,
political linguistics, business communication, and methodology of linguistic
analysis. His monographic and (co-)edited publications include Explorations in
Political Discourse (Lang, 2002), Pragmatics Today (Lang, 2005), Legitimisation
in Political Discourse (Cambridge Scholars Publishing, 2006/2008), New Approaches to Discourse and Business Communication (Palgrave Macmillan, 2009),
and Perspectives in Politics and Discourse (Benjamins, 2010). He is Managing
Editor of International Review of Pragmatics (Brill) and member of editorial and
advisory boards of several journals in pragmatics and discourse analysis.
Elena Collavin is a Lecturer in the Department of Communication at the University
of California San Diego, USA. Her work has focused on Italian political debate and
on discourse surrounding food biotechnologies in Italy. More recently she has
studied everyday family conversations among Friulian speakers in San Giorgio di
Nogaro, in north eastern Italy. Her recent publications include Participation in
Italy: The public and genetically modified crops, Democratic Transgressions of
Law, ed. by A. Bora and H. Hausendorf (Brill, 2010); Marketing concern in food
products, IF, Journal of Italo-Finnish Studies, 2008; Tavano, A. and E. Collavin,
Pragmatiche furlane, Manul di Lenghistiche Furlane, ed. by F. Fabbro (Udine,
Forum Edizioni, 2007); Food Biotechnologies in Italy: A Social Psychological
Study (Social Psychology Department, Helsinki University, PhD dissertation, 2007).
About the authors
683
Manfred Consten is Lecturer for discourse linguistics and linguistic pragmatics at

the Institute for Germanic Linguistics, University of Jena, Germany. His main research areas are anaphoric reference, cognitive approaches to text coherence (esp.
the discourse function of demonstratives), and meta-linguistics. His doctoral thesis
(2004) suggests a unified model of anaphora and deixis as instances of domain
bound reference. He is co-editor of the volume Anaphors in Text. Cognitive, Formal and Applied Approaches to Anaphoric Reference (Benjamins, 2007). Fuller information is provided at www.textlinguistik.net.
Maeve Cooke is Professor and Head of the School of Philosophy at University
College Dublin, Ireland, and a member of the Royal Irish Academy. Her current
research focuses on re-conceptualising truth, freedom and power and on issues
in current debates on religion and politics. Her principal book publications are
Language and Reason: A Study of Habermass Pragmatics (MIT Press, 1994) and
Re-Presenting the Good Society (MIT Press, 2006). She is editor and translator of
Habermas: On the Pragmatics of Communication (MIT Press, 1998) and has published numerous articles in scholarly journals and books, mainly in the areas of
social and political philosophy.
Elizabeth Couper-Kuhlen is Finland Distinguished Professor at the University
of Helsinki, Finland. Her main research areas are situated in the field of language
use in interaction, including prosody, grammar and most recently, affect and emotion. Among her main publications are the monographs An Introduction to English Prosody (Niemeyer and Edward Arnold, 1986), English Speech Rhythm (Benjamins, 1993), Language in Time (with P. Auer & F. Mueller, Oxford University
Press, 1999) and the edited volumes Prosody in Conversation (with M. Selting,
Cambridge University Press, 1996), Cause, Condition, Contrast, Concession:
Cognitive and Discourse Perspectives (with B. Kortmann, de Gruyter Mouton.
2000), Studies in Interactional Linguistics (with M. Selting, Benjamins, 2001) and
Sound Patterns in Interaction (with C.E. Ford, Benjamins, 2004).
Saskia Daalder teaches Dutch linguistics at the Vrije University Amsterdam, The
Netherlands. Her main research areas are the pragmatics of verbal communication (Relevance Theory), the theory of grammar in the light of pragmatic theory,
and the history of twentieth-century linguistics (in particular, the Prague School
of functional linguistics and its reflections in the Netherlands). Selected publications are Conditional constructions: The special case of modern Dutch mits,
Journal of Germanic Linguistics 21, 2009; H.J. Pos (18981955): Studies over zijn
filosofie van taal en taalwetenschap (2 vols, Mnster: Nodus-Publikationen,
1999).
684
About the authors
Arnulf Deppermann is head of the Pragmatics Department at the Institut fr

Deutsche Sprache and Professor for German Linguistics at Mannheim University,
Germany. His areas of research cover conversation analysis, cognition and understanding in interaction, multimodal interaction, positioning in narratives, medical
and therapeutic interaction, construction grammar and semantics in talk-in-interaction. Main publications are Gesprche analysieren (Wiesbaden: VS, 42008);
Grammatik und Semantik aus gesprchsanalytischer Sicht (de Gruyter, 2007); Verstehen in professionellen Handlungsfeldern (with U. Reitemeier, R. Schmitt and
Th. Spranz-Fogasy, Narr, 2010); Data and transcription, (together with W.
Schtte), Handbook of Interpersonal Communication, ed. by G. Antos et al (de
Gruyter, 2008); Constructions vs. lexical items as sources of complex meanings,
Emergent constructions, ed. by P. Auer and S. Pfnder (de Gruyter, 2010); How
does cognition matter to the analysis of talk-in-interaction?, Language Sciences
33, 2011.
Christine Domke is a research assistant at the department of German studies at the
Technical University Chemnitz, Germany. She holds a doctorate in linguistics
from the University of Bielefeld with a thesis on the process of decision-making in
business meetings. Currently she is working on her habilitation about the characteristics of place bound communication in railway stations, airports and inner
cities. Her major research interests are in combining organisation theory and organisational communication based on qualitative approaches like conversational
analysis (CA), and also in the semiotics of mass media texts and recent forms of
advertising. Her publications include Besprechungen als organisationale Entscheidungskommunikation (de Gruyter, 2006), Name: Deutsch, Alter: 1200, Befund: Gesund, (with J. Kilian), Aisthesis 2009; Wissen in (Inter)Aktion (with
U. Dausendschn-Gay and Sren Ohlhus (de Gruyter, 2010).
Anita Fetzer is Full Professor of English linguistics at Wrzburg University, Germany. She received her doctorate from Stuttgart University in 1993 and her habilitation in 2003, and is currently engaged in research projects on evidentiality, the
strategic use of pronouns and attitudinal markers in political discourse, and on the
form and function of the theme zone in present-day English. She has had a series of
articles published on rejections, context, political interviews and intercultural
communication. Her most recent publications are Pragmatic and discourse-analytic approaches to present-day English, (with Karin Aijmer), Journal of Pragmatics 2008, Political Discourse in the Media (with Gerda Lauerbach, Benjamins,
2007), Lexical Markers of Common Grounds (with Kerstin Fischer, Elsevier,
2007), Context and Appropriateness (Benjamins, 2007), and Recontextualizing
Context: Grammaticality Meets Appropriateness (Benjamins, 2004).
About the authors
685
William Hanks earned the Joint PhD in Anthropology and Linguistics at The University of Chicago, 1983. Since 2000, he has held the Berkeley Distinguished
Chair in Linguistic Anthropology at University of California, Berkeley, USA. His
research bears on semantics and pragmatics, with a special focus on indexicality in
interaction, ordinary referring and ritual speech in Yucatec Maya. His relevant
books include Referential Practice, Language and Lived space among the Maya
(University of Chicago Press, 1990), Language and Communicative Practices
(Westview Press, 1995), and Converting Words, Maya in the Age of the Cross (University of California Press, 2010, winner of the Edward Sapir Book Prize, Society
for Linguistic Anthropology).
Werner Holly is Full Professor of Germanistische Sprachwissenschaft at the
Technical University Chemnitz, Germany. His main research areas are: pragmatics, linguistics of text and conversation, language in politics, language and the
media, audiovisuality. Major publications include Imagearbeit in Gesprchen
(Niemeyer, 1979); Politikersprache (de Gruyter, 1990); Politische Fernsehdiskussionen (with Peter Khn and Ulrich Pschel, Niemeyer, 1986); Der sprechende
Zuschauer (ed. with Ulrich Pschel and Jrg Bergmann, Westdeutscher Verlag,
2001); Einfhrung in die Pragmalinguistik (Langenscheidt, 2001); Fernsehen
(Niemeyer, 2004); ber Geld spricht man (with Stephan Habscheid et al.,Verlag
fr Sozialwissenschaft, 2006); Linguistische Hermeneutik (ed. with Fritz Hermanns, Niemeyer, 2007).
Yan Huang (PhD Cambridge, DPhil Oxford) is Professor of Linguistics at the University of Auckland, New Zealand. He has previously taught linguistics at the
universities of Cambridge, Oxford, and Reading, UK, where he was Professor of
Theoretical Linguistics. His main research interests are in pragmatics, especially
the pragmatics-semantics interface and the pragmatics-syntax interface including
anaphora. His books include The Syntax and Pragmatics of Anaphora (Cambridge
University Press, 1994, re-issued in 2007), Anaphora: A Cross-Linguistic Study
(Oxford University Press, 2000), and Pragmatics (Oxford University Press, 2007).
His Pragmatics is being translated into a number of languages, and the Chinese
edition and the Korean translation have already been published. He has also published a number of articles and reviews in leading international journals of linguistics. He has been invited to lecture in around 90 universities and research institutes in many countries in Europe, North America, Asia, Australasia, and North
Africa.
Axel Hbler is Professor of English Linguistics; before his retirement in 2009, he
held the chair of English Linguistics at the University of Jena, Germany. His main
research areas are Semantics, (Historical) Pragmatics and Nonverbal Communication. Main publications include Understatements and Hedges in English (Benja-
686
About the authors
mins, 1983), The Expressivity of Grammar: Grammatical Devices Expressing

Emotion across Time (de Gruyter, 1998), Das Konzept Krper in den Sprachund Kommunikationswissenschaften (A. Francke, 2001), The Nonverbal Shift in
Early Modern English Conversation (Benjamins, 2007), Introducing Metapragmatics in Use (with W. Bublitz), Metapragmatics in Use, ed. by W. Bublitz and
A. Hbler (Benjamins, 2007).
Nikola Kompa is Assistant Professor of Theoretical Philosophy at the University
of Bern in Switzerland where she teaches philosophy. She works in philosophy of
language, epistemology, metaphysics, logic and the history of philosophy. Within
philosophy of language, she is particularly interested in theories of meaning, the semantics/pragmatics-interface, the connection between meaning and categorization,
vagueness, and context-sensitivity. She is author of Wissen und Kontext (mentis,
2001) and co-editor of The Apriori and its Role in Philosophy (mentis, 2009).
Wataru Koyama is Professor of Pragmatics and Linguistics at the Graduate
School of Intercultural Communication, Rikkyo (St. Pauls) University, Tokyo,
Japan. He received his doctorate in linguistics from The University of Chicago in
2003. His particular research interests focus on language in culture, social semiotics, and socio-historical pragmatics. Some of his publications have appeared
in Journal of Pragmatics, Studies in Language, Rask, Language and Communication, and the Encyclopedia of Language and Linguistics (2nd edition).
Roger Kreuz is Professor of Psychology at the University of Memphis, USA, a
position that he has held since 1988. He is also the director of the departments
masters program. He earned his masters and doctoral degrees in cognitive psychology at Princeton University, and undergraduate degrees in psychology and linguistics at the University of Toledo. Before joining the University of Memphis faculty, he held a post-doctoral position at Duke University. His primary research
areas are discourse processing and pragmatics, nonliteral language production and
comprehension, and computer-mediated communication. With Susan Fussell, he
edited Social and Cognitive Approaches to Interpersonal Communication (Erlbaum, 1998).
Sophia Marmaridou is Professor of Linguistics at the Faculty of English Language and Literature of the University of Athens, Greece. She holds a BA in English and Greek Language and Literature from the University of Athens and a PhD
in Linguistics from the University of Cambridge. Her main research areas are Semantics, Pragmatics, Cognitive Linguistics, Construction Grammar, and Lexicography. Some of her recent publications are: Pragmatic Meaning and Cognition
(Benjamins, 2000), On the conceptual, cultural and discursive motivation of
Greek pain lexicalizations, Cognitive Linguistics 17, 2006; Cognitive, cultural,
About the authors
687
and constructional motivations of polysemy and semantic change. The case of the
Greek , Pragmatics and Cognition 18, 2010. She is one of the editors of
Pragmatics, a member of the Editorial Board of the Pragmatics & Beyond New
Series (Benjamins), and a Board Member of the International Pragmatics Association. She is also a member of the International Cognitive Linguistics Association.
Georg Meggle is Professor Emeritus of Philosophy, Logic, Anthropology, Cognitice Sciences and Ethics at the universities of Saarbrcken, Mnster and Leipzig,
Germany. His main research topics include Theory of Action, Communication and
Meaning, Collective Intentionality, Ethics and Applied Ethics, in particular of War
and Terrorism. His publications include: Grundbegriffe der Kommunikation (2nd
ed. de Gruyter, 1997); Sprache und Ethik (with G. Grewendorf, Suhrkamp, 1974);
Analytische Handlungstheorie (Suhrkamp, 1977, re-issued 1985), Handlung,
Kommunikation, Bedeutung (Suhrkamp, 1979; Sprachphilosophie Philosophy of
Language La philosophie du langage (ed. with M. Dascal, D. Gerhardus, K. Lorenz, de Gruyter, vol 1: 1992, vol. 2: 1995); Analyomen 1 Proceedings of the 1st
Conference on Perspectives in Analytical Philosophy (ed. with U. Wessels, de
Gruyter, 1994); Zum moralischen Denken (with C. Fehige, Suhrkamp, 1994, 2
vols); ANALYOMEN 2 Proceedings of the 2nd Conference on Perspectives in
Analytical Philosophy (de Gruyter, 1997, 3 vols); Actions, Norms, Values: Discussions with Georg Henrik von Wright (de Gruyter, 1998); Der Sinn des Lebens
(ed. with C. Fehlige and U. Wessels, dtv, 2000); Speech Acts, Mind, and Social
Reality (ed. with G. Grewendorf, Kluwer, 2002); Social Facts & Collective Intentionality (German Library of Sciences, 2002); Saying, Meaning, Implicating (ed.
with C. Plunze, Leipzig Universittsverlag, 2003); Terror und Der Krieg gegen ihn
(mentis, 2003); Humanitre Interventionsethik (mentis, 2004); Ethics of Terrorism
and Counter-Terrorism (ontos, 2005); Deutschland Israel Palstina. Streitschriften (Europische Verlagsanstalt, 2007).
Jacob L. Mey is Professor Emeritus of Linguistics at the University of Southern
Denmark. Previous appointments include the University of Oslo, the University of
Texas at Austin, Georgetown University, Tsukuba University, The National Language Research Institute, Tokyo, City University of Hong Kong, University of
Frankfurt, University of Campinas, University of Braslia, University of Haifa,
Shanghai International Studies University, University of Graz, University of
Science and Technology, Kuwait, as well as numerous other institutions of research and higher learning. His research interests concern all areas of pragmatics,
with an emphasis on the social aspects of language use, the pragmatic impact of
computer technologies, and the pragmatic use of literary devices. Among his most
recent publications in these areas are Pragmatics: An Introduction (Blackwell,
2001); When Voices Clash: A Study in Literary Pragmatics (de Gruyter, 2000); As
Vozes da Sociedade (The Voices of Society; Campinas, S.P.: Mercado de Letras,
688
About the authors
2004); Cognition and Technology (with B. Gorayska, Benjamins, 2004); Intention,

Common Ground and the Egocentric Speaker-Hearer (with I. Kecskes, de Gruyter,
2008). He edited the 1200-page Concise Encyclopedia of Pragmatics (Elsevier,
2009). In 1977, he founded (with Hartmut Haberland) the Journal of Pragmatics,
of which he was Editor-in-Chief until Elsevier terminated him in 2009. That same
year, he founded the new journal Pragmatics and Society (Benjamins), the first two
issues of which have appeared in 2010. He holds honorary Dr. Phil. degrees from
the Universities of Zaragoza, Spain (1993) and Bucharest, Romania (2006). In
2008, he was presented with a Life Time Award from the University of Southern
Denmark for his work in pragmatics.
Andreas Musolff is Professor of Intercultural Communication at the University of
East Anglia in Norwich, UK. He has published on the History of Political Discourse, Metaphor Theory and the History of Pragmatics. His monographs include
Metaphor, Nation and the Holocaust (Routledge, 2010), Metaphor and Political
Discourse (Palgrave Macmillan, 2004), Mirror Images of Europe (Iudicium,
2000), and Kommunikative Kreativitt. Karl Bhlers Zweifelderlehre als Ansatz zu
einer Theorie innovativen Sprachgebrauchs (1990). He is a member of the Executive Committee of the Societas Linguistica Europaea and of the editorial boards
of Aptum, Zeitschrift fr Sprachkritik und Sprachkultur, Journal of Germanic
Studies, German as a Foreign Language, Metaphor and the Social World.
Winfried Nth is Emeritus Professor of Linguistics and Semiotics at the University of Kassel, Germany, and Visiting Professor at the Catholic University of So
Paulo. His 250 articles and 27 authored or edited books are on topics of English linguistics (esp. semantic opposites), semiotic aspects of language, literature, the
image, maps, the media, systems theory, culture, and evolution. His Handbook of
Semiotics (Bloomington, 1990, transl. into Bahasa, and in a rev. transl. into German, Metzler, 2000, and Croatian) was awarded the Choice Outstanding Academic
Book prize. Among his books are: Literatursemiotische Analysen zu Lewis Carrolls Alice-Bchern (Narr, 1980), Origins of Semiosis (ed., de Gruyter, 1994), Semiotics of the Media (ed., de Gruyter, 1997), Crisis of Representation (ed. with C.
Ljungberg, de Gruyter, 2003), Imagen: Comunicacin, semitica y medios (Reichenberger, 2003), Comunicao e semitica (Hacker, 2004), Self-Reference in the
Media (ed. with N. Bishara, de Gruyter, 2007), Mediale Selbstreferenz: Grundlagen und Fallstudien zu Werbung, Computerspiel und Comics (with B. Bishara
and B. Neitzel, Halem, 2008), and Estratgias semiticas da publicidade (with L.
Santaella, Cengage, 2010).
Neal R. Norrick holds the chair of English Linguistics at Saarland University in
Saarbrcken, Germany. His research specializations include conversation, narrative, verbal humor and formulaicity. His authored publications include Conver-
About the authors
689
sational Joking: Humor in Everyday Talk (Indiana University Press, 1993) and
How Proverbs Mean (Mouton, 1985); his recent edited volumes are Phraseology:
An International Handbook of Contemporary Research, HSK vol. 28, 12 (ed. with
H. Burger, D. Dobrovolskij, and P. Khn, de Gruyter, 2007) and Humor in Interaction (ed. with Delia Chiaro, Benjamins, 2009). His monograph Conversational
Narrative: Storytelling in Everyday Talk (Benjamins, 2000) was reprinted in a paperback edition in 2010. He acts as Co-Editor in Chief of the Journal of Pragmatics, and serves on the advisory boards of the journals Discourse Processes,
Humor, International Review of Pragmatics and Text & Talk.
Monica A. Riordan earned her PhD in Experimental Psychology at the University of Memphis, USA. She earned her masters degree at the University of Memphis and her undergraduate degree at the University of Missouri-Columbia. Her
primary research interest is the expression of emotion in computer-mediated communication. Representative publications include: A survey of emotion expression
in CMC: Reasons for channel choice, (with R.J. Kreuz), Computers in Human Behavior 2010, Cues in online conversation: A corpus analysis, (with R.J. Kreuz),
Computers in Human Behavior.
Christoph Rhlemann is a full-time secondary school teacher and a part-time
Lecturer in modern linguistics at Ludwig-Maximilians-University, Mnchen, Germany. He is the author of Conversation in Context: A Corpus-driven Approach
(Continuum, 2007) and Narrative in English Conversation (Cambridge University
Press, forthcoming). He has published on different topics relating to conversational
English in edited collections and journals such as Applied Linguistics, the ICAME
Journal, the International Journal of Corpus Linguistics, and the Journal of English Linguistics. His main interests are in corpus linguistics, pragmatics and sociolinguistics. He is currently co-editing (with Karin Aijmer) a research volume on
Corpus Pragmatics, covering the fast-growing field of pragmatic studies based on
corpus methods. Further, he is involved in the construction, annotation and analysis of the Narrative Corpus, a corpus of conversational narratives extracted from
the British National Corpus.
John Saeed is a Fellow of Trinity College Dublin, Ireland, and Associate Professor of Linguistics in its School of Linguistic, Speech and Communication
Sciences. His PhD at the School of Oriental and African Studies, London was on
information structure in Somali and his subsequent research has concerned the
grammar and pragmatics interface in a number of languages, most recently in Irish
Sign Language. His books include Somali Reference Grammar (2nd ed. Dunwoody
Press, 1993), Somali (Benjamins, 1999) and Semantics (3rd ed. Wiley-Blackwell,
2009).
690
About the authors
Monika Schwarz-Friesel is Full Professor of General Linguistics at the Technical University Berlin, Germany. Her research focuses on cognitive semantics/pragmatics, the interaction of language, cognition and emotion and discourse comprehension processes, especially the problem of anaphor resolution and coherence.
She is the author and editor of several books on semantics (Kognitive Semantiktheorie und neuropsychologische Realitt, Niemeyer, 1992; Semantik. Ein Arbeitsbuch, Narr, 5th ed. 2007, with J. Chur; Metapher, Winter, 2007, with H. Skirl), cognitive linguistics (ed., Cognitive Semantics, Narr, 1994; Einfhrung in die kognitive
Linguistik, Francke, 3rd ed. 2008) and discourse linguistics (Indirekte Anaphern in
Texten, Niemeyer, 2000; Sprache und Emotion, Francke, 2007; Anaphors in Text,
ed. with M. Consten and M. Knees, Benjamins, 2007; Aktueller Antisemitismus
ein Phnomen der Mitte, ed. with E. Friesel and J. Reinharz, de Gruyter, 2010).
Name index
691
Name index
A
Aarsleff, H. 152
Aarts, J. 592
Abbott, C. 333
Abelson, R.P. 480
Abraham, W. 599
Abse, D.W. 246
Adolphs, S. 539, 542543, 547, 549, 609,
648650
Adorno, T.W. 303
del, A. 115, 589, 609
Agha, A. 143, 390
Agirre, E. 464
Ahlsn, E. 670
Aijmer, K. 39, 592594, 597602, 651
Aitchison, J. 443
Akman, V. 31, 3435
Algeo, J. 591
Allan, W. S. 602
Allwood, J. 194, 670
Alston, W.P. 69, 221, 379
Altenberg, B. 591595
Altmann, G.T.M. 568, 574
Andersen, C. 194
Andersen, G. 17, 26, 364, 547, 590,
593, 595, 597599, 605606, 610,
636637
Anderson, A. 572
Anderson, R.C. 544
Anderson, S.R. 319320, 326, 328332,
334335, 463
Anderson, T. 663
Andrade, M.J. 329, 332333
Anton, C. 108, 117
Anward, J. 496
Apel, K.-O. 147, 154, 156, 290, 307
Apperly, I.A. 571
Archer, D. 10, 87, 97, 100
Ariel, M. 39, 41, 43, 353, 433, 445, 447
Armstrong, L.E. 491492
Arnold, G.F. 491492, 664
Arnovick, L.K. 95, 124
Athitsos, V. 661
Atkinson, J.M. 267, 663
Atlas, J.D. 69, 400, 402, 405, 416, 562
Attardo, S. 61
Ashby, W.J. 431
Asher, N. 358, 363
Aston, G. 592
Auer, P. 26, 116, 156, 271, 273274,
429430, 435, 438, 446, 496497,
499500, 504, 665
Austen, J. 512, 518, 521522, 524, 531
Austin, J.L. 23, 5, 7, 11, 24, 31, 3940,
128, 143149, 153, 156159, 177, 190,
213, 220221, 223224, 272, 294296,
298, 373384, 390392, 461, 539, 561,
687
Averintseva-Klisch, M. 354
Aya, R. 266, 272
Ayers, G.M. 504
B
Bach, K. 28, 31, 4243, 69, 209210, 221,
386, 411413, 464
Baker, P. 549553, 609
Bakhtin, M.M. 16, 113, 517, 531532
Bal, M. 514, 521
Bald, W.D. 593
Baldry, A. 650651
Ball, M.J. 670
Ballmer, T. 120, 130
Bally, C. 113, 118119, 141, 158, 240
Bamberg, M. 544
Bamford, J. 130
Bang, J.C. 81
Banjo, A. 592
Bar-Hillel, Y. 13, 25, 141142, 148,
167
Barclay, R.R. 349
Barden, B. 665
Bardovi-Harlig, K. 9091, 552
Barker, C. 465, 552
Barlow, M. 429
692
Name index
Barr, D.J. 571574

Barth-Weingarten, D. 450
Barwise, J. 213
Baszanger, I. 545546
Bates, E. 81, 429
Bateson, G. 36, 53, 130, 156, 271
Bauer, L. 593
Bauman, R. 56, 69, 144, 157
Bavelas, J.B. 661, 666667
Bazzanella, C. 52
Beaugrande, R. de 35, 364, 588
Beaver, D. 59
Beavin, J.H. 176
Beck, G. 247
Beckman, M.E. 504
Bednarek, M. 16, 26, 538539, 544,
548551, 608, 632633
Beebe, L.M. 89
Behaghel, O. 444
Behne, T. 579
Benhabib, S. 306
Bennett, J. 11, 217, 220
Benseler, F. 131
Benveniste, . 141142, 147, 158, 245,
319320, 334
Benwell, B. 540
Benz, A. 412
Berens, F.-J. 274
Berenz, N. 665
Berger, P. 149, 492
Bergman, M. 182, 188, 194
Bergmann, J.R. 265, 268274, 276277,
665
Bergsland, K. 333
Berkum, J.J.A. van 580
Bernicot, J. 125
Bertolet, R. 63
Beule, J. de 450
Bezuidenhout, A. 209, 212
Biber, D. 56, 9596, 443, 588589,
592593, 609, 640641, 651
Bickel, B. 334
Biondi, M. 130
Bird, C.M. 672
Birdwhistell, R.L. 650
Birner, B. 27, 39, 434
Bishop, D. 576577
Bjrge, A.-K. 598, 606
Blackburn, P. 34
Blakemore, D. 413, 472, 475
Blommaert, J. 153
Bloom, L. 669
Bloom, P. 212
Bloomfield, L. 317
Blum-Kulka, S. 89, 542
Blutner, R. 412
Boas, F. 142, 144, 150, 326, 328329,
332333, 545
Bod, R. 429
Bodman, J. 9091
Bolinger, D. 147, 153, 491, 502
Bonvillain, N. 333
Borg, E. 205, 208211, 466, 574
Bornkessel-Schlesewsky, I. 450
Bosch, P. 364
Bott, L. 564, 577, 580
Bou-Franch, P. 87, 9293
Bouquet, P. 31, 34, 36, 43
Bourdieu, P. 149
Brandom, R.B. 27, 36, 204, 223
Bransford, J.D. 349
Brazil, D. 276, 591, 607
Breheny, R. 6, 16, 26, 544, 564565,
567570, 574, 579
Brennan, S.E. 446, 572573
Brennenstuhl, W. 120, 130
Brinton, L.J. 95, 599, 610
Brock, J. 190, 192, 195
Broek, P. van den 364
Brodda, B. 592
Brown, P. 25, 2728, 31, 40, 89, 144, 157,
277, 386, 538
Brown G. 247, 544
Brown J.D. 94
Brown-Schmidt, S. 571, 573
Brugmann, K. 158, 234, 328, 334
Brumfit, C. 85
Bruner, J. 448
Buba, M. 325, 330, 334
Bublitz, W. 23, 25, 3334, 45, 5253,
109, 111, 147, 275277, 443, 539,
547, 591, 593, 598, 604605,
631633
Bucholtz, M. 660, 671672
Buchstaller, I. 636637
Bhler, C. 247
Name index
Bhler, K. 2, 7, 1112, 68, 141142, 146,
149, 156, 158159, 170, 229230,
232239, 241249, 277, 298, 309, 315,
318, 328331, 334, 428, 440, 442, 525,
531
Bring, D. 433434, 464
Bull, P.E. 668
Bultinck, B. 411
Burenhult, N. 334
Burkhardt, A. 194
Burks, A.W. 244, 248
Burnard, L. 592
Burton-Roberts, N. 411
Busse, W. 247
Butler, J. 146, 549
Butterfill, S.A. 571
Buttrick, S. 322, 334, 348
Buyssens, E. 175
Bybee, J. 429, 445, 447
Byron, D.K. 364
C
Caffi, C. 107, 120, 127129, 131, 155156
Calvo Perez, J. 324, 333
Cameron, D. 110, 146
Cameron, L.J. 672
Cameron-Faulkner, T. 448
Camhy, D.G. 247
Candlin, C.N. 85, 146
Cap, P. 9, 61
Caple, H. 550
Cappelen, H. 15, 412, 466
Caramazza, A. 349
Carnap, R. 1, 7, 11, 142, 156, 159, 167,
206207, 561
Carston, R. 6, 15, 27, 69, 211, 214,
410412, 464, 467, 470, 473474,
476478, 574
Carter, D.M. 364
Carter, R. 594, 610, 634, 650
Cassirer, E.A. 149
Caton, S.C. 142, 157
Cermk, F. 537
Chafe, W.L. 364, 431433, 553, 587588,
592, 596
Channell, J. 631
Chapman, S(iobhan) 142, 154, 158
Chapman, S(teve) 537, 545
693
Chemla, E. 580
Chen, L. 594
Chien, Y.-C. 577
Chierchia, G. 400, 411, 565
Chilton, P. 67
Chomsky, N. 81, 188, 292293, 387, 426,
448, 466467, 484, 596
Chovil, N. 666667
Cicourel, A. 319
Clark C. 118
Clark E.V. 212, 448449
Clark, H.H. 16, 34, 36, 322, 334, 348349,
359, 364, 382, 432, 446, 561, 567, 572,
660
Clarke D.D. 140141, 147148, 150151,
158159, 236, 246249
Claudi, U. 444
Clayman, S. 272
Clift, R. 496
Coates, J. 592
Coffa, J.A. 148
Coffin, C. 549
Cohen, J. 672
Cohen, J.F. 672
Cohen, J.L. 213
Cohen, P.R. 28, 32, 36, 43, 387388
Colapietro, V. 177, 183184, 194
Collins, P. 588
Comrie, B. 147
Comte, A. 289
Connelly, G. 668
Conroy, A. 577578
Consten, M. 13, 354, 358359, 361364,
637
Conte, M.E. 247
Cook, G. 598, 630, 660
Cook, H.M. 94
Cooke, M. 12, 31, 246, 298299, 302, 304,
306309
Cooper, R.M. 565
Cooren, F. 192195
Cornish, F. 359, 364
Corriveau, J.-P. 463
Coulmas, F. 277, 644
Coulthard, M. 273276, 388, 598, 606608
Couper-Kuhlen, E. 15, 278, 430, 435436,
450, 491, 494, 496500, 503504,
597598, 602, 665
694
Name index
Coupland, N. 110, 113, 146, 159, 247

Courts, J. 192
Cousins, W.E. 334
Crain, S. 400, 575
Cresti, E. 593
Croft, W. 211, 214, 425, 428429, 603
Crowdy, S. 592
Cruse, A.D. 211, 214
Cruttenden, A. 603
Crystal, D. 450, 491492, 591
Csibra, G. 579
Culpeper, J. 10, 87, 9697, 100
Cummings, L. 23, 36, 44, 52, 416
Curl, T.S. 494, 499
Cutler, A. 499
D
Dahl, M. 551
Dahl, . 357
Damasio, H. 349
Damico, J.S. 662, 669670
Danes, F. 248
Danon-Boileau, L. 334
Dariusz, G. 159
Dascal, M. 7778, 8182
Dausendschn-Gay, U. 272
Davidson, B.D. 591
Davidson, D. 192, 309, 461
Davies, M. 595, 631
Davis, P.W. 333
Davis, S. 142, 381
Davis, W.A. 69, 223
Deacon, T. 178
Delaney, S.M. 125
Deledalle, G. 194
Dell, H. 12, 245, 275, 546
Denny, J.P. 326, 328329, 332333
Deppermann, A. 14, 278, 437, 450
Derrida, J. 177
Derry, S.J. 350
Deutschmann, M. 609
Dez, J. 334
Diessel, H. 324, 327332, 334335
Diewald, G. 438, 445
Dijk, T.A. van 2526, 40, 65, 67, 6970,
275, 364
Dik, S.C. 427
Dilthey, W. 289
Dilts, P. 651
Dingwall, R. 118
Dixon, C. 672
Dixon, R.M.W. 329, 334
Dodier, N. 545546
Drnyei, Z. 537, 542544, 546, 551, 553
Dr, J. 81
Domke, C. 12, 269, 430, 547
Donnellan, K.S. 142, 348349
Downing, A. 593
Dowty, D.R. 348
Dressler R.A. 662, 666
Dressler, W.U. 35, 364
Drew, P. 269270, 272
Driel, H. 194
Du Bois, J.W. 144, 159, 431, 504, 661, 664
Ducrot, O. 156
Dunn, J.A. 333
Duranti, A. 26, 34, 52, 143144, 156, 159,
244
E
Easton, K.L. 670
Eco, U. 334
Edmonds, P. 464
Edwards, D. 435
Edwards, J.A. 597, 608, 659661, 669
Eelen, G. 89
Eggins, S. 428, 608
Ehlers, K.-H. 248
Ehlich, K. 31, 247, 275, 278, 355, 428, 659,
665, 668, 670
Eisele, C. 183
Eisenstein, M. 9091
Ekman, P. 650, 667
Elffers, E. 246247
Emirbayer, M. 263
Enfield, N.J. 319, 323, 334335, 503
Enkvist, N.E. 591
Erk, F. 359
Erman, B. 591, 601
Eschbach, A. 247
Evans, G. 320
Evans, N. 334
F
Fabbri, P. 187, 192, 195
Faber, D. 592
Name index
Faerch, C. 93, 591
Fairclough, N. 9, 26, 67, 110, 275276, 552
Fang, A.C. 592
Fanshel, D. 275276
Farncescotti, R.M. 413
Farroni, T. 579
Fauconnier, G. 15, 352, 398, 479483
Federmeier, K.D. 450
Feilke, H. 437, 446
Feng, G. 412, 414
Fetzer, A. 8, 2527, 3435, 40, 45, 5253,
56, 68, 99
Feyerabend, P. 224
Fiehler, R. 278
Fill, A. 593
Fillmore, C.J. 57, 68, 141, 316, 324, 326,
328329, 334335, 428, 450, 480, 596
Finegan, E. 9596, 592, 651
Firth, J.R. 2, 7, 144, 491, 606
Fisch, M.H. 181, 194
Fischer, K. 438, 599, 604, 610
Fitzmaurice, S.M. 98
Flader, D. 276
Fletcher, J. 602
Foolen, A. 595
Ford, C.E. 438, 440, 447, 450, 495496,
598
Fortescue, M. 328329, 333
Foucault, M. 151, 275, 308
Fox, B.A. 437438, 440, 450, 495496
Fox Tree, J.E. 660
Francis, G. 364
Francis, W.N. 591, 596
Franozo, E. 7778, 8182
Franks, J.J. 349
Franquiz, M. 672
Fraser, B. 277, 364
Fraurud, K. 357, 359
Freese, J. 499
Frege, G. 11, 68, 143, 148149, 206, 220,
309, 347, 379, 401402, 412, 461
French, P. 495
Freitas, S. 364
Fretheim, T. 364, 434, 599
Freud, S. 82
Friederici, A.D. 450
Friedman, M. 148, 159
Friedrich, P. 334
695
Fries, N. 437
Fries, U. 95
Friesen, W.V. 667
Fritz, G. 95, 275
Fronek, J. 248
Furiasso, C. 595
G
Gabelentz, G. von der 229, 240
Galasinski, D. 541, 552
Gale, R.M. 320
Garcs-Conejos, P. 87, 9293
Gardiner, A.H. 2, 7, 232, 234, 246247
Gardner, R. 438, 606
Garfinkel, H. 2, 7, 12, 26, 34, 141, 145,
149, 261269, 272274, 319, 323, 334
Garnham, A. 364, 591592
Garret, M. 411
Garrod, S.C. 359, 364, 446, 572
Garvey, C. 329, 334
Garvin, P.L. 237, 247
Gazdar, G. 68, 143, 406407, 411, 562,
566
Gee, J.P. 120
Geeraerts, D. 123
Geissner, H. 130
Geluykens, R. 592
Gergely, G. 579
Gernsbacher, M.A. 350, 355, 364
Geurts, B. 400, 411, 563, 567
Giannakidou, A. 400
Gibbon, D. 668
Gibbs, R.W. 16, 125
Giddens, A. 149
Gilliom, L.A. 364
Givn, T. 25, 28, 32, 38, 352353, 355,
364, 446
Glassner, B. 540, 542
Gloning, T. 275
Glucksberg, S. 16, 580
Goffman, E. 2, 7, 1213, 28, 34, 36, 53,
122, 144145, 153, 157, 269271, 273,
277, 390391, 499
Golato, A. 552
Goldberg, A. 428, 450
Goldberg, J. 498
Goldin-Meadow, S. 667
Goodman, N. 213214
696
Name index
Goodwin, C. 6, 34, 52, 143, 156, 268, 272,

319321, 323, 335, 391392, 438, 442,
499500, 502503
Goodwin, M.H. 438, 499500
Goossens, L. 121
Gordon, P.C. 364
Gotti, M. 124125
Goudge, T.A. 186
Graczyk, R. 326, 329, 333
Grady, J. 122, 124
Graesser, A.C. 561
Graf, C. 112
Graffi, G. 246, 248
Granger, S. 592593
Graumann, C.F. 247
Greatbatch, D. 118
Green, G.M. 69, 274
Green, J. 672
Greenbaum, S. 592, 596
Greenberg, R. 670
Grefenstette, G. 595
Greimas, A.J. 190, 192195
Grice, H.P. 2, 57, 11, 24, 29, 31, 3637,
4143, 61, 69, 89, 109, 114, 129130,
142, 145, 147, 153154, 157159, 170,
175176, 178180, 187188, 208,
217219, 223, 273, 319, 381383, 387,
407410, 412414, 416, 440, 448, 461,
467469, 471472, 474, 539, 561562,
566567, 578, 580
Gries, S.T. 449
Grimm-Vogel, I.H. 246
Grodner, D.J. 411, 565566, 569570
Grodzinsky, Y. 577
Grnqvist, L. 670
Grondelaers, S. 123
Grosz, B. J. 36, 364
Gruber, H. 547
Grnbaum, A.A. 236237, 239, 247
Grundy, P. 6869, 274
Guasti, M.T. 576
Guendouzi, J.A. 670
Gnthner, S. 278, 439, 450, 665
Gundel, J.K. 359, 364, 433434
Gunnarsson, M. 670
Gumperz, J.J. 2627, 34, 56, 117118, 120,
122, 131, 144, 156, 274275, 319, 334,
462, 492, 546, 665
Gustafsson, M. 591
Gut, U. 538, 551, 668
Guy, G. 602
H
Haan, F. de 593,
Haastrup, K. 542
Habel, C. 364
Haberland, H. 158
Habermas, J. 2, 7, 12, 31, 111, 116, 148,
171, 178, 246, 289310
Haegeman, L. 591
Hagemann, J. 107, 130
Hagen, K. 598
Hager, J.C. 667
Haiman, J. 444
Hajicov, E. 248
Halliday, M.A.K. 12, 25, 38, 52, 55, 114,
117, 130, 141, 147, 156, 190, 229, 232,
246, 317, 357, 427, 433434, 440,
491492, 538, 596, 603, 607, 609
Hancher, M. 155
Hanks, W.F. 6, 13, 26, 28, 36, 58, 68, 143,
316, 320, 323, 325327, 330, 333335,
463, 651
Hanna, J. 571
Haravon, A. 670
Harlow, L.L. 9293
Harnish, R.M. 221, 386, 411
Harris, R.A. 143, 147
Hart, H.L.A. 222
Hartford, B.S. 91, 552
Hasan, R. 52, 55, 114, 117, 130, 147, 156,
317, 357
Haslerud, V. 593
Haspelmath, M. 444
Hasselgrd, H. 364, 594
Hausendorf, H. 269, 275276
Have, P. ten 70, 268
Haverkate, H. 79
Haviland, J.B. 334335
Haviland, S.E. 364
Hawkins, J. 359, 364
Hay, J. 429
Hayashi, M. 436
Heath, C. 268, 272, 548
Heath, J. 325, 328, 330, 332334
Hedberg, N. 433
Name index
Hedstrm, K.E. 591
Heger, K. 247
Heim, I. 359, 363364, 407
Heine, B. 444, 590
Heller, D. 571
Hellermann, J. 496
Hellmann, C. 357
Hemingway, E. 518519
Hengeveld, K. 427
Henrichsen, P.J. 670
Henzler, S. 247
Hepburn, A. 504
Heritage, J. 26, 34, 145, 261, 264,
267269, 272, 278, 335, 526, 663
Herrmann, T. 247
Hervey, S.G.J. 176
Herzog, M.I. 240
Heusinger, K. von 353, 414
Heuvel, T. van den 592
Hill, C. 325, 334
Hilpinen, R. 179, 188, 190, 194195
Himmelmann, N.P. 334335
Hindelang, G. 275
Hintzman, D.L. 355
Hirschberg, J. 410411, 597
Hobbs, J.R. 364
Hrmann, H. 247
Hoey, M. 549, 632
Hoffmann, L. 428
Hoffmann, S. 630
Hofland, K. 591592, 595
Hoiting, N. 596
Holdcroft, D. 63
Holenstein, E. 149, 151, 248249
Holly, W. 12, 277, 430, 547
Holmes, J. 87, 551552, 592593
Holt, E. 496
Honneth, A. 303, 307308
Hood, S. 550
Hopper, P.J. 429430, 439, 444445, 447
Horn, L.R. 6, 2728, 3839, 69, 400, 405,
407408, 410414, 416, 464, 470471,
473, 562, 566
House, J. 93
Houser, N. 183, 194
Huang, Y. 3, 14, 2728, 32, 39, 43, 60,
6869, 140, 381, 397398, 400, 402,
405413, 470474, 580
697
Huang, Y.T. 569570

Hudson, R. 590, 637
Hbler, A. 10, 109, 111, 124, 147, 547
Hlzer, H. 246
Hnnemeyer, F. 444
Hundsnurscher, F. 275
Hundt, M. 695
Hunston, S. 549, 608, 633
Hurewitz, F. 411
Husserl, E. 148149, 158159, 233, 236,
247, 261, 291, 319, 394
Hutchby, I. 267268
Hutchins, E. 393
Hyland, K. 115
Hymes, D.H. 2, 7, 1112, 56, 144,
244245, 274275, 336, 546
I
Ide, S. 31
Imo, W. 278, 446, 450
Innis, R.E. 247
Irvine, J.T. 144
Israel, M. 398
J
Jackendoff, R. 350, 480
Jackson, D.D. 176
Jacobs, A. 9596, 100
Jaffe, A. 672
Jaggar, P. 325, 330, 334
Jakobson, R. 2, 7, 1113, 129130, 134,
141142, 146147, 149152, 156, 158,
170171, 229230, 232, 240, 242246,
248249, 309, 318, 334
Jannedy, S. 429
Janssen, T. 334
Jasperson, R. 499
Jaszczolt, K. 2729, 45
Jaworski, A. 110, 159
Jefferson, G. 70, 145, 265267, 274, 335,
379, 388, 391, 435436, 492, 498, 504,
531, 663
Jeffrey, R.C. 217
Jespersen, O. 13, 130, 141, 240, 244, 318,
334
Jewitt, C. 547
Johannessen, J.B. 598
Johansen, J.D. 181, 188, 195
698
Name index
Johansson, S. 443, 591592, 594, 596

Johnson, K. 85
Johnson, M.H. 121122, 178, 214
Johnson-Laird, P.N. 364, 366
Johnston, T. 549
Jonge, C.C. de 407
Jrgensen, A.M. 593
Joshi, A.K. 364
Joswick, H. 195
Juchem, J.G. 232, 246247
Jucker, A.H. 7879, 9597, 100, 599, 630,
646648, 651
K
Kallmeyer, W. 274
Kamide, Y. 568, 574
Kamio, A. 525
Kamp, H. 363
Kamp, R. 247
Kaplan, D. 11, 142, 207, 209, 223, 462, 466
Kappner, S. 176
Karkkinen, E. 446
Karttunen, L. 68, 143, 402, 406407
Kasper, G. 8890, 9293, 543, 551, 591
Katsos, N. 576577
Kay, P. 429, 450
Keenan, E.L. 89, 319320, 326, 328332,
334335, 463
Kehoe, A. 595
Keller, R. 214, 446
Kelly, J. 493, 495, 498
Kemmer, S. 429
Kendon, A. 321322, 598
Kennedy, C. 465
Kennedy, G. 587588
Kenwood, C. 661, 667
Keysar, B. 571
Kibble, R. 364
Kilgarriff, A. 595, 638639
King, J.C. 412, 462
Kingdon, R. 664
Kintsch, W. 70, 364
Kipp, M. 661
Kirk, J. 608
Kirschbaum, I. 214
Kissmann, U.T. 272
Kita, S. 335
Kitzinger, C. 504
Kjellmer, G. 606
Klein, W. 248
Klewitz, G. 496
Knees, M. 358
Knoblauch, H. 272
Knobloch, C. 246247
Knowles, G. 592593, 608
Koerner, E.F.K. 247
Kohnen, T. 97, 651
Kohrt, M. 275
Koktov, E. 533
Koons, R. 209
Kowal, S. 659, 662, 670671
Koyama, W. 10, 23, 143, 223
Krafft, U. 272
Kress, G. 442443
Kreuz, R.J. 18, 537, 547, 590, 662, 666,
672
Krifka, M. 434
Kripke, S. 142
Kronmuller, E. 573
Kroskrity, P.V. 159
Krzyzanowski, M. 551
Kubo, S. 3133
Khn, P. 277
Kuklick, H. 145
Kulick, D. 146
Kuno, S. 334
Kumpf, L.E. 431
Kurylowicz, J. 328, 335
Kussmaul, P. 591
Kutas, M. 450
Kutschera, F. von 217, 220
Kyburg, A. 465
Kyratzis, A. 658
Kyt, M. 97, 592
L
Labov, W. 240, 275276, 444, 544, 546
Ladd, D.R. 499
Ladusaw, W.A. 399, 658
Lakoff, G. 15, 121122, 143, 178, 214,
428, 478, 480
Lakoff, R.T. 38, 146
Lambrecht, K. 39, 364, 369, 431
Lampert, M.D. 597, 608
Langacker, R.W. 15, 53, 428429, 441,
447, 478479, 483
Name index
Langford, D. 608
Lasater, A. 324
Lasersohn, P.N. 465
Lasnik, H. 426
Lass, R. 444
Lauridsen, K.M. 592
Laval, V. 125
Lawrence, L. 592
Lazarus, M. 246
Laziczius, J. von 247
Leder, D. 108
Lee, D. 550
Leech, G.N. 9, 38, 7778, 80, 8286, 89,
92, 99100, 109, 114, 147, 158, 170171,
186188, 443, 538, 588, 597, 642, 661
Leeuwen, T. van 442443
Lehrer, A. 410
Leicher, S. 630
Lenk, U. 364, 443, 599, 651
Lenz, F. 358
Leonard, R. 334
LePage, R.B. 444
Lepore, E. 15, 412, 461, 466
Lerner, G. 436
Levelt, W.J.M. 668
Levenston, E. 542
Levinson, S.C. 6, 10, 23, 2729, 3132, 36,
38, 4043, 5657, 61, 6869, 79, 83, 89,
141, 143145, 153, 155159, 167170,
175, 186187, 194, 223, 242, 244, 247,
273277, 319, 334335, 383, 386,
388390, 392, 398, 400, 402403, 408,
410413, 416, 465, 469473, 475, 484,
538, 562567, 642
Lewis, D.K. 11, 23, 33, 36, 204, 209, 217,
220, 223, 481, 561
Liberman, A. 242
Liebal, K. 579
Liedtke, F. 436
Lieven, E. 448
Lindquist, H. 590
Lindsay, J. 671
Linell, P. 56, 493
Lipton, P. 223
Liszka, J.J. 194
Lizskowski, U. 578
Local, J.K. 493, 495498, 501, 504
Locke, J. 140, 149150, 154, 159
699
Loehr, D.P. 598

Lof, G.L. 658
Lohmann, J. 247
Loll, A. 354, 364
Loman, B. 591
Looks, K. 668
Lopes, J.G. 364
Louw, B. 539, 549, 608, 632633, 635
Louwerse, M.M. 667, 672
Luckmann, T. 56, 69, 149, 492
Lucy, J.A. 111, 116, 129130, 154, 156
Luff, P. 268, 272
Luhmann, N. 178, 269, 308
Luzio, A. di 26, 156
Lycan, W.G. 638
Lyons, J. 141, 242, 244, 320, 335, 358, 464,
642
M
MacDonald, M.C. 566
Macfarlane, S. 495
Mackenzie, J.L. 427
MacWhinney, B. 349, 429, 660, 665666,
669670
Maes, A. 364
Mahlberg, M. 551
Mahon, B. 349
Mair, C. 590
Makin, V.S. 572
Malinowski, B. 2, 7, 144145, 150, 232,
243, 246, 248, 545
Manelis Klein, H.E. 333
Manes, J. 646647
Mann, W. 55
Manning, P. 57
Manor, R. 205
Markman, A.B. 572
Marmaridou, S.S.A. 9, 25, 38, 45, 54, 57,
69, 79
Marshall, C. 334, 348
Marslen-Wilson, W. 349
Martens, E. 195
Martin, J.R. 548
Martinet, A. 149, 229, 319
Marty, A. 158159, 233, 247
Masaryk, T.G. 240
Mathesius, V. 229230, 239242, 245, 248
Matisoff, J. 329, 334
700
Name index
Matthiessen, C.M.I.M. 550, 552

Maturana, H.R. 127
Mauranen, A. 589
Mauro, T. de 594
Mautner, G. 547, 551552
Maynard, C. 630
Maynard, D.W. 263, 499
Mazeland, H. 496
McCarthy, M.J. 594, 606, 610, 634, 639
McCarthy, T. 289
McComish, J.F. 670
McDonald, J.L. 349
McEnery, T. 18, 587589, 594
McIntyre, D. 642645
McKee, C. 575
McKoon, G. 361
McNamara, T. 10, 9294
McNeill, D. 115, 127, 598, 668
McNeill, P. 537, 540, 545, 551
Meggle, G. 11, 217, 220, 223, 539
Meibauer, J. 274, 450
Meier, C. 665
Meijs, W. 592
Meillet, A. 158, 319
Merin, A. 412
Merlan, F. 334
Merleau-Ponty, M. 148, 152, 319
Metzing, C. 572573
Mey, J.L. 9, 15, 2425, 54, 6869, 7778,
80, 128129, 140141, 143144,
146149, 153, 155156, 158159, 168,
274, 526, 528, 532, 629
Meyer-Hermann, R. 112, 130
Mill, J.S. 407
Miller, J. 540, 542
Milroy, L. 596
Ming, T. 594
Mishler, E.G. 672
Mitterer, H. 503
Moeschler, J. 275
Mondana, L. 334
Moneglia, M. 593
Montague, R. 141142, 167, 207, 484
Monville-Burston, M. 248249
Moore, G.E. 158159, 379, 416, 528
Moravcsik, J.M. 212213
Morel, M.-A. 334
Morgan, J. 28, 32, 36, 43
Morley, J. 631633
Morpurgo Davies, A. 246
Morreau, M. 465
Morris, C.W. 1, 2, 7, 11, 13, 24, 79, 142,
148, 150, 158, 167169, 178, 183, 186,
190, 203204, 248, 316, 334
Mosegaard Hansen, M.-B. 587, 597,
599600, 610
Moxey, L.M. 364
Muck, B. 124
Mller, F. 496, 499500, 504
Mller, N. 662, 670
Mulder, J. 441
Mulligan, K. 150, 159
Murphy, G.L. 364, 477478
Murphy, J.P. 183
Musolff, A. 11, 247, 309, 428
Musolino, J. 575576
Myers, G. 661
N
Nadin, M. 184, 194
Nsslin, S. 591
Neidle, C. 661
Nekula, M. 248
Nerlich, B. 140141, 147148, 150152,
158159, 236, 246249
Nevala, M. 98
Nevalainen, T. 96, 592593
Newman, J. 651
Newmeyer, F.J. 426
Niedzielski, N.A. 120
Nth, W. 11, 168170, 172, 174, 176,
182185, 189190, 194195
Norrick, N. 604, 610, 651
Norris, S. 442, 660
Noveck, I.A. 411, 544, 563564, 575577,
580
Novick, D.G. 657
Nunberg, G. 68, 316
O
OConnell, D.C. 659, 662, 670671
OConnor, J.D. 491492, 664
OConnor, M.C. 450
ODonnell, M. 548
OHalloran, K. 549, 633
OSullivan, M. 650
Name index
Oakhill, J. 364
Obler, L.K. 670
Ochs, E. 278, 335, 450, 600, 660, 663, 667,
669
Oehler, K. 179, 194195
stman, J.-O. 591, 600
Ogden, R. 494495, 497, 503
Olsen, S.E. 80, 100
Olshtain, E. 89
Ono, R. 542543, 552
Ono, T. 450, 504
Orestrm, B. 591, 604
Ortner, H. 247248
P
Palen, L. 393
Paltridge, B. 553
Panizza, D. 411
Pankow, C. 194
Papafragou, A. 575576
Pape, H. 186, 188, 195
Paradis, C. 591
Parks, E. 516517
Parmentier, R.J. 177
Parrot, D.J. 672
Parret, H. 176, 184, 194
Parsons, T. 261, 265
Partee, B. 205, 207
Partington, A. 589, 631633
Paul, I. 108
Peirce, C.S. 1, 7, 11, 13, 23, 141142,
148150, 158159, 168169, 171174,
176177, 179186, 188195, 244, 263,
316317, 334, 382
Pekarek Doehler, S. 439
Penco, C. 36
Pepp, S. 495
Perrault, C.R. 387388
Perrott, M.A. 672
Perry, J. 213, 462
Peter, G. 466
Petrus, K. 566, 580
Pfnder, S. 429430
Phillips, B. 661, 667
Pickering, M.J. 446
Piercy, M. 522523
Pietarinen, A.-V. 179180, 182, 188, 190,
194
701
Pietroski, P.M. 400, 467

Pike, K.L. 144, 491
Pinker, S. 448
Plato 172, 233
Pollak, A. 547
Pollack, M.E. 28, 32, 36, 43
Pomerantz, A. 267
Portner, P.H. 205, 407
Posner, R. 167, 194
Potter, J. 70, 504
Potts, C. 413414
Pouscoulous, N. 563, 567, 576
Power, R. 364
Powers, E. 529
Powers, W.R. 666
Pressman, J.F. 142
Preston, D.R. 120, 672
Preyer, G. 466
Prieto, L.J. 176, 194
Prince, E.F. 364, 433, 439
Prucha, J. 9, 77, 80
Psathas, G. 267, 274, 663
Pschel, U. 277
Pullum, G.K. 658
Putnam, H. 142
Pye, C.A. 672
Q
Quasthoff, U. 665
Quinn, N. 131
Quirk, R. 114, 130, 591, 596, 664
R
Rahilly, J. 670
Ransdell, J. 182
Ratcliff, R. 361
Rauch, E. 277
Raumolin-Brunberg, H. 96
Rawls, J. 124
Raymond, G. 441
Raymond, W.D. 597
Reber, E. 499
Reboul, A. 411
Recanati, F. 27, 29, 3132, 155, 209, 214,
364, 411412, 464, 466, 484
Redder, A. 276, 428
Reddy, M. 29, 121, 124
Rees, M.A. van 390
702
Name index
Rehbein, J. 275, 278

Reichenbach, H. 320, 335
Reinhart, T. 364, 577578
Rellstab, D. 194
Renouf, A. 592, 595
Reppen, R. 589, 609
Richardson, A. 148
Rickheit, G. 364
Ricur, P. 157
Rieber, S. 413
Riley, P. 8486
Rim, B. 650
Rissanen, M. 592, 596
Risselada, R. 364
Roberts, C. 672
Roberts, F. 672
Robinson, J.D. 672
Robinson, M.A. 90
Roever, C. 10, 87, 9294
Romero, J. 603604
Rommetveit, R. 319
Rorty, R.M. 147
Rosaldo, M.Z. 89, 144, 159
Rose, K.R. 88, 90, 9293, 542543, 552
Routarinne, S. 495
Rubio-Fernandez, P. 580
Rudy, S. 243, 248
Rhlemann, C. 17, 26, 587, 590, 599, 634,
651
Ruesch, J. 156
Ruiter, J.P. de 503
Russell, B(enjamin) 400
Russell, B(ertrand) 143, 156, 309, 320, 335,
348, 405406, 461, 638, 640
S
Saba, W.S. 463
Sacks, H. 2, 7, 12, 70, 143, 265269,
272274, 316, 319, 335, 379, 388, 391,
433, 435, 492, 503, 531, 663
Sadock, J.M. 61, 69, 143, 159, 464
Saeed, J.I. 15, 29, 68, 223, 400, 416, 430
Sajavaara, K. 591
Sanders, T. 351, 364
Sandford, D. 588, 609
Sanford, A.J. 16, 359, 364
Sankoff, D. 601
Santaella, L. 169, 189, 194195
Sapir, E. 142, 151, 545

Sarangi, S. 146
Sarno, M.T. 670
Sauerland, U. 400, 411, 566
Saul, J. 416, 566, 580
Saunders, R. 333
Saussure, F. de 5, 126, 152, 154, 168, 174,
234236, 308
Savan, D. 185
Savigny, E. von 221222
Saville-Troike, M. 56
Sayette, M. A. 672
Sbis, M. 28, 3132, 187, 192193, 195,
373, 382384, 387
Schallenberger, S. 247
Schank, G. 277
Schank, R.C. 480
Scheerer, T.M. 247
Schegloff, E.A. 12, 70, 145, 261, 265269,
273274, 278, 319, 323, 330, 334335,
379, 388, 390391, 433, 435436, 440,
450, 492493, 496, 498, 500, 503504,
531, 547, 663, 668
Schelfhout, C. 593
Schenkein, J.N. 267268, 663
Schiaratura, L. 650
Schieffelin, B.B. 159
Schiffer, S. 217
Schiffrin, D. 244, 438, 658
Schlieben-Lange, B. 247
Schlobinski, P. 443, 665
Schmid, H.-J. 35, 130, 357, 439, 609
Schmitt, R. 272
Schmitz, U. 443
Schneider, S. 593
Schneider, W.L. 269
Schnurr, S. 551552
Scholes, R.E. 513, 519, 532
Scholz, O.R. 223
Schreuder, R. 322, 334, 348
Schubiger, M. 492
Schtz, A. 261, 264, 273
Schtze, F. 274
Schunn, C.D. 350
Schutz, A. 148, 319321, 335
Schwarz, F. 580
Schwarz, M. 350352, 354, 357359,
362364
Name index
Schwarz-Friesel, M. 13, 354, 358, 363, 637
Schwitalla, J. 274, 277278
Sclaroff, S. 661
Searle, J.R. 2, 5, 7, 11, 24, 2728, 3133,
3537, 3941, 69, 93, 107, 127128,
143145, 147148, 154, 157, 171,
175176, 180, 184192, 212, 220, 223,
236, 272, 294296, 316, 334, 349, 373,
381388, 390, 392, 432, 436, 461, 466,
539, 641, 646, 648, 658
Sedivy, J.C. 565567, 569570, 577
Sefi, S. 526,
Seidenberg, M.S. 566
Selinker, L. 88
Selting, M. 278, 435, 450, 495, 497, 504,
598, 602, 665
Semino, E. 121122, 548
Sengul, C.J. 349
Senft, G. 335
Shannon, C.E. 170, 242
Shastri, S.V. 592
Sherzer, J. 144, 334
Short, M. 548
Short, M.H. 642
Short, T.L. 179
Shriberg, L.D. 658
Sidnell, J. 334335
Sidner, C.L. 36
Sifianou, M. 87
Silverman, D. 537
Silverstein, M. 116, 142, 147, 153154,
156157, 159, 318, 323, 333334
Singer, M. 364
Simmons-Mackie, N.N. 669
Simon-Vandenbergen, A.-M. 122123,
594, 599
Simpson, G.B. 464
Sinclair, J.M. 274275, 388, 550551,
558589, 592, 598, 606609, 631,
639
Siren, K.A. 672
Skaffari, J. 549
Slobin, D.I. 669
Smith, B.C. 461
Smith, B.L. 123
Smith, H.L.J. 491
Smith, S. 599
Smits, R. 335
703
Snedeker, J. 569570
Snow, C. 669
Soames, S. 69
Sontag, S. 523524, 533
Sorjonen, M.-L. 438
Southgate, V. 579
Sperber, D. 6, 15, 27, 3234, 36, 4142, 68,
112, 129, 157, 168, 173, 175, 179, 351,
382, 410412, 473475, 477, 544, 564,
567, 580, 594, 599, 636
Spinks, C.W. 194
Spooren, W. 351, 364
Spranz-Fogasy, T. 278
Stalnaker, R.C. 11, 3133, 35, 6869, 223,
406407, 484
Stanley, J. 206, 210, 412, 462, 484
Steels, L. 450
Stefanowitsch, A. 449
Stenning, K.D. 349, 364
Stenstrm, A.-B. 588, 591593, 597598,
603, 607609, 644, 651
Steinbach, M. 450
Stelma, J.H. 672
Stetter, C. 174
Stiles, W.B. 630
Stivers, T. 335
Stocking, G.W. Jr. 151
Storm, T. 514
Strawson, P.F. 2, 5, 7, 11, 68, 143, 147, 153,
187, 309, 348, 382, 403, 406, 416, 461,
466, 638, 656
Streeck, J. 272, 278, 443
Stckl, H. 443
Stokoe, E. 540
Stubbe, M. 548
Stubbs, M. 549, 607, 609, 631, 633634
Sudnow, D. 262, 265, 267
Svartvik, J. 591592, 596, 664
Swanton, J. 326, 333
Sweetser, E.E. 444
Swiggers, P. 247
Szab, Z.G. 467, 484
Szczepek Reed, B. 495, 497, 500
T
Taavitsainen, I. 95, 100
Tabouret-Keller, A. 444
Tagliamonte, S. 590, 637
704
Name index
Tajfel, H. 118
Takahashi, S. 90
Takahashi, T. 89
Talmy, L. 478
Tanaka, H. 495
Tanenhaus, M.K. 565
Tannen, D. 144, 277, 544545, 664665
Tao, H. 325, 334
Tarplee, C. 494
Taylor, L. 592
Taylor, S. 537, 548, 551
Tenchini, M.P. 246
Tesnire, L. 194
Thavenius, C. 591
Thibaud, P. 195
Thibault, P. 56, 69, 549, 650651
Thies, A. 668
Thomas, J. 9, 69, 7778, 80, 82, 8587, 99,
155, 592, 661
Thomason R. H. 33
Thompson, G. 549
Thompson, S.A. 55, 278, 335, 435436,
438441, 446447, 450, 495
Thornton, R. 577
Thrane, T. 348
Tilley, S.A. 672
Titscher, S. 537, 545546, 548, 553
Tognini-Bonelli, E. 538, 588
Toman, J. 242243, 248
Tomasello, M. 212, 448450, 567,
578579
Toobin, J. 377
Tottie, G. 591, 593, 598, 601, 605
Tracy, K. 277
Trager, G.L. 491
Traugott, E.C. 39, 100, 124, 444445,
590
Travis, C. 212
Trippel, T. 668
Trnka, B. 248
Trosborg, A. 89
Trotha, T. von 276
Tsohatzidis, S. 69
Turner, K. 414
Turner, M. 479, 483
Turner, R.M. 262265, 268
Tynjanov, J. 243
Tyson, S. 593
U
Uhmann, S. 278, 434, 496, 665
Ungeheuer, G. 112, 232, 247
Urban, G. 153
V
Vachek, J. 248
Vanderveken, D. 3133, 436
Vanparys, J. 121
Venditti, J.J. 597
Ventola, E. 190, 275, 443
Verschueren, J. 25, 52, 108, 116, 120,
127128, 130, 147, 156, 158159, 588
Vicher, A. 601
Virtanen, T. 450, 587588
Volosinov, V.N. 157
Vonk, F. 247
Vonwiller, J. 602
Voormann, H. 538, 551
W
Walker, G. 496498, 504, 597
Walsh, S. 589590
Waltereit, R. 594
Walton, L. 657
Ward, G. 27, 39, 434
Ward I.C. 491492
Ward, K. 657
Wardhaugh, R. 546, 592
Warner, R. 566
Warnock, G.J. 147, 374, 381, 383
Wasow, T. 444
Watts, R. 31
Watzlawick, P. 130, 176
Waugh, L.R. 243, 248249
Weaver, W. 170, 242
Webber, B.L. 358
Weber, M. 150, 154, 223, 273, 289, 300
Wegener, P. 2, 7, 158, 230232, 234,
240243, 245247
Weil, H. 241
Weinback, L. 89
Weinreich, U. 240
Weissenborn, J. 247
Welker, K. 410
Wells, B. 495
Wells, W.H.G 495
Welte, W. 130
Name index
Werlen, I. 277
Wertz, J.M. 672
Westney, P. 592
Wetherell, M. 541
Wettersten, J. 247
Wexler, K. 577
White, P.R.R. 548
Whitsitt, S. 632
Wichmann, A. 496, 593, 597598,
602603
Wichmann, S. 333
Widdowson, H.G. 35, 85
Widmer, J. 269
Wierzbicka, A. 89, 100, 599
Wilcox, K.A. 672
Wilkes-Gibbs, D. 446
Wilkinson, S. 504
Williams, G. 146
Wilson, A. 587588, 592
Wilson, D. 6, 15, 27, 3234, 36, 4142, 68,
112, 129130, 157, 168, 173, 175, 179,
351, 382, 410412, 473475, 477479,
564, 567, 594, 599, 636
Wittgenstein, L. 2, 7, 11, 23, 100, 144, 147,
156, 172, 216, 222, 224, 248, 273,
293296, 374, 389, 461
Wodak, R. 67, 100, 275, 537, 547, 551
Wolf, K. 247
Wolfson, N. 89, 646647
Wood, J.L. 97
Woodfield, H.P. 542543, 551
Woods, D.K. 661
Woods, J. 415
Wooffitt, R. 118, 267
Woolard, K. 159
Woolf, V. 531
Wootton, A. 269270
Wright, G.H. von 154
Wright, M. 498
Wunderlich, D. 247
Wundt, W. 150, 233, 247
Wynne, M. 538, 551
X
Xiao, R. 594
Y
Yaeger-Droh, M. 500
Yang, Y. 325
Yngve, V. 604
Yuan, Y. 543, 546, 551552
Yule, G. 24, 100, 247, 364, 544
Z
Zacharsky, R. 433
Zelinsky-Wibbelt, C. 364
Zide, N. 325326, 329, 333334
Ziv, Y. 599, 651
705
706
Subject index
Subject index
A
anaphora 4, 7, 13, 51, 55, 156, 205, 238,
317318, 325, 327, 330333, 347,
353358, 363, 412, 428, 433, 440, 462,
599600, 603, 607, 640641, 644
indirect ~ 359360, 364, 640641
anaphoric see anaphora
appropriateness 910, 25, 27, 29, 77, 92,
94, 99100, 143144, 146, 268, 414
C
cataphora 111, 238, 317, 330, 640641,
644
cataphoric see cataphora
cognition 14, 185186, 291, 305, 307,
349350, 391, 393, 430, 474, 480, 566
cognitive
~ models 121, 123, 480
~ pragmatics 26, 34, 68, 566, 574
~ process 15, 43, 82, 347, 352, 355356,
358, 360361, 363, 431, 450,
479481, 483, 542, 566
~ processing see ~ process
~ science 4, 2334, 38, 44, 127, 350,
578, 588
~ semantics 15, 363, 478479
coherence 116, 351, 353, 356357, 363,
438, 651
common ground 8, 31, 3335, 38, 58, 69,
120, 178, 180, 184, 188189, 269, 320,
327, 330332, 348, 407, 415, 432, 589,
599660
compositionality 2728, 45, 206, 325326,
406
Construction Grammar 14, 278, 428, 430,
450
context
cognitive ~ 9, 25, 29, 3536, 39, 52, 54,
62, 68, 70
extralinguistic ~ see non-linguistic ~
linguistic ~ 9, 25, 2829, 3536, 56,
565566
nonlinguistic ~ 55, 59, 66, 206, 317,

324, 398, 408, 642
~ sensitive see ~ sensitivity
~ sensitivity 205, 207211, 215, 414,
449, 465467, 548, 629
social ~ 9, 2529, 32, 3536, 39, 52, 62,
70, 7980, 96, 99, 246, 278, 304, 306,
491, 552, 590, 609, 630, 660
Conversation Analysis 14, 16, 26, 65,
144145, 149, 153, 155, 265270,
272278, 319320, 390391, 430, 492,
531, 537, 547548, 587, 589, 597,
663
cooperative principle 24, 3738, 4143,
61, 83, 382, 387, 407408, 413, 468,
648
corpora see corpus
corpus 56, 1718, 95100, 359, 364, 429,
449, 538, 545552, 587599, 601603,
605610, 629631, 636640, 642, 644,
646651, 664665
~ based 6, 8, 17, 436, 538, 551,
587593, 595598, 600, 602606,
609610, 638, 648, 651
~ driven 538, 588590, 610
Corpus Analysis see Corpus Linguistics
Corpus Linguistics 8, 10, 1618, 56,
7778, 95, 538539, 547549, 552,
587590, 593, 595597, 610, 629,
630631, 633, 638, 646647, 649651
Corpus linguistic see Corpus Linguistics
Critical Discourse Analysis 9, 26, 6770,
9697, 100, 144, 146, 275, 548, 552
Critical Social Theory 12, 148, 290, 300,
303, 305, 307308
D
deictic center 142143, 642
deictic field 141142, 147, 237238, 247,
315316, 322323, 328
deictics 13, 58, 117, 315318, 320,
324333, 428, 435, 444, 642, 644645
Subject index
deixis 45, 7, 9, 13, 15, 18, 28, 31, 51,
5559, 6164, 68, 79, 95, 98, 141,
184, 186, 204205, 235, 238, 244,
246247, 316324, 326, 329, 331332,
354, 440, 462, 481, 638, 642, 644645,
651
discourse ~ 28, 36, 58, 315, 327,
358
participant ~ see person ~
person ~ 36, 315, 644
referential ~ 315, 319, 644
social ~ 36, 98, 315, 392
spatial ~ 186, 238, 315, 317, 327330
temporal ~ 36, 186, 315, 324, 327
Discourse Analysis 9, 2627, 120, 153,
155156, 274275, 373, 390, 547548,
551552, 587, 597, 607, 609
discourse marker 57, 28, 95, 119, 350,
438439, 443, 446, 448, 476, 526,
587, 590591, 594595, 598599,
601602, 604, 609610, 637, 651,
658660
E
entailment 4, 7, 14, 397401, 405, 408,
410, 413416, 470, 566
ethnography 8, 12, 16, 27, 55, 120, 144,
229, 263, 274275, 319, 322, 388,
390392, 537, 545546, 596
ethnographic(al) see ethnography
Ethnomethodology 2, 7, 12, 34, 145, 149,
261265, 269, 272, 274, 319
explicature 1415, 42, 397, 410412,
430, 475, 477
F
face 28, 31, 40, 94, 144, 270, 277, 543, 597,
603
falsity see truth
felicity 23, 357, 379
~ condition 8, 3233, 35, 4041, 62, 69,
143144, 172, 178, 383385, 390,
403, 576
frame 4, 14, 34, 36, 53, 9697, 111,
144, 153, 178, 271, 277, 322, 360,
362, 441442, 449, 480, 514, 544,
640
Frankfurt School 146, 290, 303, 307
707
function of language
conative ~ 146, 152153, 157, 171, 243,
248
emotive ~ 146, 152153, 157, 170171,
243
expressive ~ 170, 241, 243
metalinguistic ~ 146, 152153, 159,
244
phatic ~ 146, 152, 171, 244, 248
poetic ~ 146, 152153, 156, 159, 171,
243
referential ~ 146147, 152153, 155,
157158, 170, 243
I
illocution 9, 15, 32, 40, 51, 54, 57, 62,
77, 8385, 8889, 109, 148, 171, 177,
190194, 220222, 224, 296298, 349,
380381, 383387, 389390, 643, 648,
658
illocutionary act see illocution
illocutionary force see illocution
illocutionary force indicating devices
383385, 436, 646
implicature 4, 67, 9, 1416, 28, 3133,
37, 39, 4244, 51, 5564, 6769, 79,
94, 98, 139, 142, 145, 153, 157158,
179180, 188, 203205, 208, 319, 323,
382, 387, 397, 400, 404, 407416,
429430, 445, 467470, 473, 475476,
484, 561570, 574578, 580, 630
conventional ~ 14, 28, 37, 4243, 69,
397, 412415, 469
conversational ~ 14, 28, 37, 4243, 69,
179, 208, 382, 387, 397, 404,
407416, 445, 467470, 473, 580, 630
default ~ 42, 563564, 566567
generalized ~ 562, 564, 566, 569
scalar ~ 39, 400, 411, 562564, 567,
570
impliciture 28, 397, 411412, 430
indexical see indexicality
indexicality 45, 7, 13, 3133, 35, 5556,
116117, 119, 139, 141143, 146147,
159, 167, 183186, 207, 209210,
223, 238, 244245, 248249, 264265,
276, 315319, 320321, 323, 326328,
331333, 462, 466467, 600
708
Subject index
inference 45, 7, 1415, 2728, 3233,

35, 4144, 56, 59, 61, 63, 98, 223, 231,
266, 278, 319, 351352, 359, 361363,
397398, 401, 403, 411, 415, 425, 445,
448, 462, 465, 470, 474, 477, 484, 500,
565, 576, 579, 633, 638, 640, 646,
648
pragmatic ~ 51, 203, 205, 208215, 387,
412, 416, 469, 473, 475476, 563,
566567, 569570, 573575, 603, 636
intention 911, 13, 28, 3133, 3639, 41,
43, 44, 51, 54, 62, 6467, 69, 7879, 86,
95, 108112, 116, 124125, 144155,
169, 171, 173177, 180182, 187189,
193, 206, 209211, 213, 215223, 236,
271, 289291, 293, 295297, 299, 302,
307309, 322, 347, 349, 377, 379380,
382388, 392, 427, 445446, 448, 462,
467, 475, 526527, 567, 570, 572575,
578580, 594, 596597, 605
intentionality see intention
introspection 5, 8, 1617, 192, 461, 491,
539, 563, 596, 631
K
knowledge
metapragmatic ~ 119120, 122, 125,
432
L
language acquisition 87, 90, 92, 212, 236,
373, 448, 542, 544
lifeworld 291, 295, 298, 300302, 304,
306
locution 62, 296, 298, 380, 383, 390
locutionary see locution
Logical Positivism 159, 167, 374
M
maxim of conversation
exploiting ~ 4243, 208209, 409
flouting ~ 43, 114, 208, 409
manner ~ 37, 109, 114, 408, 468
quality ~ 37, 109, 114, 409410, 468
quantity ~ 37, 109, 114, 440, 468,
471472
relation ~ 37, 109, 114, 468
~ of relevance see relation ~
meaning
context-dependent ~ 24, 26, 34, 462
conventional ~ 41, 186188, 208,
219220, 222, 414, 467, 479, 570
hearer ~ 187188
intended ~ 2425, 33, 39, 176, 178, 187,
571
intersubjective ~ 219, 291292
lexical ~ 27, 352, 357, 477, 479
pragmatic ~ 9, 31, 3738, 77, 96, 187,
426, 437, 549, 552, 630631
sentence ~ 2, 170, 187, 222, 382, 385,
461, 484
speaker ~ 170, 187188, 411, 416, 561,
578, 635
subjective ~ 219, 223, 290291, 293,
309
utterance ~ 3, 27, 29, 31, 81, 183, 187,
322, 385
metacommunication 107109, 111113,
116, 119, 129130, 156157
metalanguage 108, 116, 129130, 140,
156, 171
metaphor 1516, 121124, 131, 213214,
232, 319, 321, 331, 385, 429, 444445,
478480, 483, 580, 609610
metapragmatics 10, 107108, 112, 117,
119, 125, 129131, 139, 147, 156
multimodal 1415, 18, 442, 449450,
502503, 543, 547549, 610, 629,
650651
multimodality see multimodal
N
narration 513514, 517519, 522, 530, 532
O
Ordinary Language Philosophy 27, 39,
153155, 159, 272
Organon model 141, 170, 232233, 235,
237238
origo of deixis see deictic center
P
participation framework 13, 271, 321
performative 25, 31, 3940, 55, 124, 139,
146, 148, 155159, 194, 294, 296,
375380, 390, 413, 436, 651
Subject index
~ verb 112, 120, 297, 377, 384
performativity see performative
perlocution 9, 23, 27, 40, 62, 127, 143, 148,
154, 177, 192193, 220222, 224,
298381, 387, 392
perlocutionary see perlocution
politeness 15, 27, 31, 38, 79, 81, 83, 8689,
9293, 95, 100, 109, 114, 157, 191, 277,
386, 441, 446, 538, 547, 597
polysemy 211214, 464
Positivism see Logical Positivism
pragmatic
~ competence 85, 89, 561, 566567,
574576, 578579
~ development 78, 84, 9091, 95, 99,
578
~ enrichment 27, 464, 475, 477
~ failure 43, 7778, 8488, 9193, 95,
99
~ intrusion 205, 412, 465, 473
~ transfer 7778, 84, 8792, 95, 99
pragmatics
development of ~ 12, 461, 561
empirical ~ 26, 292293, 552, 587588
formal ~ 12, 2627, 141, 292, 294,
299300, 302303, 305308
functional ~ 275, 278, 428, 548
general ~ 8, 25, 31, 38, 44, 8284, 89
historical ~ 17, 26, 7778, 84, 95,
98100
inferential ~ 467, 473
interlanguage ~ 26, 8890
literary ~ 16, 511
micro ~ and macro ~ 7, 9, 51, 5357,
6061, 63, 6668, 70
universal ~ 12, 292
presupposition 45, 79, 1516, 28, 31, 35,
51, 5556, 6065, 69, 79, 117, 119, 130,
139, 143, 153, 174, 192, 204, 297, 299,
304307, 397, 401405, 413416, 427,
465, 470, 481, 544, 561, 565, 580, 609,
638
pragmatic ~ 14, 3233, 57, 62, 397,
406407
~ trigger 32, 59, 402, 406
proposition 29, 3233, 35, 37, 40, 51,
55, 68, 112113, 147148, 150, 158,
172, 179, 188192, 203, 205206, 215,
709
296297, 304305, 349, 351352,

358, 379380, 383385, 397399,
401, 403407, 410, 413, 430431,
461467, 469, 473, 475476, 484,
567, 577, 589, 599, 601, 604, 607,
635636, 642643
propositional see proposition
prosodic see prosody
prosody 6, 15, 116118, 430, 438, 449,
491495, 497503, 539, 548549, 590,
592, 598, 600603, 608609, 629,
631635, 650, 664, 670671
semantic ~ 18, 539, 549, 608609, 629,
631635
R
reference 7, 11, 13, 28, 37, 40, 5859, 98,
109, 112, 117, 142, 148, 158159,
183185, 206, 315, 317318, 320,
322323, 327328, 330331, 333,
347350, 357, 359, 363364, 380,
442444, 462, 465, 481, 483, 501, 512,
519520, 544, 568569, 572, 637638,
640646
co-~ 317, 355356, 432, 447, 577578
~ resolution 355, 475, 646
referential see referentiality
referentiality 155, 127, 155, 157, 269,
350
reflexive (language) see reflexivity
reflexivity 111, 116, 130, 139, 155, 157,
217218, 264265, 267, 307, 320
relevance
conditional ~ 35, 267
principle of ~ 129, 179, 351, 474,
594
Relevance Theory 15, 33, 4142, 129, 157,
410413, 473, 475, 477479, 484, 567,
599, 636
rheme 39, 157, 229, 241, 248, 380,
433434, 603
rule
constitutive ~ 35, 39, 107
regulative ~ 35, 39
S
second language acquisition see language
acquisition
710
Subject index
semantic
~ preference 608609, 631, 635
~ prosody 18, 539, 549, 608609, 629,
631635
shifter 13, 116117, 129130, 141,
244246, 249, 316, 318, 327
sincerity condition 40, 125, 384385
stance 15, 320, 425, 437438, 442, 463,
493, 499500, 503, 517, 600, 651
T
theme 39, 157, 229, 241, 433434, 603
Theory of Communicative Action 178,
246, 290, 295, 298, 302303, 308309
transcription 6, 8, 1718, 391, 503,
590591, 597598, 601602, 629
systems of ~ 537, 547, 657672
truth
~ conditional see ~ conditions
~ conditions 29, 167, 172, 203,
205208, 211, 215, 348, 358, 374,
384, 401, 412, 414415, 466467,
469, 475, 479480, 484, 574
~ value 32, 207, 348, 358, 374376, 378,

398, 406, 466, 575, 638
turn 85, 94, 110, 113, 115, 117, 268, 430,
436, 438, 440442, 446, 494, 498501,
504, 517, 523, 605, 608, 650, 658660,
665, 668
~ constructional see ~ construction
~ construction 15, 435436, 441, 493,
496, 503504
~ continuation 436, 493, 497
~ -taking 15, 35, 66, 122, 266267, 272,
321, 392, 425, 435, 495496, 503,
531, 543, 548, 587, 591592, 597,
602604
U
underspecification 15, 41, 350351, 364,
464465, 477, 484
underspecified see underspecification
Z
Zeigfeld see deictic field

Foundations of Pragmatics PDF

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Foundations of Pragmatics PDF

Uploaded by

Copyright:

Available Formats

Foundations of Pragmatics

Bibliographic information published by the Deutsche Nationalbibliothek

Preface to the handbook series

Wolfram Bublitz, Andreas H. Jucker and Klaus P. Schneider

Preface to the handbook series

Preface to the handbook series . . . . . . . . . . . . . . . . . . . . . . . .

Introduction: the burgeoning field of pragmatics

Part I Conceptual foundations

Pragmatics as a linguistic concept

Micropragmatics and macropragmatics

Pragmalinguistics and sociopragmatics

Part II Theoretical foundations

The rise of pragmatics: a historiographic overview

Semiotic foundations of pragmatics

Pragmatics in modern philosophy of language

Foundations of pragmatics in functional linguistics

Foundations: ethnomethodology and Erving Goffman

Pragmatics in Habermas Critical Social Theory

Part III Key topics in pragmatic description

Deixis and indexicality

Reference and anaphora

Types of inference: entailment, presupposition, and implicature

Part IV The place of pragmatics in the description of discourse

Pragmatics and grammar

Pragmatics and semantics

Pragmatics and prosody: prosody as social action

Pragmatics and literature

Part V Methods and tools

Approaching the data of pragmatics

Corpus-based pragmatics I: qualitative studies

Corpus-based pragmatics II: quantitative studies

The transcription of face-to-face interaction

About the authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 681

Introduction: the burgeoning field of pragmatics

The aim of the handbook

The history of pragmatics

Wolfram Bublitz and Neal R. Norrick

Introduction: the burgeoning field of pragmatics

The definition of pragmatics

Wolfram Bublitz and Neal R. Norrick

Pragmatics is fundamentally concerned with communicative action in any kind of

Introduction: the burgeoning field of pragmatics

That pragmatics overcomes any narrow focus on language as a system of signs

The purview of pragmatics

Wolfram Bublitz and Neal R. Norrick

in a small corpus and suggesting hypotheses to be tested. For statistical purposes,

Introduction: the burgeoning field of pragmatics

Organization of this Handbook

To distinguish pragmatics as a linguistic discipline from pragmatics as an attitude or

Wolfram Bublitz and Neal R. Norrick

Part I Conceptual foundations

Introduction: the burgeoning field of pragmatics

Wolfram Bublitz and Neal R. Norrick

Part II Theoretical foundations

The opening chapter 5 in this section, The rise of pragmatics: a historiographic

Introduction: the burgeoning field of pragmatics

International Pragmatics Association, Journal of Pragmatics etc.). Finally, he

Wolfram Bublitz and Neal R. Norrick

Introduction: the burgeoning field of pragmatics

Part III Key topics in pragmatic description

Wolfram Bublitz and Neal R. Norrick

Part IV The place of pragmatics in the description of discourse