Argument and Rhetoric.

Adverbial Connectors in the History of English

Topics in English Linguistics
64
Editors
Bernd Kortmann
Elizabeth Closs Traugott
De Gruyter Mouton
Argument and Rhetoric.
Adverbial Connectors
in the History of English
by
Ursula Lenker
De Gruyter Mouton
ISBN 978-3-11-020558-9
e-ISBN 978-3-11-021606-6
ISSN 1434-3452
Library of Congress Cataloging-in-Publication Data
Lenker Ursula.
Argument and rhetoric : adverbial connectors in the history of
English / by Ursula Lenker.
p. cm. (Topics in English linguistics ; 64)
Includes bibliographical references and index.
ISBN 978-3-11-020558-9 (alk. paper)
1. English language Adverbials. 2. English language Con-
nectives. 3. English language History. 4. English language
Grammar, Historical. 5. English language Grammar. I. Title.
PE1326.L46 2010
4251.76dc22
2009051326
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.
2010 Walter de Gruyter GmbH & Co. KG, 10785 Berlin/New York
Cover image: Brian Stablyk/Photographers Choice RF/Getty Images
Typesetting: OLD-Media OHG, Neckarsteinach
Printing: Hubert & Co. GmbH & Co. KG, Gttingen
Printed on acid-free paper
Printed in Germany
www.degruyter.com
Acknowledgements
Adverbs are an oppressed part of speech, and probably bitter about their ill
treatment. If they were ever organized the upper class of nouns and verbs had
better look out (Landau 1984: 7879). The present book, which was accepted
as a Habilitationsschrift at the Ludwigs-Maximilians-Universitt Munich in
the winter term of 2006/2007 (this date basically also marks the end of the
full coverage of literature quoted), tries to remedy the ill treatment of at least
one class of adverbs: adverbial connectors. I am deeply indebted to the men-
tors and referees of this thesis, Professor Hans Sauer, Professor Ulrich Detges
and Professor Richard Janney, who provided invaluable information, advice
and encouragement while I was writing this book. I have to thank Dick Jan-
ney in particular for drawing my attention to the relevance of (your) little
pointing words, a remark which acted as a catalyst for my ideas on the im-
portance of different means of discourse deixis for the subject of adverbial
connectors. I am also very grateful to Hans Sauer and my father, Willibald
Lenker, without whose unyielding perseverance this monograph might never
have been written.
I am very deeply indebted to Helmut Gneuss, who gave me as good a start
in academic research as anyone could wish for: from my early student days, he
has always supported me in my various strands of academic interest (even in
felds he would never have researched himself), unfailingly providing a pletho-
ra of references and critical comments. I would also like to thank Alfred Bam-
mesberger, Manfred Markus, Judith Huber and Friedrich Heberlein for their
very helpful critical comments on an earlier version of this book, and the series
editors, Bernd Kortmann and Elizabeth Traugott, for accepting the book for
the Topics in English Linguistics. A number of readers and reviewers of earlier
versions of selected chapters, in particular Cynthia Allen, Walter Hofstetter,
Lucia Kornexl, Andreas Mahler, Anneli Meurman-Solin and Elizabeth Trau-
gott have substantially improved sections in this book with their pertinent re-
marks. At the very beginning of my research on this topic, conversations with
Otto Gsell and Klaus Schubert on trains and in the language centre corridor,
which seemed casual at the time, were decisive in encouraging me to focus on
the mixed bag of the word class adverb.
In the fnal stages of completing and proof-reading the manuscript, Stefanie
Beckstein, Veronika Bischofberger, Christina Griel, Christine Haunz and, at
Mouton de Gruyter, Wolfgang Konwitschny and Birgit Sievert gave wonderful
support. Steff Beckstein made these tedious weeks bearable with her infnite
patience and, in particular, her upbeat personality.
vi Acknowledgements
Only connect, from Howards End, was the slogan of a telephone company
some years ago and so, above all, I would like to thank all my friends and
family who helped me to stay connected to a world outside the realm of adver-
bial connectors during all these years. In particular, of course, I have to thank
Michael Burghart, who provided so much more than the music.
Munich/Eichsttt, December 2009
Contents
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
List of abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv
1. The framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1. Particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2. Earlier research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3. Aims of the study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4. Rhetoric and stylistic aspects: a sample study . . . . . . . . . . . . . . 6
1.5. Early Modern and Late Modern English . . . . . . . . . . . . . . . . . . 9
1.6. The inventory of adverbial connectors . . . . . . . . . . . . . . . . . . . . 9
1.7. The text corpus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.8. Presentation of corpus fndings . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.9. Outline of the present study . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2. Clausal connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.1. Clausal connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.2. Connectors connects: defnitions . . . . . . . . . . . . . . . . . . . . . . . 23
2.3. Connectors: coordinating conjunctions, subordinating
conjunctions, adverbial connectors . . . . . . . . . . . . . . . . . . . . . . . 25
2.4. Connectors and information processing . . . . . . . . . . . . . . . . . . . 28
3. The category adverb . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.1. Adverbs and adverbials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.2. Classifcation of adverbials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.2.1. Circumstance adverbials or adjuncts . . . . . . . . . . . . 36
3.2.2. Stance adverbials or disjuncts
(content/attitudinal and style disjuncts) . . . . . . . . . . . . . 36
3.2.3. Adverbial connectors
(linking adverbials or conjuncts) . . . . . . . . . . . . . . 37
3.3. Semantic categories of adverbial connectors . . . . . . . . . . . . . . . 39
3.4. Pure and impure connectives . . . . . . . . . . . . . . . . . . . . . . . . 41
3.5. Adverbial connectors in Present Day English: corpus fndings . 42
3.5.1. The corpus of the Longman Grammar . . . . . . . . . . . . . 42
viii Contents
3.5.2. Different types of adverbials over the core registers . . . 42
3.5.3. Syntactic realizations of adverbials . . . . . . . . . . . . . . . . 43
3.5.4. Positions of adverbials . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.5.5. Linking adverbials: distribution of semantic categories 44
3.5.6. Summary: Present Day English corpus fndings . . . . . . 45
3.6. Adverbial connectors discourse markers . . . . . . . . . . . . . . . . . 46
4. Adverbs and conjunctions in earlier metalinguistic thought 49
4.1. Introductory remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.2. The Greek and Latin tradition . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.3. lfrics Old English Grammar 51
4.4. Grammatical treatises of the Middle English period . . . . . . . . . 53
4.5. The prologue to the revision of the Wycliffte Bible . . . . . . . . . . 54
4.6. Early Modern English dictionaries and grammars . . . . . . . . . . . 56
5. Connectors in Old English . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
5.1. Semantic and syntactic polyfunctionality . . . . . . . . . . . . . . . . . . 58
5.2. Circumstance adverbs with connective force (Group B-a):
temporal and spatial adverbs . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
5.2.1. OE eft again . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
5.2.2. OE nu now (Group B-a, Group C) . . . . . . . . . . . . . . . . 61
5.2.2.1. Temporal nu . . . . . . . . . . . . . . . . . . . . . . . . . . 61
5.2.2.2. Text-deictic nu . . . . . . . . . . . . . . . . . . . . . . . . . 61
5.2.2.3. Adverbial connector nu transition/result 62
5.2.2.4. Adverb/conjunction nu in correlative
constructions . . . . . . . . . . . . . . . . . . . . . . . . . . 62
5.2.2.5. Conjunction (subordinator) nu . . . . . . . . . . . . . 63
5.3. Ambiguous adverbs/conjunctions (Group C) . . . . . . . . . . . . . . . 64
5.3.1. OE a 64
5.3.2. OE eah, swaeah, (swa)eahhwere
contrast/concession . . . . . . . . . . . . . . . . . . . . . . . . . 66
5.4. Position of adverbial connectors . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.4.1. Nacherstposition post-frst-position . . . . . . . . . . . . . 67
5.4.2. Present Day German adverbial connectors in
post-frst-position . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.4.3. Post-frst-position of adverbial connectors in Old English 70
5.5. Pronominal connectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
5.5.1. Pronominal connectors in Old English . . . . . . . . . . . . . 72
Contents ix
5.5.2. Pronominal connectors in the history of the Romance
languages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
5.5.3. Connectors in Louisiana French and French-based creoles 74
6. Adverbial connectors in the history of English . . . . . . . . . . . 76
6.1. Introductory remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
6.2. The diachrony of coordinating and subordinating conjunctions:
general tendencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
6.3. The diachrony of adverbial connectors: general tendencies . . . . 77
6.4. New adverbial connectors in the sub-periods of English:
methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
6.5. The inventory of Present Day English connectors: donor periods 80
6.6. Middle English: a period of experiment . . . . . . . . . . . . . . . . . . . 85
7. Adverbial connectors: morphology . . . . . . . . . . . . . . . . . . . . . 88
7.1. The expansion of the English lexicon . . . . . . . . . . . . . . . . . . . . . 88
7.2. Major morphological changes . . . . . . . . . . . . . . . . . . . . . . . . . . 92
7.3. Simple adverbs, compounds and derivations . . . . . . . . . . . . . . . 94
7.3.1. Simple adverbs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
7.3.2. Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
7.3.3. Derivations: adverbs in -lice, es and ways/wise . . . . . 95
7.4. Pronominal connectors lexicalized phrases . . . . . . . . . . . . . . 96
7.5. Pronominal connectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
7.5.1. Pronominal connectors in Old English . . . . . . . . . . . . . 98
7.5.2. Pronominal connectors in the history of English . . . . . . 100
7.6. Lexicalized (prepositional) phrases . . . . . . . . . . . . . . . . . . . . . . 102
7.6.1. General tendencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
7.6.2. Lexicalization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
7.6.3. Verbal and nominal phrases . . . . . . . . . . . . . . . . . . . . . . 104
8. Cognitive source domains . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
8.1. General tendencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
8.2. Source domain time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
8.3. Source domain place/space . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
8.4. Source domain truth/fact . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
8.4.1. General tendencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
x Contents
8.4.2. Conversational implicatures: concession/contrast . . 114
8.4.3. transition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
8.4.4. OE solice ME soothly . . . . . . . . . . . . . . . . . . . . . . . . 117
8.4.5. ME forsooth(e) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
8.4.6. OE treowlice ME/EModE/PDE truly . . . . . . . . . . . . . 124
8.4.7. PDE indeed and in fact . . . . . . . . . . . . . . . . . . . . . . . . . 127
8.4.8. Regularities in change . . . . . . . . . . . . . . . . . . . . . . . . . . 128
8.4.9. Truth, facts and communicative principles . . . . . . . . . . 128
9. Shifting deictics in the history of English causal connection 131
9.1. Earlier research on causals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
9.2. Causal connectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
9.2.1. The relation cause: cause result vs. result cause 133
9.2.2. Present Day English causal connectors: corpus fndings 134
9.2.3. Causal connectors: word classes and topology . . . . . . . 134
9.2.4. Semantic and pragmatic parameters . . . . . . . . . . . . . . . 136
9.2.5. Information processing . . . . . . . . . . . . . . . . . . . . . . . . . 138
9.3. OE form, foron, fory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
9.3.1. Forms and functions of form, foron, fory . . . . . . . 140
9.3.2. Expressions for causal relations in Early and Late West
Saxon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
9.4. Discourse deixis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
9.4.1. Form: morphological make-up and discourse deixis 147
9.4.2. Pronominal connectors in Present Day German . . . . . . 149
9.4.3. Deictic elements in English causal connectors . . . . . . . 151
9.5. Causal connectors in the history of English . . . . . . . . . . . . . . . . 152
9.5.1. Causal connectors in English translations of Boethius
De Consolatione Philosophiae . . . . . . . . . . . . . . . . . . . 152
9.5.2. Causal adverbial connectors in the history of English . . 154
9.5.3. Deixis in new adverbial connectors . . . . . . . . . . . . . . . . 157
9.5.4. ME for that . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
9.5.5. Recursive for Latin nam/enim . . . . . . . . . . . . . . . . . 161
9.5.6. Subordinators: for as much as, since, because . . . . . . . 164
9.6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
10. contrast/concession . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
10.1. cause condition contrast: affnities . . . . . . . . . . . . . . . . 168
10.2. concession contrast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
Contents xi
10.2.1. concession . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
10.2.2. contrast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
10.2.3. concession/contrast: information structure . . . . . . . 171
10.2.4. Subordinators adverbial connectors: mood distinctions 173
10.3. Cross-linguistic patterns in the origin of concessive connectors 173
10.4. The diachrony of English contrastive/concessive adverbial
connectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
10.4.1. General tendencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
10.4.2. Antithetic/reformulatory adverbial connectors . . . . . . . 175
10.4.3. Contrastive/concessive connectors in English:
general tendencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
10.4.4. Shifting deictics in English contrastive/concessive
connectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
10.4.5. Patterns in the origin of English concessive connectors 181
10.5. OE eah PDE though 184
10.5.1. Long-term developments: grammaticalization . . . . . . . 184
10.5.2. OE eah 186
10.5.3. Contrastive adverbial connectors from Middle to
Present Day English . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
10.5.3.1. Though 190
10.5.3.2. Yet and still . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
10.5.3.3. However 194
10.6. Sentence-fnal connectors in Present Day English . . . . . . . . . . . 197
10.6.1. Corpus fndings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
10.6.2. Information structure . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
10.6.3. Sentence-fnal connectors in Present Day German . . . . 198
10.6.4. Sentence-fnal connectors in the history of English . . . . 200
10.7. Present Day English sentence-fnal though . . . . . . . . . . . . . . . . 200
10.7.1. Earlier research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
10.7.2. PDG obwohl 202
10.7.3. PDE but 203
10.7.4. Although and though in the LLC 207
10.7.4.1. Quantitative fndings . . . . . . . . . . . . . . . . . . . . 207
10.7.4.2. Functions of the subordinator (al)though 208
10.7.4.3. Functions of sentence-fnal though 210
11. addition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
11.1. General tendencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
11.2. The diachrony of reinforcing adverbial connectors . . . . . . . . . . 214
xii Contents
11.2.1. General tendencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
11.2.2. Also 215
11.2.3. New coinages: item, plus, too 219
11.2.4. Iconic principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
11.3. Equative connectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
11.4. Summative connectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
11.5. Appositive connectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
11.6. Enumerative (listing) connectors . . . . . . . . . . . . . . . . . . . . . . . . 222
11.7. Genre-dependency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
12. transition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
12.1. Preliminary considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
12.2. Source domain uncertainty/doubt:
peradventure, perchance, perhaps . . . . . . . . . . . . . . . . . . . . . . . 228
12.3. Interrogatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
12.4. The diachrony of transitional connectors . . . . . . . . . . . . . . . . . . 230
12.4.1. Paradigm shift: truth fact . . . . . . . . . . . . . . . . . . . . 230
12.4.2. Regularities in semantic change . . . . . . . . . . . . . . . . . . 231
13. Perspicuity and the New Rhetoric . . . . . . . . . . . . . . . . . . . . 233
13.1. Collocations vs. medial position of adverbial connectors . . . . . . 233
13.2. Sentence-initial collocations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
13.3. Medial positions of adverbial connectors . . . . . . . . . . . . . . . . . . 235
13.4. Corpus fndings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
13.4.1. Present Day English . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
13.4.2. Old English to Late Modern English . . . . . . . . . . . . . . . 238
13.5. Copia perspicuitas 241
13.5.1. Perspicuity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
13.5.2. Punctuation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
13.5.3. The Scottish Rhetoricians: The New Rhetoric . . . . . . 243
13.5.4. George Campbells The Philosophy of Rhetoric (1776) . 243
14. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
1. Corpora . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
2. Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
Contents xiii
3. Dictionaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
4. Secondary sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
Appendix
Appendix A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
A.1. Adverbial connectors: items . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
A.2. Alphabetical index of adverbial connectors . . . . . . . . . . . . . . . . 282
Word index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
Subject and name index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
Appendices on CD-ROM
Appendix B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
B.1. Listing/additive adverbial connectors . . . . . . . . . . . . . . . . . . . . . 295
B.2. Summative adverbial connectors . . . . . . . . . . . . . . . . . . . . . . . . 307
B.3. Causal/resultive adverbial connectors . . . . . . . . . . . . . . . . . . . . 308
B.4. Contrastive and concessive adverbial connectors . . . . . . . . . . . . 317
B.5. Transitional adverbial connectors interjections . . . . . . . . . . . . 329
Appendix C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
C.1. Corpus texts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
C.2. Selected corpus texts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
C.2.1. Treatises and Homilies . . . . . . . . . . . . . . . . . . . . . . . . 353
C.2.2. Translation of Boethius
De Consolatione Philosophiae . . . . . . . . . . . . . . . . . . . 358
List of abbreviations
A Adverbial
BT Bosworth & Toller (18821898)
ByrM Byrhtferths Enchiridion (ed. Baker & Lapidge 1995)
CE- Initial Letters of Corpus Texts from Early Modern English
CH Clark Hall 1984
CL- Initial Letters of Corpus Texts from Late Modern English
CLMET A Corpus of Late Modern English Texts (de Smet 2005)
CM- Initial Letters of Corpus Texts from Middle English
CME Corpus of Middle English Prose and Verse
CO- Initial Letters of Corpus Texts from Old English
Conj. Conjunction
DOE Dictionary of Old English
DOEC Dictionary of Old English Corpus
EETS Early English Text Society
EMEDD Early Modern English Dictionary Database
EModE Early Modern English
EModE1 (15001570), EModE2 (15701640), EModE3
(16401710)
Eustace Life of St Eustace; ed. Skeat 1900: 190219
HC Helsinki Corpus
Ind. Indicative
Lat. Latin
LLC London Lund Corpus of Spoken English
LModE Late Modern English
LModE1 (17101780), LModE2 (17801850), LModE3
(18501920)
LOB LancasterOslo/Bergen Corpus of British English
M Merkmal = property
ME Middle English
ME1 (11501250), ME2 (12501350), ME3 (13501420),
ME4 (14201500)
MED Middle English Dictionary
O Object
OE Old English
OE1 ( 850), OE2 (850950), OE3 (9501050),
OE4 (10501150)
xvi List of abbreviations
OED Oxford English Dictionary
ON Old Norse
O. S. Old Series
PDE Present Day English
PDG Present Day German
PP Prepositional Phrase
Prs. Person
S Subject
Sg. Singular
Subj. Subjunctive
V Verb (Predicate)
V2 Verb-second
Symbols
ambiguous adverb/conjunction
[ ] in Appendix A.1: not attested in the corpus texts
1. The framework
1.1. Particles
In his Essay Concerning Human Understanding, one of the frst linguistic ac-
counts explicitly addressing what we now call textual organization or cohesion,
1

John Locke asserts that it is the right use of Particles which is crucial for the
clearness and beauty of good style:
These Words, whereby it signifes what connection it gives to the several Affrma-
tions and Negations, that it unites in one continued Reasoning or Narration, are
generally calld Particles: and tis the right use of these, that more particularly con-
sists the clearness and beauty of good Stile. And to express well such methodical
and rational Thoughts, he must have words to shew what Connexion, Restriction,
Distinction, Opposition, Emphasis, etc. he gives to each respective part of his Dis
course (Locke [1690] 1975: 471; emphasis by Locke).
As is evident from this description, Lockes Particles are co-referential with
the linguistic elements we now call clausal connectives.
2
Locke highlights
that he regards these elements not merely as an adornment or frequent supple-
mentary device to uncover textual organization,
3
but as indispensable for the
understanding of an utterance:
To mistake in any of these, is to puzzle, instead of informing, his Hearer: and there-
fore it is, that those words, which are not truly, by themselves, the names of any
Ideas, are of such constant and indispensable use in Language (Locke [1690]
1975: 471).
The novelty and originality of this view that Particles play a central role in
language is made explicit when Locke proceeds by criticizing previous gram-
mars and grammarians for their ill-treatment or even neglect of these linguistic
elements:
1 For details on the treatment of connectives in European and English language
scholarship, see Lenker (2003) and below, Chapter 4.
2 The term connective is used as an umbrella term for all kinds of linguistic items
signalling a linkage of sentences or chunks of discourse. The term connector more
specifcally refers to paratactic connectives, in particular adverbial connectors.
3 For the repeatedly expressed view of adverbial connectors and so-called discourse
markers as supplementary or even superfuous features of language, and for a discus-
sion of the overlap between coordinators, subordinators, adverbial connectors and
discourse particles as well as cases of polyfunctionality, see Chapters 2 and 3.
2 The framework
This part of Grammar has been, perhaps, as much neglected, as some others over-
diligently cultivated. Tis easy for Men to write, one after another, of Cases and
Genders, Moods and Tenses, Gerunds and Supines: But though Prepositions
and Conjunctions, etc. are names well known in Grammar, and the Particles con-
strained under them carefully ranked into their distinct subdivisions; yet he who
would shew the right use of Particles, and what signifcancy and force they have,
must take a little more pains, enter into his own Thoughts, and observe nicely the
several Postures of his Mind in discoursing (Locke [1690] 1975: 471472).
1.2. Earlier research
More than 300 years after Lockes attempt to draw attention to the importance of
connectives and to foster their attention with grammarians, we fnd that this feld
is still neglected in linguistics.
4
In 2003, the compendium on German connectives,
the Handbuch der deutschen Konnektoren, has to state in an almost identical way:
Mit der Wahl dieses Gegenstandsbereiches behandelt das Handbuch Einheiten des
deutschen Wortschatzes [= Konnektoren connectives], die weder in Grammati-
ken noch Wrterbchern noch in den texttheoretischen und konversationsanalyti-
schen Arbeiten befriedigend beschrieben sind. Wie nicht anders zu erwarten ist,
werden in Grammatiken nur die systematischen Eigenschaften der betreffenden
Einheiten beschrieben. Dabei stehen dort jedoch traditionell in der Regel semanti-
sche Klassenbildungen im Vordergrund (Pasch et al. 2003: xv).
The subject area chosen by the handbook are items of the German vocabulary
[=connectors, U. L.] which have neither been satisfactorily described in dictionar-
ies, nor in studies on text theory and conversation analysis. As expected, it is only
the systemic properties of the respective items which are described in grammars.
Traditionally, however, these descriptions generally focus on the distinction of vari-
ous semantic classes.
This summary testifes to a lack of synchronic in-depth studies in the feld of
connectives.
5
It is, however, in particular the history of connectives that has
received only very little attention. With regard to the history of the English
language and the establishment of modern English prose, there are only two
studies which explicitly approach this feld diachronically in a wider and sys-
tematic perspective:
4 For the frst fuller treatment of connectives in Campbell ([1776] 1963), see below,
Chapter 13.5.4.
5 For English, notable exceptions are the very comprehensive studies by Altenberg,
in particular Altenberg (1984, 1986).
Earlier research 3
In a study situated at the interface between literary studies
6
and linguistics, Syl-
via Adamson (1999) relates the emergence of new forms of sentential connection
to major changes in literary style and text production in Early Modern England,
namely to the evolution of the plain style as an indicator of perspicuitas an
idea connected with the stylistic ideals of the Royal Society in contrast to earlier,
very different stylistic ideals such as copia (for details, see below, Chapter 13.5).
In the feld of linguistics proper, Bernd Kortmanns study of adverbial subor-
dinators (1997), though predominantly concerned with typological and cross-
linguistic data (see the material collected in 1997b), also gives a short outline
of the history of adverbial subordinators, one of the most important groups of
connectives in English. The results of Kortmanns study of the history of sub-
ordinators will here be used as a contrastive plane for the comparison with the
fndings for adverbial connectors.
7
In addition to these more general accounts, there is a small number of stud-
ies examining individual items, concepts and relations in the feld, such as
Jucker 1997 (well), Enkvist & Wrvik 1987 (OE a; and many further studies
on OE a such as, e. g., Kim 1992), Brinton 1996 (on various pragmatic mark-
ers), Lenker 2000 (OE solice, witodlice), Lenker 2003, 2007a (forsooth),
Markus 2000 (ME wherefore, therefore, etc.), sterman 1997 (there-com-
pounds), Rissanen 1999a (rather), Schleburg 2002 (OE swa), Stanley 2000 (OE
hwt), Traugott 1997 (after all), Traugott and Dasher 2002 (indeed, in fact) and
Fischer 2007 (instead; indeed, in fact, solice and witodlice).
The semantic relation analysed most thoroughly is that of concessives which
features prominently in studies of the semantic-pragmatic approach to gram-
maticalization (see Knig and Traugott 1982; Knig 1985a, 1985b, 1988; Barth-
Weingarten and Couper-Kuhlen 2002). In addition to the still central study on co-
hesion in Present Day English by Halliday and Hasan (1976), there are a number
of studies on Present Day English discourse markers, which refer to the history
of some of the adverbial connectors in question (e. g., Schiffrin 1987; Lenk 1998).
With the exception of Halliday and Hasans systematic analysis of conjunctive
elements in Present Day English (1976: 226273), these studies concentrate on
single connectives or patterns of language change, such as grammaticalization,
and do thus not endeavour to provide a comprehensive treatment of connectives.
A more integrated view of clausal connection in the history of English is
offered in most of the papers presented at the workshop Clausal Connection in
6 We also fnd some relevant information on connectives scattered over more general
works on the evolution and establishment of English prose (see, for instance, Muel-
ler 1984, McIntosh 1998 and Robinson 1998).
7 The present study is very much indebted to Kortmanns study and was set off by my
review of his book. For the details of the comparison, see in particular Chapter 6.
4 The framework
the History of English at the 13th International Conference on English Histori
cal Linguistics at Vienna in 2003. This workshop was initiated because the con-
venors Anneli Meurman-Solin and Ursula Lenker regarded it as a drawback that
the typologies of clause-combining devices in English as well as other European
languages are widely discussed in recent literature (Devriendt, Goossens, and
van der Auwera 1996; Kortmann 1997; van der Auwera 1998), but have chiefy
been construed by using secondary sources such as dictionaries and grammars
(for this grammar-cum-dictionary-method, see, e. g., Kortmann 1997: 53).
Most of the workshop papers are therefore corpus-based studies of various con-
nectives in the history of English, focusing on subordinators (while, lest, since,
albeit) or specifc semantic domains, such as concessives or conditionals (see
Lenker and Meurman-Solin 2007). There is as yet, however, no comprehensive,
corpus-based treatment of adverbial connectors in the history of English.
1.3. Aims of the study
The present study tries to fll at least a segment of this large gap in historical
(English) linguistics by corpus-based analyses of the development of a particular
word class in connector function namely adverbial connectors, which are also
called conjuncts (Quirk et al. 1985) or linking adverbials (Biber et al. 1999).
8

The focus of the study therefore rests on the inventory and use of linguistic ele-
ments which explicitly mark textual cohesion on a level higher than the phrase.
9

Even more specifcally, the focus is on connectives which signal textual organi-
zation on a more global level in discourse, i. e. a level higher than the sentence.
10
In Lehmanns universal typology of clause linkage (Lehmann 1988), ad-
verbial connectors belong together with subordinators and coordinators (i. e.
coordinating and subordinating conjunctions) to the little integrated, explicit
markers of clausal connection (in contrast to embedded constructions, rela-
tive connectives, non-fnite verb forms, absolute constructions, etc.).
11
While
8 For the terminology employed in various grammars, see below, Chapter 3.
9 Some of these items can also mark relations on the level of the phrase; these are
only included, however, if they are polyfunctional and if they can also function on
the higher level of textual organization.
10 This position at the interface between the sentence and the text (paragraph) may
be one of the reasons why they are a neglected subject in linguistics. Theories of
syntax, such as all kinds of generativist approaches and also many functional per-
spectives, ignore them because they are above the level of the core sentence.
11 Most of these other strategies fulflling a connective function, such as linking by
non-fnite constructions (present participles, infnitives etc.) extra-textual links and
non-linguistic structuring devices, which are not in the focus of this study, do not play
Aims of the study 5
the distinction between coordinators, subordinators and adverbial connectors
is rather clear-cut for Modern English,
12
it has to be stressed here right at the
beginning of this study that there was no such clear distinction in Old English,
which predominantly uses so-called ambiguous adverbs/conjunctions (for
details, see below, Chapter 5).
This frst of all means that the whole system of clausal connection has been
re-structured in the history of English. While other, in particular typologically
related, languages such as German, show much more stability in the morpho-
logical make-up and use of subordinating conjunctions and adverbial connectors,
English has only very few remnants of the older Germanic system present in
Old English. The Old English inventory mainly consisted of polyfunctional items
comprising an explicitly deictic element. In addition to forms comprising swa so
(cf. swaeah or swaeahhwere although; nevertheless), we very frequently
fnd morphologically complex connectors comprising a pronominal element,
such as OE form, foron or fory for, because; therefore (< for [preposition]
because of + m/on/y [demonstrative; dative m, instrumental on/y]).
13
This scenario suggests that the frst essential motivating force for the dramat-
ic changes in the inventory of adverbial connectors starting in the early Middle
English period are structural constraints and therefore typological properties
of English.
14
The break-down of the case and gender infections of both of
the Old English demonstratives and, in particular, the use of that as a demon-
strative, complementizer and general subordinating particle (replacing OE e)
triggered new structures which signal anaphoric relations on the surface, dis-
ambiguating subordinators from adverbial connectors. The complex problems
involved in drafting this new inventory marking discourse deixis are refected
an important role in the history of the development, because they are either rather
rare (e. g. full prepositional phrases) or emerge very late in the history of English, i. e.
in the Early Modern English period when non-fnite constructions become increas-
ingly grammaticalized. Only from LModE2 onwards, infnitives (e. g. to begin with,
to conclude, to proceed) become more popular. In Present Day English, they do not
amount to more than four per cent of linkage constructions (Altenberg 1984: 47).
In the present sample from LOB-D and FROWN-D (68,000 words each), the token
numbers are: infnitives (3), present participles (9), and past participles (7).
12 A notable exception is the status of PDE for; see Chapter 9.
13 The changes in the morphological make-up of these forms and their relevance are
discussed in Chapter 7. In Raible (1992: inserted foldout), these pronominal con-
nectors form a typologically separate group: Junktion durch Wiederaufnahme
(eines Teils) des vorhergehenden Satzes (II).
14 For similar typological changes involving the loss of deictic elements in French
creoles see Raible (1992: 172177) and below, Chapter 5.5.
6 The framework
in the manifold forms in Early Middle English (ME 1/2). The innovations in
these periods, however, still show symptomatic regular, pronominal patterns,
comprising the new demonstratives this and that (cf. additive, reinforcing over
that / over this or resultive for that / this). Early Middle English thus emerges as
a period of experiment and transition, and hence as the frst decisive period for
the re-structuring of the system of adverbial connectors in the history of Eng-
lish. This re-structuring also led to the loss of polyfunctionality in connectives
and a distinction between coordinators and subordinators, which will, as one
exemplary case, be followed in the development of the system of causal con-
nectives in the history of English in Chapter 9 (i. e. the development of because,
since and as in certain positional variants and contexts); the other sections of
the book focus primarily on adverbial connectors.
1.4. Rhetoric and stylistic aspects: a sample study
The following introductory example of an (unfortunately rather small) experi-
ment conducted by Mauranen (1993: 159168) shows that a study of adverbial
connectors cannot be confned to morphological, semantic and syntactic analy-
ses. Mauranen presented the following text (Text 1) to a group of English native
speakers with professional linguistic training. All of the participants worked
as English lecturers at universities or as lectors or revisers for technological re-
search reports and were selected because they were considered to be extremely
sensitive to language and textual organization.
The subjects were frst invited to read the text and were then, in a second
step, placed in groups of two and three and asked to come to a group decision
on the following issues: whether the text was correct as it stood, and, secondly
and, more specifcally, if they felt that connectives should be added in any of
the sentences.
Text 1 (Mauranen 1993: 164)
In a recent study on 5- to 6-year-old children, Astington confrmed that they see
a strong link between promising something and actually doing it: To promise
means you do it. This link is much stronger for children than for adults, which
leads children to assert that an unfulflled promise was not a promise, but a lie.
For young children promising is not simply a speech act but something that includes
execution of the promised action as well.
It seems likely that children of 5 years or so understand that the use of the word
promise entails commitment, but this understanding may be based on a simple rule
If you have said I promise then you must do what you said you would. Although
the use of the word promise may often be a suffcient condition for becoming com-
mitted, it is certainly not a necessary condition. It is seldom used in everyday ex-
Rhetoric and stylistic aspects: a sample study 7
changes between adults, who tend to say I will meet you at 6 oclock, Ill return
your book tomorrow, and so forth. The commitment, although informal, is binding.
What distinguishes a commitment (I will return your book) from a statement of
intention (I will stay in tonight) or a prediction of a future event (I will get wet) is
the knowledge that someone else is relying on one to carry out the commitment and
the knowledge that the other person is aware that one has made a commitment. The
interesting question is whether young children can recognize commitment without
the help of the word promise. Our aim in this study was to investigate whether child-
ren of between 5 and 10 years are aware of reliance as the essential and necessary
condition for commitment.
After having read and discussed Text 1, the groups unanimously concluded that
the text was good as it stood. They reported that they felt it was clear, easy
to read and that it required no additions (Mauranen 1993: 166). They also
agreed that no connectives needed to be added at all.
In a second phase, however, the groups were shown the authentic version of
the text, which had been altered by Mauranen for the frst test phase. For Text
1, Mauranen had deleted connectives which were dispensable, i. e. which are
not crucial for the grammatical correctness of the text because they are peri-
pheral to the sentence structure and act on a level higher than the local phrase,
marking textual cohesion globally. The authentic text (Text 2) with the restored
connectors reads as follows:
Text 2 (Mauranen 1993: 164)
In a recent study on 5- to 6-year-old children, Astington confrmed that they see a
strong link between promising something and actually doing it: To promise means
you do it. However, this link is much stronger for children than for adults, which
leads children to assert that an unfulflled promise was not a promise in the frst
place, but, rather, a lie. In other words, for young children promising is not simply
a speech act but something that includes execution of the promised action as well.
It seems likely, then, that children of 5 years or so understand that the use of the
word promise entails commitment, but this understanding may be based on a sim-
ple rule, such as If you have said I promise then you must do what you said you
would. Although the use of the word promise may often be a suffcient condition
for becoming committed, it is certainly not a necessary condition. Indeed, it is
seldom used in everyday exchanges between adults, who tend to say I will meet
you at 6 oclock, Ill return your book tomorrow, and so forth. The commitment,
although informal, is nonetheless binding.
What distinguishes a commitment (I will return your book) from a statement of
intention (I will stay in tonight) or a prediction of a future event (I will get wet) is
the knowledge that someone else is relying on one to carry out the commitment
and, furthermore, the knowledge that the other person is aware that one has made
a commitment. The interesting question is, therefore, whether young children can
8 The framework
recognize commitment without the help of the word promise. Our aim in this study
was to investigate whether children of between 5 and 10 years are aware of reliance
as the essential and necessary condition for commitment.
The linguistic elements which had been removed in Text 1 are underlined here
for reasons of transparency. They are spread over the whole text with a ratio
of about one deletion per sentence. No deletions were undertaken in the frst
and the last sentence of the text, as these sentences set the frame for the textual
cohesion of the passage and therefore do not show any employment of connec-
tives on the level above the sentence.
After having read the original version with all the connectives restored (Text
2), Mauranens subjects were asked to comment on the text again and evaluate
it. The immediate reaction of the subjects was that they felt a dramatic differ-
ence between the two versions.
15
More specifcally, the emphasis was seen to
have changed, and the text was said to be not only easier to read, but also to
be more logical and more convincing. It was also perceived to have more
authority (Mauranen 1993: 167168).
The fact that the test groups registered a dramatic difference between the
two versions of the text with the connectives having an overall effect of mak-
ing Text 2 more authoritative, logical and convincing shows that a study
of the history of connectives in English must certainly not be confned to their
etymology, morphology and semantics, but has to be situated at the interface
between syntax, semantics, pragmatics, text linguistics and rhetoric. The cen-
tral role or even indispensability of these connectives in the construction of
a text, as highlighted by Locke, is corroborated by Mauranens experiment,
which also shows that it is necessary to view the employment of these items in
the wider perspective of the evolution of an English prose style. Yet, the prefer-
ence of certain means of textual organization and the employment of linking
adverbs as connectives is certainly not the only crucial driving force in the
developments of adverbial connectors in the history of English. In Present Day
English, for example, an author may choose among the adverbials therefore,
thus and hence for marking the semantic relation result, and their use indeed
exhibits some variability by author: most academic texts show a clear prefer-
ence for either thus or therefore, usually using one item at least twice as often
as the other (Biber et al. 1999: 889). This choice, however, is merely a choice in
the lexical material employed. In both cases, the authors do not choose to mark
the organization of the text by coordinating or subordinating conjunctions (be
cause, since, as or for), but by adverbial connectors. Although some changes
15 For the importance of stylistic considerations, see also Lehmann (1988: 210213).
For an early discussion, see Campbell (1776: 384415) and below, Chapter 13.5.4.
Early Modern and Late Modern English 9
and choices in the use of adverbial connectors may indeed be due to stylistic
predilections of a period or of individual authors (see below, Chapter 13), the
more crucial changes in the history of English connectors thus seem to have
been triggered by the typological and structural changes which set English
also in this respect apart from other Germanic languages.
1.5. Early Modern and Late Modern English
After the period of experiment and variation in early Middle English, there are
two further periods which are decisive for shaping the system and the use of
adverbial connectors in English. In Early Modern English (EModE 1/2), when
English develops into an Ausbausprache gradually being used as a national
language also in the written medium, a number of new connectors are formed.
In contrast to those coined in the Early Middle English period, however, these
are not replaced again, but for the most part remain in the language until Pres-
ent Day English (see Chapter 6.5).
The English of the late eighteenth and nineteenth centuries (Late Modern
English 2 and 3; 17801850 and 18501920) then sees crucial changes in (a)
the preference of adverbial connectors over coordinating conjunctions and their
collocations, and (b) the sentence position of adverbial connectors (see Chapter
13.113.3). Instead of sentence-initial position very often in collocation with a
conjunction such as and or but or a connective such as for , adverbial connec-
tors become increasingly used in medial position. This medial position, how-
ever, is virtually only found in the written medium and is attested in less than 2.5
per cent of instances in the spoken medium (see Biber at al. 1999: 891): it thus
sets the English of the written medium apart from that of the spoken medium
(see Chapter 13.4). In spoken interaction, however, we also fnd a new position
for adverbial connectors: adverbial connectors such as however and, in particu-
lar, though are increasingly placed at the very end of a sentence from the middle
of the twentieth century onwards. The motivating forces and consequences of
the sentence-fnal positioning of adverbial connectors will be exemplifed by
a contrastive analysis of the use of PDE although, though, and sentence-fnal
though and German obwohl with main clause word order in Chapter 10.5.
1.6. The inventory of adverbial connectors
The present study covers many of the issues suggested by these introductory
examples. It gives an as complete as possible account of the single-word or
lexicalized linguistic items which have been used as adverbial connectors in the
10 The framework
history of English. Furthermore, it describes their employment as well as their
meaning and their text structuring and information processing functions in rela-
tion to other coordinators, such as coordinating and subordinating conjunctions.
The approach employed is therefore a decidedly diachronic and functional
one. The study is mainly focussed on single-word adverbs because corpus fnd-
ings for Present Day English show that it is single adverbs which are predomi-
nantly employed in this function (Biber et al. 1999: 887, and Figure 10.26), al-
though this adverbial linking function may theoretically be realized not only by
single adverbs, but also by prepositional phrases, such as for that reason, in other
words, or on the contrary. The distinction between prepositional phrases and
single adverbs, however, is, even in Present Day English, by no means clear-cut;
in Present Day American English, for example, the lexicalization of the phrase
after all is indicated in its spelling afterall as one word (see OED, s. v. after; 10.):
(1) Afterall, the movement of people, not vehicles, is what counts. (1976
Billings [Montana] Gazette 1 July)
Similarly, in Middle English, the phrase at /on /of last fnally is attested as alast
(see MED, s. v. a-last). Even longer phrases such as over and above further-
more, a popular reinforcing connector in Early Modern English, could be spelt
as a single word hoverendebuv (see OED, s. v. over and above).
Furthermore, since many of todays single word items are univerbated and lexi-
calized prepositional (e. g. indeed) or verbal (e. g. albeit, howbeit, notwithstanding)
phrases and since a distinction, especially in periods without a fxed orthography,
is not unambiguously possible, lexicalized prepositional phrases were included
in the detailed analysis of this book. Only fully transparent, i. e. non-lexicalized
phrases, such as in other words, for that cause, or the sum is, have been excluded
(for the respective morphological and syntactic criteria, see Chapter 7.4).
1.7. The text corpus
The inventories of adverbial connectors of the respective periods were frst
gathered by a search in the relevant dictionaries (OED for the history of Eng-
lish; CH and BT for Old English, MED for Middle English and the EMEDD
for Early Modern English) as well as in the grammars of the earlier periods of
English (Mitchell 1985; Mustanoja 1960; Kerkhof 1982; Franz 1939) and of
Present Day English (Quirk et al. 1985; Biber et al. 1999; Huddleston and Pul-
lum 2002; Halliday and Hasan 1976). While it was thus possible to gain a good
overview over the system of adverbial connectors in Old, Middle, Early Mod-
The text corpus 11
ern and Present Day English, the Late Modern English Period, i. e. the English
of the eighteenth and the nineteenth century, remained a neglected period.
The OED, for example, whose frst edition was compiled in the nineteenth
century, often does not have special entries for phrases which were at that time
not yet lexicalized or had only recently been lexicalized, such as after all, above
all, in all events, at any rate (see OED, s.vv. after, all, event, rate). Generally,
the varieties of the late eighteenth to the early twentieth century had until re-
cently to be described as the Cinderellas of English historical linguistic study
(Jones 1989: 279). Although the general interest in the language of an increas-
ingly more distant past and the change of emphasis within historical linguistics
to socio-historical and corpus-based approaches has led to a surge of interest in
Late Modern English, the more recent publications on the feld (Bailey 1996;
Romaine 1998; Lass 1999; Grlach 1999; Grlach 2001; Beal 2004) do not treat
textual cohesion in any detail. This lack of material on Late Modern English,
together with the plan to study not only the inventory, but also the changes in
use and position of adverbial connectors in the history of English, created the
need for a diachronic corpus-based approach, including texts from Late Mod-
ern English, which are not covered by the Helsinki Corpus of English Texts.
In order to study not only changes in the inventory, but more generally de-
velopments in the use of adverbial connectors (also in contrast to conjunctions)
from Old English to Present Day English, the fndings of the present study are
based on a corpus of texts from all periods of English (see Appendix C.1) and,
for the quantitative analyses, on two smaller corpora of selected comparable
texts (see Appendix C.2.1. Treatises and Homilies and C.2.2. Translations
of Boethius De Consolatione Philosophiae).
The complete corpus is modelled on the Helsinki Corpus, which comprises
texts from the earliest Old English period until 1710 (see Kyt 1996). For the
periods no longer covered by the Helsinki Corpus, I have compiled a corpus
in 70-year sub-periods (equivalent to the periods from ME3 to EModE3 of the
Helsinki Corpus) until 1920 (LModE1 from 1710 to 1780, LModE2 from 1780
to 1850, LModE3 from 1850 to 1920). For each of these periods, I extracted
5,000-word passages from texts of the Project Gutenberg and the Literature
Online (LION) collections.
16
In order to provide a comparable text basis, the
16 Most of the Late Modern English texts I had selected are now publicly available in
the CLMET (Corpus of Late Modern English Texts), also modelled on the Helsinki
Corpus in its 70-year sub-periods (see De Smet 2005). In Appendix C.1, I only list
the exact sources of these corpus texts which are not part of the CLMET (these
texts are marked as Project Gutenberg with text number or LION in appendix C.1).
While the CLMET usually gives the full texts, I, as a rule, took the frst ca. 5,000
12 The framework
selection of texts was guided by the principles of the Helsinki Corpus with
respect to the length of the passages and, as far as possible, the text types (for
the problems concerning text type comparability, see Kohnen 2004: 81128
and below). The target number was 100,000 words per period; furthermore, the
respective corpus texts should comprise as many complete texts (e. g. homilies)
by as many different authors as possible.
17
For a frst survey of the sub-periods and the design of the corpus, see the
following table (for details on the texts, see Appendix C.1):
Period Dates Number
of Texts
Words
OE1 850 1 1860
OE2 850950 69 48780
OE3 9501050 13 76350
OE4 10501150 12 34390
ME1 11501250 10 73370
ME2 12501350 3 1440
ME3 13501420 13 100540
ME4 14201500 22 136630
EModE1 15001570 13 97310
EModE2 15701640 16 108770
EModE3 16401710 20 136040
LModE1 17101780 13 102890
LModE2 17801850 18 99840
LModE3 18501920 18 102410
PDE 19201990 49 103900
TOTAL 1,224,520
This table shows that the texts which have come down to us do not allow a de-
tailed quantitative analysis on a large scale (for the periods from 1150 to 1700,
see the detailed account in Kohnen 2004: 81128). Apart from the general
words from these full texts to create a corpus comparable to the texts chosen for the
Helsinki Corpus. Page numbers are thus only given when the passages are not taken
from the beginning of the respective texts.
17 A selection of text as in the BROWN, LOB, F-LOB and FROWN corpora, which
choose 2,000 words from the beginnings of texts, would not have been suitable for
the present study because certain kinds of adverbial connectors, i. e. listing or, in
particular summative ones, only occur at certain passages of the text as a whole (in
the middle or the end, respectively).
The text corpus 13
problems of the lack of documents for the periods OE1, OE4, ME1 and ME2,
which are a hazard to all corpus studies,
18
the present study is confronted with
a more specifc problem concerning conjunctions and adverbial connectors.
First of all, only prose texts could be chosen because poetry employs very
different means of text structuring (see, e. g., Brinton 1996: 6879 on text struc-
turing devices such as ME gan + infnitive in Chaucers Troilus & Criseyde).
More importantly, Present Day English corpus fndings also demonstrate that
adverbial connectors are extremely rare in narrative texts belonging to the
Longman Grammars registers fiction and news (Biber et al. 1999: 882; see
also Chapter 3.5). In narrative texts, authors often choose to leave the relation-
ships unmarked, since relations such as cause and result may be inferred
from a chronological sequence. Authors of narrative fction even avoid being
too explicit about the relations because they want to keep readers in suspense.
Thus the only adverbials which are frequently used in narrative prose are tem-
poral circumstance adverbials (Biber et al. 1999: 822), which in particular in
the categories History, Biography, Travelogue or Diaries of the Hel
sinki Corpus are used as chronological signposts. In the following passage
from the Diaries of Samuel Pepys, for example, the majority of sentences start
with a temporal adverbial:
(2) 1516 DECEMBER 1666. Lords day. Lay long, talking with my wife
in bed. Then up with great content, and to my chamber to set right a
picture or two Lovett having sent me yesterday Santa Claras head
varnished, which is very fne. And now my closet is so full stored and so
fne, as I would never desire to have it better. Dined without any strang-
ers with me which I do not like on Sundays. Then after dinner by wa-
ter to Westminster to see Mrs. Martin, whom I found up in her chamber
and ready to go abroad. I sat there with her and her husband and others
a pretty while; and then away to White-hall and there walked up and
down to the Queens side, and there saw my dear Lady Castlemayne,
who continues admirable methinks and I do not hear but that the King
is the same to her still as ever. Anon to chapel, by the Kings closet,
18 There are almost no texts surviving from the earliest period of English (OE1) and
the late thirteenth century (ME2). The texts collected in OE4 are mainly copies of
texts by authors whom we also fnd in OE3 (lfric in COTEMPO, COAELHOM,
COAELET3, COAEPREF, COAEPREG for OE3 and COAELET4 for OE4; Wul-
fstan with COWULF3 for OE3 and COWULF4 and COINSPOL for OE4). The
frst Middle English period (ME1) has some texts which are copies of Old English
manuscripts and should defnitely be re-allocated to OE3 or OE4 (in particular
CMBODLEY, which is typical of Late Old English).
14 The framework
and heard a very good Anthemne. Then with Lord Brouncker to Sir W.
Coventrys chamber, and there we sat with him and talked. He is weary
of anything to do, he says, in the Navy (CEDIAR3A, p. VII, 409).
There is, to take another example, not a single adverbial connector in the 6070
word passage taken from John Evelyns Diary (CEDIAR3B). With respect to
the text type categorization of the Helsinki Corpus, we are furthermore con-
fronted with the problem that many texts which are labelled as different cat-
egories are in fact predominantly narrative in style. See, for example, the fol-
lowing passage from the Book of Margery Kempe (which is categorized as a
Religious Treatise in the Helsinki Corpus), in which every single sentence
starts with a temporal adverbial (an then, whan when, so long for such a
long time, sithyn then):
19
(3) an sche, hauyng trust of hys a-mendyng & compassyon of hys in-
frmyte, wyth scharp wordys of correpcyon promysyd to fulfllyn hys
entent gyf God wolde grawntyn it. Whan sche cam to hir meditacyon,
not forgetyng e frute of hir wombe, sche askyd forgeuenes of hys synne
& relesyng of e sekenes at owr Lord had gouyn hym gyf it wer hys
plesawns & profte to hys sowle. So longe sche preyid at he was clene
delyueryd of e sekenes and leuyd many gerys aftyr & had a wife &
a childe, blissyd mote God ben, for he weddyd hys wife in Pruce in
Dewchelonde. Whan tydyngys cam to hys modyr fro ouyr e see at
hir sone had weddyd, sche was ryth glad & thankyd God wyth al hir
hert, supposyng & trustyng he xulde leuyn clene & chast as e lawe of
matrimony askith. Sithyn, whan God wolde, hys wife had a childe, a
fayr mayde-child. an he sent tydingys to hys modyr in-to Inglond how
gracyowsly God had visityd hym & hys wife (CMKEMPE p. I, 223).
Similarly, long narrative passages in which primarily the coordinator and
and temporal adverbials are employed are found in many of the private and
offcial letters across all periods; for example, in the following passage from a
letter by Margaret Paston to her husband John:
(4) And Jamys Gloys come with his hatte on his hede betwen bothe his men,
as he was wont of custome to do. And whanne Gloys was a-yenst Wy-
19 Similarly, in her Revelations of Divine Love also categorized as a Religious
Treatise Julian of Norwich virtually only uses the coordinators and, for and
but (CMJULNOR). Only the following adverbial connectors are employed (token
counts): therefore (4), wherefore (1), furthermore (1), nevertheless (1).
The text corpus 15
mondham he seid us, Couere thy heed! and Gloys seid ageyn, So I
shall for the. And whanne Gloys was forther passed by e space of iij
or iiij strede, Wymondham drew owt his dagger and seid, Shalt ow so,
knave? And erwith Gloys turned hym and drewe owt his dagger
And anne Haweys ran into Wymondhams place and feched a spere and
a swerd, and toke his maister his swerd. And with e noise of is a-saut
and affray my modir and I come owt of e chirche from e sakeryng;
and I bad Gloys go in to my moderis place ageyn, and so he dede. And
thanne Wymondham called my moder and me strong hores, and seid e
Pastons and alle her kyn were (1448, 19 MAY; CMPRIV, p.223).
These test analyses for the respective periods show that the paucity of adverbial
connectors found for the Present Day English category fiction is also attested
for the earlier periods of English, mainly in texts labelled History, Bio-
graphy, Lives, Fiction, Travelogue, or Diaries. Apart from a few sample
texts (see above, and Appendix C.1), these were not analysed because they only
use very few, if any, tokens of adverbial connectors.
This means that a large number of the texts which have survived from the
Old and Middle English period (in particular passages from Bible translations,
Chronicles and Saints Lives) had to be excluded. Furthermore, texts which
obviously did not ft their respective categories were excluded from the quan-
titative parts of the present study. Generally, text type consistency and compa-
rability is one of the most problematic issues in diachronic corpus linguistics.
In his study of English participle and gerund constructions from 1100 to 1700,
which had set out to correlate the respective occurrences with text-type pat-
terns, Kohnen shows in a detailed analysis of the texts and their contexts
that only a certain number of texts of the Helsinki Corpus can be employed
for comparable analyses of patterns according to text types (see Kohnen 2004:
81128; as a result, his corpus consists of Chronicles, Religious Treatises,
Homilies, Laws/Documents, Narrative Prose and Private Letters).
The main text type chosen for the present analysis are thus argumentative
texts with their focus of attention on what is commonly called academic or
scientifc language (Biber et al. 1999), i. e. homilies or religious, philosophi-
cal, educational, and literary treatises, some of which also allow more detailed
quantitative analyses (for the principles of selection, see the introduction to
Appendix C.2). This choice of texts causes another well-known problem for
the historical analysis: there are not only few texts of this type extant from ear-
lier periods of English, but most of them are translations from Latin or French.
For the earliest periods of English, there are thus solely some Pre faces
(COPREFCP, COAEPREF, COAEPREG, CMPURVEY, CMCAXPRO) and
16 The framework
some of the Letters (COAELET3, COAELET4, CMPRIV, CMOFFIC3,
CMOFFIC4) which are relatively free from foreign infuence.
20
1.8. Presentation of corpus fndings
The fndings of the study are summarized in Appendices A.1 (Adverbial Con-
nectors: Items) and A.2 (Alphabetical Index of Adverbial Connectors) and
Appendix B (to be found on the accompanying CD-ROM). Appendix A1 pro-
vides a synopsis of all the linguistic items which have been used as adverbial
connectors in the history of English in alphabetical order. The data collected
in this appendix have served as the basis for analysing the developments in the
inventories of adverbial connectors from Old English to Modern English in
Chapters 6 to 12.
Appendix B more specifcally lists the occurrences of adverbial connectors
in the corpus texts. For each semantic category B.1 addition, B.2 summa-
tion, B.3 result/cause, B.4 contrast/concession, B.5 transition the
relevant adverbial connectors are listed alphabetically. The respective entries
for the individual connectors then give the precise text and the sub-periods in
which the items occur in adverbial connector function. An idea about quanti-
ties, i. e. of the number of the occurrences in the respective text, is signalled by
different fonts: underlining indicates that the item occurs more than fve times
per 5,000 words in the text, bold type shows that it occurs more than 20 times.
This presentation of data records the frst occurrence of the respective item
in a corpus text, but, more importantly, allows insights into the use and fre-
quency of the individual adverbial connectors over time as well as in specifc
sub-periods. It furthermore highlights genre-specifc uses of individual con-
nectors. To allow a more detailed analysis of their use, the entries also list
information on attested collocations with coordinators or other adverbial con-
nectors (e. g. and furthermore or but yet nevertheless) and on the position of the
adverbial connector in the sentence; for each item, initial, post-frst, medial and
fnal position are recorded (for the relevance of these positions, see Chapters
5.5 [post-frst-position], 10.5 [fnal position] and 13 [medial position and col-
locations]. Since the study will obviously not discuss the developments of all of
the individual connectors in detail (most of which would merit a monograph),
some illustrative examples will show how this appendix is used in the present
study and how it could be exploited for further research.
20 All corpus fndings have been cross-checked and supplemented by analyses of the
machine-readable corpora, the Dictionary of Old English Corpus (DOEC) and the
Corpus of Middle English Prose and Verse (CME).
Presentation of corpus fndings 17
In the section for additive (reinforcing) adverbial connectors (Appendix
B.1.2), we see in the entry for overmore furthermore that it was only used in
texts of the two later sub-periods of Middle English. It is only found clause-ini-
tially; in one text, it is attested in collocation with the conjunction and, i. e. and
overmore.
overmore Additive (Reinforcing)
ME3 CMDOCU3, CMAELR3
ME4 CMOFFIC4
Collocations: and overmore
ME3 CMAELR3
Eft and eftsona also, furthermore, on the other hand, were only used in the
Old and Early Middle English period.
eft Additive (Reinforcing) also; furthermore
OE2 COLAECE
OE3 COQUADRU, COAELHOM, COBENRUL, COAE LET3,
COAEPREG
OE4 COLAW4, COLACNU, COAELET4
ME1 CMPERIDI, CMBODLEY
ME2 CMAYENBI
Collocations: and eft
OE2 COCURA
ME1 CMPERIDI, CMVESHOM, CMTRINIT,
CMLAMBET, CMBODLEY, CMVICES1
ac eft
OE2 COCURA
eftsona Additive (Equative Reinforcing) moreover, likewise
ME1 CMPERIDI, CMTRINIT
Collocations: and eftsona
ME1 CMPERIDI
Eft and eftsona are also found in collocations with the coordinator and, but
also with ac but. In two texts COLAECE and CMPERIDI they occur
more than twenty times per 5,000 words, as indicated by the bold font. Both
these texts are medical handbooks (the Laeceboc and a Middle English hand-
book entitled Peri Didaxeon) in which sentence-initial eft and eftsona are used
to list alternative medications for specifc illnesses:
(5) Eft sona wi gif eo ylca adle cilde eggelic. on geogee; Nim garluces
And eft sona gif a wunda toinda. Nim fyrs & cnuca hine. & lege
uppa at geswollene (CMPERIDI, p.89).
18 The framework
Furthermore against if the same illness harms a child in its youth.
Take garlic And furthermore, if the wounds swell up. Take bramble
and pound it and put it on the swelling .
This also shows that a large-scale, strictly quantitative comparison is not sen-
sible for the subject in question. The extract from the Laeceboc, for example,
which consists of 10,420 words, only uses six different types of connective
items: the conjunctions and (2 tokens) and ac (4 tokens), the adverbial connec-
tors eac furthermore (5 tokens), onne therefore, then and eah neverthe-
less (1 token each) compared to 6 tokens of and eft and 46 tokens of eft. The
numbers are similar for the Peri Didaxeon (22 tokens of eft sona vs. 13 tokens
of and and 2 of foran therefore).
A similar restriction to genres and uneven distribution is, for example, at-
tested for the loan item furthermore:
item Additive (Reinforcing)
ME3 CMEQUATO
ME4 CMDOCU4, CMREYNES, CMPRIV
EModE1 CELAW1
This reinforcing adverbial connector is as its entry illustrates only attested
in Handbooks, Laws, Statutes and Documents (mainly wills) in the Late
Middle English and Early Modern English sub-periods, often in lists. See, for
instance, the following passage taken from An Act against deceyptfull mak-
ing of wollen cloth:
(6) Fyrst That the Wolle whiche shalbe delyvered for or by the Clothier to
any persone or persones for brekyng kembyng cardyng or spyernyng of
the same the delyvere therof shalbe And that the breker or kember to
delyver agayn to the seid Clothier Item that the Wever whiche shall
have the wevyng of eny wollen yerne to be webbed into cloth shall weve
werk Item that no maner persone bye eny coloured Wolle or coloured
wollen Yerne of eny Carder Spynner or Wever Item that the Walker
and Fuller shall truely walke fulle thikke Item that the Clothier nor
other persone whatsoever he be after the fest of Midsomer next cuerm-
yng shall not put eny cloth to sale Item that the byer of Wollen clothes
denysen or alyen after the byeng therof shall not drawe nor cause to be
drawen in lenght (CELAW1, p. III, 28).
Since texts like these would distort a quantitative investigation based on the
whole of the texts analyzed for this study, only a sub-set of comparative texts
Presentation of corpus fndings 19
was used for the more detailed analysis concerning collocations and word or-
der frequencies in Chapter 13 (for these texts, see Appendix C.2).
Apart from information on genre patterns, the survey in Appendix B also
allows fndings on the popularity of individual adverbial connectors in specifc
sub-periods. Reinforcing also, which has been attested since the Old English
period (OE eall-swa all so) was not found as an adverbial connector in
texts of the Late Modern English period:
also Additive (Reinforcing)
OE3 COBYRHTF, COTEMPO
OE4 COAELET4
ME1 CMTRINIT, CMLAMBET, CMANCR, CMVICES1
ME2 CMAYENBI
ME3 CMDOCU3, CMASTRO, CMHORSES, CMPHLEBO,
CMCTPROS, CMWYCSER, CMAELR, CMPURVEY,
CMOFFIC3
ME4 CMLAW, CMDOCU4, CMREYNES, CMCHAULI,
CMROYAL,
CMINNOCE, CMAELR4, CMKEMPE, CMHILTON,
CMROLLTR, CMPRIV, CMOFFIC4
EModE1 CELAW1, CESCIE1A, CESCIE1B, CEEDUC1A, CEEDUC1B,
CEBOETH1, CEPRIV1, CETRI1
EModE2 CELAW2, CEHAND2B, CESCIE2A, CESCIE2B,
CEEDUC2A,
CETRI2A
EModE3 CESCIE3B, CEBOETH3, CEPRIV3, CEAUTO3
PDE FROWN-D
Interestingly, the OED states that also generally, i. e. also in other functions,
was [n]ot common in 16th c.; Shakes., according to Schmidt, has it only 22
times (OED, s. v. also, Etymology; for a more detailed analysis of this restric-
tion and its causes, see below, Chapter 11).
A further example illustrating the design and, more importantly, the results
from the data analysis provided in Appendix B, is given by the entry for the
concessive/contrastive adverbial connector though, which illustrates the dis-
continuous path of this connective (Appendix B.4).
though Contrastive/Concessive (OE eah/ ME eh)
OE2 COLAECE, COBOETH
OE3 COQUADRU
OE4 CODICTS, COINSPOL
ME1 CMVESHOM, CMBODLEY, CMLAMBET, CMANCR,
CMHALI
ME3 CMASTRO
EModE1 CESCIE3A
20 The framework
eah/eh, second position
OE2 COBOETH, COCURA
OE3 COTEMPO, COWULF3, COBENRUL, COAELET3
OE4 COLAW4, CODICTS, COINSPOL, COPREFSO
ME1 CMSAWLES, CMHALI
though, second position in questions
EModE1 CEHAND2A, CEBOETH2
though, medial position
PDE FROWN-D
This entry shows that though as an adverbial connector was used in a wide
variety of texts in the Old and Early Middle English periods (OE, ME1). It is
attested as an adverb in later Middle and Early Modern English, but only in a
limited number of texts and in very specifc contexts (questions). In the corpus
texts, a use as an adverbial connector is not recorded in any of the texts from
Late Modern English. By contrast, al(though) has been common as a subordi-
nator in all periods of English (see OED, s. v. though, II). This shows that there
is no unidirectional path of though from adverb to conjunction (as grammati-
calization theories would suggest), but rather a zigzag path from ambiguous
adverb/conjunction to conjunction to (a) initially placed conjunction and (b)
medially or fnally placed adverb (see below, Chapter 10.5). This path is neither
evident from the entry in the OED nor from other meta-linguistic texts, but
only emerges as a result of the analysis of the corpus material.
1.9. Outline of the present study
Chapters 2 to 4 prepare the ground for the ensuing analyses by summarizing the
main fndings of previous and contemporary scholarship on clause linkage and
the classifcation of connectives (Chapter 2), and on the problems of the clas-
sifcation and hierarchy of the word class adverb (Chapter 3). In this context,
Chapter 3.5 summarizes those corpus fndings for contemporary English adver-
bial connectors in the Longman Grammar which are relevant for their history.
In an excursus on earlier meta-linguistic thought, Chapter 4 gives an account of
the treatment of conjunctions and adverbial connectors and their status until the
times of Locke and the Scottish Rhetoricians, who, as is mentioned above, are
the frst to explicitly accept adverbial connectors as a separate class.
The main part of the study (Chapters 5 to 13) then provides an in-depth treat-
ment of adverbial connectors in the respective historical periods of English.
For an understanding of the distinctive developments English has undergone
in its history, it is important to lay a foundation by describing the system of
Outline of the present study 21
clausal connection in Old English (Chapter 5). Chapter 6 the heart of the
present study summarizes the general patterns in the diachrony of adver-
bial connectors and contrasts their long-term developments with the changes in
other connectives, i. e. coordinators and subordinators. The divergent histories
of adverbial connectors and subordinators are then examined in more detail in
Chapters 7 to 10. Chapter 7 analyses the morphology of adverbial connectors
(in particular the innovations coined after the loss of pronominal connectors).
Chapter 8 then summarizes the fndings on innovations from the cognitive
source domains time and place/space which have been crucial in the coin-
age of English adverbial connectors from Middle English onwards. In a more
detailed investigation, Chapter 8.4 establishes the source domain truth/fact
for adverbial connectors signalling the relation transition. Chapters 9 to 12
then concentrate on the individual semantic relations: these investigations be-
gin with an analysis of developments in the ccc-relations (cause, contrast/
concession) because they in contrast to the more superfcial changes in the
relations addition and transition (Chapters 1112) also mirror structural
and typological changes. Chapter 13 then resumes the topic of the infuences of
different schools of rhetoric, in particular the impact of the ideas of John Locke
and the Scottish Rhetoricians.
In all of the chapters, overarching principles of long-term developments
in language change will be addressed along the following key questions: Are
there recurrent patterns of change which affect the individual connectors? Are
there different patterns for the different semantic relations? Do the changes in
the system of adverbial connectors correspond to more general cross-linguistic
patterns in language change, such as principles of iconicity or grammaticaliza-
tion clines? Do the changes, more particularly, correspond to general tenden-
cies in the typological history of English and the expansion of the English lexi-
con? And fnally, to what extent are it stylistic or rhetorical preferences which
have directed the paths of change rather than more deeply-rooted typological
properties of English?
2. Clausal connection
2.1. Clausal connection
Clausal connection which can basically be defned as a relation of depen-
dency or sociation between clauses (Lehmann 1988: 182) is one of the key
building blocks of language in use. It is thus a subject which relates not only
to syntactic, but also to a wide range of semantic, pragmatic and, most im-
portantly, textlinguistic phenomena. Traditionally, a frst distinction is drawn
between asyndetic (or unlinked) and syndetic (or linked) clausal connection.
1

Asyndetic connection is not overtly marked (the standard quoted example is
Caesars veni, vidi, vici I came, saw, conquered). It is much rarer than syn-
detic linkage, usually stylistically marked and used, for instance, for dramatic
intensifcation (Quirk et al. 1985: 918; category I in Raible 1992).
Thus, cross-linguistic accounts base their classifcation of clause linkage on
the absence or presence of an explicit linking device; see, for instance, Leh-
manns criterion isolation vs. linkage (Lehmann 1988: 210213; subsection
explicitness of linking). In Raibles scheme of the dimension of linkage as
realisation from aggregation to integration (cf. the title of Raible 1992: Junk
tion Eine Dimension der Sprache und ihre Realisierungsformen zwischen
Aggregation und Integration), explicit linkers are used in categories II to IV
(diagram on the inserted foldout): category II Junktion durch Wiederaufnah-
me (eines Teils) des vorhergehenden Satzes (Linkage by resumption of (a
part of) the preceding sentence), category III Explizit verknpfte Hauptst-
ze (Explicitly linked main clauses) and category IV Verknpfung durch
subordinierende Konjunktionen (Linkage by subordinating conjunctions).
Syndetic connection is thus characterized by the use of overt signals, such
as clause-integrated linkage (Thats why; The result was ) or prepositional
phrases (because of ). Most frequently, however, connectives such as con-
junctions or adverbial connectors are used.
2
1 The following summary rests on Quirk et al. 1984: chapters 1315, Altenberg 1984,
Biber et al 1999: chapters 10 and 11. For overall conceptions of clausal linkage
comprising all kinds of non-explicit linking options, such as non-fnite construc-
tions, the ablative absolute, etc., see Lehmann 1988 and Raible 1992.
2 For corpus fndings for Present Day English, see Altenberg 1984 and 1986 (for
causal and contrastive/concessive connection) and Biber 1999: chapter 10.4 (for all
kinds of linking adverbials). These studies are summarized and discussed below in
Chapters 3.5, 9.2.2 and 10.5.
Connectors connects: defnitions 23
2.2. Connectors connects: defnitions
The most recent, comprehensive study of connectors in Present Day German by
a project group at the Institut fr deutsche Sprache at Mannheim, the 800-page
Handbuch der deutschen Konnektoren, defnes connectors by the following
fve properties (M = Merkmal property; Pasch et al. 2003: 16 and a refned
version on 331334):
M1 x ist nicht fektierbar (x is not infectable)
M2 x vergibt keine Kasusmerkmale an seine syntaktische Umgebung (x
does not govern case)
M3 die Bedeutung von x ist eine zweistellige Relation (x signals a two-place
semantic relation)
M4 die Argumente der Bedeutung von x sind propositionale Strukturen
(the arguments of the meaning of x are propositional structures)
M5 die Relate der Bedeutung von x mssen Satzstrukturen sein knnen
(the elements linked by x may be expressed by sentential structures).
Basically, these characteristics of connectors are generally agreed on in the
literature (see, e. g., also Raible 1992: diagram on inserted foldout). Instead
of classifcations focussing on Present Day English (such as Quirk et al. 1985:
chapter 13 or Biber et al. 1999: chapters 10 and 11), these specifc criteria were
chosen as the basis of the present investigation because the diachronic account
requires a system which is transferable to earlier periods of the English lan-
guage. Since Old English is typologically more similar to Modern German
than to Modern English, the criteria of the Handbuch der deutschen Konnek
toren provide a solid basis for comparison.
In the following section, I will briefy introduce those points of differentia-
tion from other parts of speech such as prepositions, disjuncts or modal par-
ticles which are essential for an understanding of the diachrony of English
connectors.
M1 is a morphological criterion which groups connectors together with other
indeclinable items such as prepositions.
3
Connectors may, of course, have their
origin in case forms, such as in a genitival -es, which is attested in PDE else (< OE
elles) or by way of analogy in besides (OE be sidan) and hence (OE heonan);
3 In the Greek and Latin tradition, all indeclinable items were called particles; see
the quote from Locke above, p. 1, and below, Chapter 3.5.
24 Clausal connection
these forms, however, were petrifed and are now indeclinable in their uses as
connectors. In earlier stages of English, we regularly also fnd pronominal con-
nectors, similar to PDG deshalb or demnach therefore. In Old English, these
consist of, for instance, a preposition and the case form (mainly dative or instru-
mental) of a demonstrative, such as OE form (< preposition for + dative m)
or OE foron and fory (< preposition for + instrumental on or y). Analyzing
these forms, it is important to differentiate full prepositional phrases, such as
PDE because of this, from genuine connectors, such as PDE because or therefore.
For this differentiation, M2 is formulated as the syntactic criterion distin-
guishing prepositions which govern a particular case from connectors,
which do not govern case. This criterion is, however, only rarely applicable
to English after the Old English period, when English had given up case (and
gender) distinctions in articles and demonstratives.
M3 is a semantic property which differentiates connectors from other sen-
tence adverbials, such as the epistemic probably, perhaps or stance adverbials
such as frankly. These also have a scope extending over the whole sentence,
but they modify this sentence only and unlike connectors do not relate two
clauses or even larger portions of text (for details on different classes of ad-
verbs, see below, Chapter 3). They do thus not qualify as organizers of textual
structure. Connectors have to signal a two-place relation.
M4 and M5 are properties augmenting M3. They contrast connectors, again,
from prepositions and also from coordinators on the phrase level. Prepositions
may also be seen to relate two items, such as tree and house in a sentence like
PDE the tree in front of the house, but they do not relate propositions or clauses.
M5 in particular distinguishes prototypical connectors from coordinators,
such as and and or. Coordinators are not always connectors in the sense of the
present study because they may also combine elements on the phrase level, as
in Peter and / or Mary (were sitting in front of the house). There are, however,
also classifcations, which propose a gradient of connectors, ranging from co-
ordinators to subordinators. The Comprehensive Grammar (Quirk et al. 1985:
923-924) for example, explicitly states that coordinators can link clause con-
stituents; such a gradient, however, is not suitable for the present account,
which only deals with connection on the clausal or textual level.
4
Whenever
the present study wants to stress this fact of a two-place relation on the level
above the phrase, the clauses or parts of text connected will be referred to as
4 According to M5 of the Handbuch, the sentences connected have to comprise
a fnite verb or have at least to be able to comprise a fnite verb. The restriction to
clauses containing a fnite verb is only applicable to earlier periods of English, but
not to Present Day English, which allows non-fnite clauses.
Connectors 25
connect(s), a term (PDG Konnekt) coined for the Handbuch der deutschen
Konnektoren (Pasch et al. 2003).
2.3. Connectors: coordinating conjunctions, subordinating
conjunctions, adverbial connectors
By these fve criteria we can establish three different kinds of connectors, a dif-
ferentiation which is also basically agreed on in the literature: coordinating con-
junctions (working on the clausal/textual level, not on the phrasal level; see M4
and M5), subordinating conjunctions (subordinators), and adverbial con-
nectors (Quirk et al. 1985 conjuncts, Biber et al. 1999 linking adverbials).
The present study focuses on adverbial connectors. As the system of ad-
verbial connection has changed so dramatically since the Old English period,
it is, however, necessary to summarize the main syntactic and text-linguistic
properties of the other types here in some detail. Virtually all of todays ad-
verbial connectors are new coinages or have undergone some morphological
and structural change. One of the natural paths was the transfer from one of
the other classes of connectors: compare, for example, the use of the Present
Day English subordinators albeit and howbeit as adverbial connectors in the
Early Modern English period. Since Old English is typologically much closer
to Modern German than to Modern English, the following section will exem-
plify the relevant issues not only by material taken from Present Day English
(contrastive/concessive connectors; Table 2.1) but also by Present Day German
examples (causal connectors; Table 2.2).
55
Table 21: Contrastive/concessive connectors in Present Day English
Parataxis
Coordinating
Conjunction:

(7) He tried hard, but he failed

[post-position]
Adverbial Connector: (8) He tried hard, and yet he failed [post-position;
and + AdvCo]
Adverbial Connector: (9a) He tried hard However, he failed. [initial]
(9b) He tried hard He, however, failed [medial]
(9c) He tried hard; he failed, however [fnal]
Hypotaxis
Subordinating
Conjunction:
(10a) (Al)though he tried hard, he failed
(10b) He failed, though he tried hard
[pre-posed] or
[post-posed]
5 The contrastive/concessive and causal semantic relations are chosen because they
will be in the focus of the detailed diachronic studies in Chapters 10 and 9 below.
26 Clausal connection
Table 22: Causal connectors in Present Day German
Parataxis
Adverbial Connector: (11a) Sie wird gewinnen Sie ist nmlich
strker
[V2; post-posed]
(11b) Sie ist strker Deswegen wird sie
gewinnen
[V2; post-posed]
Hypotaxis
Subordinating
Conjunction:
(12a) Weil sie strker ist, wird sie gewinnen [V-fnal;
pre-posed] or
(12b) Sie wird gewinnen, weil sie strker ist [post-posed]
Correlatives
(13) Sie wird deswegen gewinnen, weil sie
strker ist
[V2 + V-fnal]
The criteria traditionally employed for distinguishing the different types of
connectors are
[1] the position of the connector,
[2] the sequence and position of the respective connects,
[3] the possibility of collocations of connectors, and
[4] in languages such as German, constituent order.
I will here frst of all summarize these traditional views. It has to be stressed,
however, that the constructions are by no means functionally equivalent, but
are different with respect to their coding of the information structure (for de-
tails, see below, Chapter 2.4).
6
[1] While coordinating and subordinating conjunctions are only found
clause-initially (see English but, and and German weil in examples 7, 8, 12),
adverbial connectors (English however, yet; German nmlich, deswegen) are,
like most kinds of adverbs, more free in their position in the sentence (see ex-
amples 9 for however).
7
Pasch et al. (2003) capture this difference by the terms
6 It is the major drawback of the novel classifcation and terminology coined for the
Handbuch der deutschen Konnektoren that connectors are distinguished only by
their position in the clause or connect, i. e. by topological criteria only. The Hand
buch thus differentiates konnektintegrierbare vs. nicht-konnektintegrierbare Kon-
nektoren and further Subjunktoren Postponierer Verbzweitsatz-Einbetter
Konjunktoren.
7 Pasch et al 2003 distinguish the following positions for adverbial connectors:
nicht positionsbeschrnkte Adverbkonnektoren such as allerdings, dagegen, frei
Connectors 27
nicht-konnektintegrierbare (initial position) vs. konnektintegrierbare con-
nectors.
8
[2] Parataxis usually requires a fxed order of connects. Accordingly, claus-
es linked by coordinating conjunctions and adverbial connectors only allow
one sequence of connects. Thus there are, for example, two different German
adverbial connectors for the causal/resultive relation, namely nmlich for the
sequence result cause, and deswegen for the sequence cause result,
which accordingly require reverse sequences of the respective connects. The
interesting fact about Present Day English clausal connection is that it lacks an
adverbial connector coding the relation result cause (cf. Latin nam, enim;
PDG denn, nmlich); it only has adverbial connectors expressing the sequence
cause result, such as therefore, hence, thus (cf. Latin igitur, PDG deshalb;
for a full analysis, see below Chapter 9).
Subordinate clauses, on the other hand, may be placed before or after the
superordinate clause (see English although or German weil in examples 10 and
12). In contrast to parataxis, where the units are constituents at the same level
of constituent structure, they form a hierarchy in subordination (or hypotaxis),
the subordinate unit being a constituent of the superordinate unit (Quirk et
al. 1985: 13.2). Since the present study concentrates on adverbial connectors,
it does not consider all kinds of subordinate structures and excludes cases of
embedding (e. g. obligatory nominal sentences functioning as subjects, objects,
subject/object complements) or relative clauses, which post-modify a noun
phrase only. Neither embedding nor relative clauses connect propositions (thus
failing properties M4 und M5).
9
lich, also, nicht nacherstfhige Adverbkonnektoren such as demnach, daraufhin,
trotzdem, deshalb, stark positionsbeschrnkte Adverbkonnektoren such as aber,
auch, nur and nicht vorfeldfhige Adverbkonnektoren such as nmlich. For the
relevance of the distinction of the post-frst-position from other medial positions,
see below, Chapter 13.
8 These compounds are so typically German that they are virtually untranslatable
into other languages.
9 Obviously, there are some polyfunctional elements, such as the Old English general
subordinator e, which may function as a relative particle but can also be part of a
complex connector (cf. OE for m e because; therefore). Similarly, the general
subordinator in Middle English, that, can introduce nominal clauses, but may also
be part of a complex relativizer (cf. ME which that) or of a lexicalized adverbial
connector (cf. Middle English causal for that or additive over that).
28 Clausal connection
[3] Coordinating conjunctions may in contrast to the other types collocate
with subordinators and also with adverbial connectors (see English and yet in
example 8).
10
[4] In contrast to Present Day English, which employs a fxed subject verb
word order in main and subordinate sentences, Present Day German distin-
guishes these sentence types by employing verb-second word order for main
clauses and verb-fnal for subordinate clauses (see the underlined verbs in ex-
amples 12 and 13; for a discussion of word order patterns in main and subordi-
nate clauses in Old English, see below, Chapters 5 and 11).
The morphological make-up of German connectors also allows more easi-
ly for so-called correlative constructions, which mark the relation of the
sentences with an adverbial connector in the frst connect and a subordinating
conjunction in the second one (cf. German deswegen weil in example 13 and
the much rarer English because therefore).
2.4. Connectors and information processing
The traditional accounts of connectors summarized above seem to imply that
these constructions are functionally equivalent. This is mainly due to the fact
that most grammars concentrate on the make-up of a single sentence (which
may also be a compound sentence). Complex sentences, however, and larger
chunks of discourse are, if at all, only treated marginally.
Starting with Thompsons analysis of initial versus fnal purpose clauses
(Thompson 1985), however, there has been an increasing and continuing inter-
est in the discourse factors determining the position of subordinate clauses (see
the literature from Haiman and Thompson 1988 to Diessel 2005). Thompson
found that there is no free choice between these two positions, so that initial
and fnal purpose clauses should not be regarded as a single construction po-
tentially occupying two different positions. More appropriately, they should be
viewed as two quite different constructions, which share the same morphology,
but behave radically different in the organization of discourse.
Initial clauses state a problem within the context of expectations raised by
the preceding discourse, to which the following material (often many clauses)
provides a solution.
10 For the popularity of these collocations in different phases of the history of English,
see below, Chapter 13.
Connectors and information processing 29
(14) To cool, place the loaf on a wire rack.
Final purpose clauses, on the other hand, play a much more local role of stating
the purpose for the action named in the immediately preceding clause.
(15) Place the loaf on a wire rack to cool.
As in the ordering of other constituents, the ordering of the linguistic ele-
ments is here seen to be primarily determined by information structure: given
information tends to precede new information because new information needs
to be grounded in information that is already in the hearers knowledge store.
This topic-forming capacity of pre-posed adverbial clauses is now common-
ly agreed on (see, for example, Givn 2001: chapter 18 and Meurman-Solin
2004).
In a more recent study, Diessel argues that both processing and information
structure are relevant to constituent order. He shows that the positional pat-
terns of conditional, temporal and causal clauses in Present Day English are
motivated by competing functional and cognitive forces, namely economy and
explicitness (Diessel 2005: 449). In an analysis based on Hawkins processing
theory, he shows that, with regard to parsing or utterance planning, complex
sentences are easier to process, and thus highly preferred, if the adverbial
clause follows the main clause. The discourse pragmatic motivations sketched
above, however, which favor initial occurrence, may override the processing
motivation for fnal occurrence: the clauses are positioned sentence-initially if
they provide a thematic ground or orientation for subsequent clauses.
Since the discourse pragmatic function of adverbial clauses interacts with
their meaning, different semantic types of adverbial clauses differ in their
distribution: Conditional clauses tend to occur sentence-initially, because they
establish a specifc framework for interpreting the following connect. Tem-
poral clauses tend to precede the main clause if they denote a situation prior
to the one in the main clause so that initial occurrence results in an iconic
order. Causal clauses mostly follow the main clause. According to Diessel,
only scientifc articles exhibit a substantial number of pre-posed causal claus-
es because in this type of discourse causal clauses are often used to provide a
common ground for a subsequent conclusion (Diessel 2005: 465).
11
Although the present study is not primarily concerned with hypotaxis,
these issues of processing and information structure are also crucial for an
11 For similar corpus fndings, see Altenberg (1984 causal clauses) and Quirk et al
(1985: 1107 causal clauses).
30 Clausal connection
understanding of adverbial connectors, and in particular their history.
12
The
functional relation between clause combining by conjunction vs. adverbial
connectors is not as widely researched as the criteria for the different posi-
tions of subordinate clauses. Generally, Quirk et al. (1985: 13.3) see a major
difference between coordination and subordination of clauses: the information
in a subordinate clause is often placed in the background with respect to the
superordinate clause. In this view, the syntactic inequality of subordination
tends to bring with it a semantic inequality which is realized by syntactic hier-
archization. Thus the information given in these clauses is usually not pursued
in the following discourse. As examples they give
(16) He has quarrelled with the chairman and resigned.
(17) Since he has quarrelled with the chairman, he resigned.
(18) He resigned because he has quarrelled with the chairman.
In this cause result relationship, the hearer is assumed to have heard about
the quarrel already in (17) (for this topic-forming capacity, see above). The in-
formation given in the subordinate clauses, however, will not be pursued in the
following discourse, irrelevant of whether it is pre-posed (17) or post-posed (18).
An analysis by Wegener (2000) for German, which like Old English
differentiates verb-second and verb-fnal word order, yields similar results,
which are supported by syntactic and prosodic criteria. In German and Eng-
lish, causal and concessive relations may be coded by either coordination or
subordination. While concessive subordinators mainly code the concessive
relation proper (counter-expectancy), adverbial connectors more often code
contrast (adversativity without any counter-expectancy; for details, see
below, Chapter 10). For conditionals, however, we never fnd parataxis nei-
ther in Present Day German nor in any periods of English.
Table 23: Semantic relations: parataxis and hypotaxis
Semantic Relation Parataxis Hypotaxis
cause PDG denn, (weil + V2)
PDE therefore
PDG da, weil (+ Verb-fnal), zumal da
PDE since, as, because
12 In pre-posed causal clauses, for instance, polyfunctional form, which may also
function as an adverbial connector, is frst replaced by unambiguous subordinators
for as moche as or since that. For a detailed analysis, see below, Chapter 9.
Connectors and information processing 31
Semantic Relation Parataxis Hypotaxis
concession /
contrast
PDG aber, zwar aber
PDE however, though,
still
PDG obwohl, obgleich, auch wenn
PDE (al)though (yet)
condition PDG wenn (dann), falls, sofern
PDE if (then)
The importance of textlinguistic and pragmatic features is nicely illustrated
by the functions of German weil with main clause word order (verb-second),
such as
(19) Er ist nach Hause gefahren weil er hatte Kopfweh.
He went home for he had a headache.
instead of verb-fnal, as in
(20) Er ist nicht nach Hause gefahren, weil er Kopfweh hatte, (sondern weil
er Besuch erwartet).
He didnt go home because he had a headache (but because he is ex-
pecting visitors)
The fairly new use of weil with main clause word order (mainly in spoken Ger-
man) is certainly not as bad as some purist language users think (see also
Keller 1993; or for that matter the Microsoft Word (version 2003) grammar
check, which marks it as wrong in example 19). It is a common though still
informal structural device in German, in which the lexical item weil replaces
the adverbial connectors denn or nmlich. Weil is not used in a wrong way, but
has acquired a second function, different to its use as a subordinator. Since this is
the path which depicts the history of many of todays English adverbial connec-
tors, it is important to summarize the main points of difference and change here.
The examples frst of all show that (19) provides two information entities,
whereas the hypotaxis in (20) gives only one focus; see the focus of the negator
nicht and the possible questions for (19) Ist Peter noch da? Is Peter still here?
and Warum sagst du das? Why do you say that?
13
in contrast to the single
Warum ist Peter nach Hause gefahren? Why did Peter go home? for (20).
13 Weil with verb-second order does not give a reason for the proposition of the sen-
tence but a reason for the speech-act of the frst connect. Weil therefore does not
relate to the proposition of the frst clause but to the speech act as a whole. For these
32 Clausal connection
As illocutionary signals, German employs syntactic and prosodic means,
i. e. different word orders (V2 vs. verb-fnal) and, for parataxis, an intonation
pattern which requires a pause, indicating that the second connect is not inte-
grated in the frst one. In subordination, the subordinate clause is syntactically
and prosodically integrated: the whole complex sentence only has one intona-
tion contour and one thematic (theme rheme) structure (see Wegener 2000:
36 and Givn 2001: 327 for English). In parataxis, the second connect also has
more assertive power, a power which may be highlighted by means of strongly
deictic pronominal adverbial connectors such as PDG deshalb or dementspre
chend or transparent lexical adverbial connectors such as consequently. The
greater illocutionary weight of the denn- or weil-sentence indicated syntac-
tically by V2 and prosodically by pause and main stress also on the second
connect gives this second connect an illocutionary force of its own. This
explains why we do not fnd both options of conditional relations there the
two connects cannot carry independent illocutionary forces.
This summary shows that the path from subordinator to adverbial connec-
tor (as attested for however in Early Modern English and though in Present
Day English) or the other way round (albeit in Early Modern English) is by no
means a superfcial transition. These changes are, conversely, induced by fac-
tors of information structure and may, on the other hand, also lead to changes
in information structure. In this view, adverbial connectors are in contrast
to subordinators very strong indicators of a great illocutionary weight of the
second connect. Its proposition may then be pursued in the following discourse.
uses, see Keller (1993), Wegener (2000) and below, chapter 10.5, on PDG obwohl
with verb-second order.
3. The category adverb
Adverb is not simply a lexical category, nor a function-
al one: rather, it emerges from the coalescence of both
viewpoints (Ramat and Ricca 1998: 188).
3.1. Adverbs and adverbials
Although adverbs are among the oldest categories in grammatical thought
(see below, Chapter 4), this category of words is most puzzling. The diver-
sity of items collected under the heading adverb even seems to evoke the
impression that if a linguistic item cannot be integrated into any of the other
word classes, it is called an adverb: the category adverb is the mixed
bag among the word classes. In the title of one of his articles, Guimier (1991)
even asks whether it is at all possible to defne the term adverb Peut-on
dfnir ladverbe?.
First of all, it is necessary to distinguish between the formal category ad-
verb as a word class and the functional category adverbial, which may be
realized by a wide range of syntactic forms in Present Day English: adverbs,
prepositional phrases, fnite and non-fnite clauses, and noun phrases. In con-
trast to other functional categories such as subject or direct object, multiple
adverbials can occur in a clause or sentence and they can be placed in a variety
of positions, sentence-initially, -medially or -fnally. This syntactic variability
is due to the fact that they perform a variety of functions and fulfl a variety of
semantic roles, such as location, time, agency, attitude or, most important
for the present study, a wide variety of linking functions. Most adverbials are,
however, optional.
An analysis of adverbs thus has to consider formal and at the same time
functional aspects. This position at the interface between the lexicon and
grammar has led to a long tradition of discussion whether the Present Day
English adverbial suffx -ly is a lexical (derivational) or grammatical (infec-
tional) suffx (see Nevalainen 1997; van der Auwera 1998: 3). More impor-
tant for the present study, there have also been various views on whether the
coinage of new adverbs by univerbation involving fusion and coalescence is
an instance of lexicalization or grammaticalization (see Brinton and Traugott
2005: 132136, 152).
Adverbs are like conjunctions, prepositions and other smaller categories
such as interjections indeclinable and are therefore not to be distinguished
34 The category adverb
from other indeclinable items on purely formal grounds.
1
For their cross-lin-
guistic account of sentence adverbs in the languages of Europe, Ramat and
Ricca (1998: 187) provide a very clear and workable frst defnition of adverbs
which takes both the formal and the functional perspective into account. Since
it also allows an integration of data from earlier periods of English, it will here
be adopted as a basic defnition:
(i) Formally, adverbs are invariable and syntactically dispensable lexemes (which
may have derivational status, e. g., Lat. simil-i-s similar, simil-iter similarly, or
even originate from infectional status: Lat. merito(d) rightly).
(ii) Functionally, adverbs are modifers of predicates, other modifers or higher syn-
tactic units. In other words they add information to other linguistic elements which
can stand on their own, semantically as well as syntactically.
This defnition frst of all highlights that the term adverb itself is mislead-
ing. The functional property of adverbs as being a modifer is shared by ad-
jectives and adverbs. Yet, while adjectives are essentially noun modifers, ad-
verbs basically modify non-nominal constituents. An adverb thus need not
necessarily modify a verb, but can also modify an adjective (22), or another
adverb (23):
(21) He <knocked> loudly at the door.
(22) He wrote an extremely <interesting> book.
(23) He started smoking very <heavily>.
In addition to these functions on the phrase level, adverbs may also have a
wider scope which extends over the whole sentence. These adverbs are accord-
ingly labelled sentence or sentential adverbials. Sentence adverbials may be
defned as a class of syntactically dispensable lexemes which affect/modify in
various ways the content of a sentence in which they occur (Ramat and Ricca
1998: 189), as in
(24) <You are> probably <right>.
(25) Fortunately, <no one complained>.
1 For recent accounts of the word class or category adverb and its problems, see
Bellert (1977), Biber et al. (1999: chapter 10), Greenbaum (1969a), Guimier (1991),
Quirk et al. (1985: 7.468.153; 438653), and, in particular, the chapter in the
EUROTYPE volume by Ramat and Ricca (1998).
Classifcation of adverbials 35
Their scope may then also extend further to the level of the paragraph or text,
such as in
(26) <This applies to men>. Likewise, <women >.
3.2. Classifcation of adverbials
To differentiate this mixed bag of adverbials, grammars usually suggest
a tri-partite classifcation based on syntactic and semantic criteria. Thus the
Longman Grammar distinguishes circumstance adverbials from stance adver-
bials and linking adverbials (Biber et al. 1999: 762892):
Circumstance adverbials add circumstantial information about the prop-
osition in the clause, or add information about the circumstances of an
activity or state described in the clause.
Stance adverbials express the speakers/writers stance towards the
clause, and either give the speakers comment on the proposition of the
clause or convey the speakers comments on what they are saying (the con-
tent of the message) or how they are saying it (the style).
Linking adverbials link the clause to some other unit of discourse and
thus serve connective functions.
These differences leading to a three-partite classifcation are widely agreed on
in contemporary grammars.
2
One of the most infuential monographs on Eng-
lish adverbs by Greenbaum (1969a) basically distinguishes between the same
three major classes as the Longman Grammar (Biber et al. 1999), but calls
them adjuncts, disjuncts and conjuncts. This classifcation is in essence
3

2 The Cambridge Grammar of the English Language, however, now suggests another
system which mainly affects the boundary between adverb and preposition (Huddle-
ston and Pullum 2002: 264). In this view, prepositions may also take other kinds
of complements than noun phrases, so that their preposition category also includes
some words without a complement. Thus outside in The basket is outside is classi-
fed as a preposition. This new categorization, however, conficts with the properties
of connectors chosen for the present study (see above Chapter 2.2). For a cross-lin-
guistic approach, see the synopsis of classifcations in Ramat and Ricca (1998).
3 The 1972 edition corresponds fully to the one described here. The 1985 edition
distinguished a fourth category, namely subjuncts (examples: We havent yet fn
ished. Would you kindly wait for me? Quirk et al. 1985: 8.888.120, pp.566612).
This fourth category is not adopted here because the whole category of subjunct
has been repeatedly criticised (for a recent assessment, see Valera 1998: 267270).
36 The category adverb
taken over by the leading grammar of English, the Comprehensive Grammar
of the English Language (Quirk et al. 1985), one of whose authors was Green-
baum:
Table 31: Classifcation of adverbials in the Comprehensive Grammar (Quirk et al.
1985) and the Longman Grammar (Biber et al. 1999)
Quirk et al. 1985 Biber et al. 1999
scope: phrase adjunct circumstance adverbial
scope: sentence content/style disjunct stance adverbial
scope: sentence/text conjunct linking adverbial
3.2.1. Circumstance adverbials or adjuncts
Circumstance adverbials are the ones which are best integrated into the clause
structure. They are also the most varied class and answer questions such as
how, when, where, how much, to what extent?:
(27) Slowly, they <walked> back home.
(28) He <spoke> to me about it briefy.
3.2.2. Stance adverbials or disjuncts (content/attitudinal and style
disjuncts)
Stance adverbials (Quirks content or attitudinal disjuncts) are epistemic in
that they focus on the truth value of the proposition, commenting on such fac-
tors as certainty, reality, sources, limitations or precision of the proposition.
They have a scope over the entire clause and are always optional. As content
disjuncts, they comment on the content of the proposition:
(29) <She> wisely <didnt attempt to apologize>.
Paraphrase: (I, the speaker, think that) it was wise of her that she didnt
attempt to apologise.
Also, subjuncts are of no signifcance to the present study, because they work like
adjuncts on the level of the phrase, and not on the level of the sentence or above.
Classifcation of adverbials 37
(30) The Yards wonder boy, appropriately, descended from the clouds.
Paraphrase: (I, the speaker, think that) it was appropriate that the
Yards wonder boy descended from the clouds.
The other kind of stance adverbials (Quirks style disjuncts) convey a speak-
ers comment on the style or form of the utterance, often clarifying how the
speaker is speaking or how the utterance should be understood:
(31) Frankly, <I am tired>.
Paraphrases: (a) Frankly speaking, Im tired.
(b) Im frank when I say that Im tired.
3.2.3. Adverbial connectors (linking adverbials, conjuncts)
The adverbs which are in the centre of the present study are termed link-
ing adverbials by Biber et al. (1999) and conjuncts, a term which explicitly
stresses their proximity to conjunctions, by Greenbaum (1969a) and the Com
prehensive Grammar (Quirk et al. 1985). In the present study, these items will
generally be called adverbial connectors. The more specifc terms will only be
used when necessary for contrast with other adverbial types.
Linking adverbials are similar to stance adverbials in two respects: they are
optional and their scope at their lowest level is the sentence. They also add
the voice of the speaker to the proposition. In their case, however, the voice of
the speaker does not affect the propositional content of the respective proposi-
tion, but reveals the structure of the text. Rather than adding information about
the hows and whens of the proposition (circumstance adverbials) or about the
truth value of the proposition as far as the speaker is concerned (stance adver-
bials), they serve a connective function making the relationship between two
units of discourse explicit.
Linking adverbials have a peripheral relationship with the rest of the clause,
and may extend the level of the sentence, so that they can connect units of dis-
course of different sizes, from sentences to large chunks of discourse. Much
more than with the other types of adverbials, this linking function entails a
conjunct-specifc set of semantic relations, such as addition and enumera-
tion, summation, result/inference/cause or, as in the following examples,
contrast/concession:
(32) All our friends are going to Paris this summer. We, however, are going
to London.
38 The category adverb
(33) He didnt invite her. She wouldnt have come, anyway.
As signals of two-place relations, they deal with various inter-sentential rela-
tionships and are not involved in the semantic structure of a single sentence.
That they do not work on the representational, but on the interpersonal or
textual levels is evident from a number of syntactic restrictions (see Quirk et
al.1985: 631634 and the full account in Greenbaum 1969a: 4144):
A linking adverbial cannot be the focus of a cleft sentence (34a), it cannot be
the basis of contrast in alternative interrogation or negation (34b), and it cannot
be focused by a focalizer such as only (34c):
(34) She may be unable to attend the meeting. You should nonetheless send
her the agenda.
(34a) *It is nonetheless that you should send her the agenda.
(34b) *Should you send her the agenda nonetheless or therefore?
(34c) *You should only <nonetheless> send her the agenda.
In agreement with their peripheral role in the sentence, they may similar to
stance adverbials be paraphrased by a separate clause with a verb of com-
munication plus a circumstance adverbial:
(35) One can say + circumstance adverbial then, now (time)
I tell you + circumstance adverbial here, further (space)
I will say + circumstance adverbial consequently, thus (result)
These paraphrases also show that there is a natural path from circumstance to
linking adverbials. Indeed, many of the new adverbial connectors which have
developed in the history of English started their life as circumstance adverbials
on the phrasal level, modifying a verbal phrase containing a verbum dicendi.
In the following passage from lfrics De Temporibus Anni (COTEMPO), for
example, furor further and nu now, which may also function as linking
adverbials when on their own, are used as circumstance adverbials modifying
the verbs sprecan speak and secgan say.
(36) We willa furor ymbe as emnihte swior sprecan on gedafenlicere
stowe; & we secga nu sceortlice t se forma dg yssere worulde is
(COTEMPO, p.19).
We will further and more correctly speak about the equinox in the
suitable place; and we say now briefy that the frst day of this world is
.
Semantic categories of adverbial connectors 39
This use of circumstance adverbials with a verbum dicendi is widely attested
for all periods of English:
(37) Also it is said that (Theriaca Athanasia) doe both resolue, breake and
digest humours, being compact and gathered together in the profun-
dity of the body. (Mercurialis) saith moreouer, that (CESCIE2A,
p.1112).
These contexts of verb of communication plus circumstance adverb of time
and space, etc. are prototypical bridging contexts, which allow the re-inter-
pretation and use of circumstance adverbials as linking adverbials, most often
by metaphorical extension. The source domains of circumstance adverbials
which are used as linking adverbials are predominantly time (cf. causal then),
space (cf. causal hence) and truth/fact (OE solice, ME forsooth; EModE
indeed, PDE in fact; for a detailed account, see below, Chapter 8).
3.3. Semantic categories of adverbial connectors
As has been mentioned above, the linking function of adverbial connectors
entails, much more than with the other types of adverbials, a conjunct(ion)-spe-
cifc set of semantic relations (see Quirk et al. 1985: 634640; Biber et al.
1999: 875879). Although the grammarians choose different labels, the cat-
egories distinguished are basically the same for different languages and have
also changed only slightly since the frst detailed account by Campbell ([1776]
1963: 404). Campbell distinguishes the following relations:
Table 32: Semantic relations marked by adverbial connectors according to
Campbell ([1776] 1963: 404)
copulative (when facts are related in continuation, or when one argument, remark,
or illustration, is with the same view produced after another)
And, now, also, too, likewise, again, besides, further, moreover, yea, nay, nor
disjunctive
a) adversative (if the sentiment in the second sentence is in any way opposed
to that which immediately precedes)
But, or, however, whereas
b) exceptive (if it is produced as an exception): yet, nevertheless
causal (if the latter sentence includes the reason of what had been affrmed in the
frst): for
illative (if it contain an inference): therefore, then
40 The category adverb
Similarly, the Comprehensive Grammar (Quirk et al. 1985: 634) distinguishes
six conjunctive roles,
4
in some cases with subdivisions. This system is adopt-
ed for the present study, so that Appendix B providing the corpus fndings
for the diachrony of adverbial connectors is structured according to these
sub-categories of semantic relations. The following survey exemplifes the se-
mantic relations by Present Day English conjuncts (as realized not only by
adverbs, but also by prepositional phrases or clauses):
Table 33: Semantic relations of adverbial connectors
Enumeration and addition (see Appendix B.1)
enumeration: frstly, secondly, thirdly (etc); to begin with
addition:
equative correspondingly, equally, likewise, similarly
reinforcing again, also, further, furthermore, moreover, in particular,
then, too, (above all)
Summation (see Appendix B.2)
in sum, to conclude, all in all, in conclusion, overall, to sum
marize
Result/Inference/Cause (see Appendix B.3)
therefore, consequently, thus, so; hence, in consequence
Contrast/Concession (see Appendix B.4)
on the other hand, in contrast, alternatively; though, anyway,
however
focus on contrast: conversely, instead, on the contrary, in
contrast
focus on concession: anyhow, besides, nevertheless, still, in
any case
Transition (see Appendix B.5)
now, meanwhile; incidentally, by the way
The only category which is not analysed in detail in the present study is appo-
sition (in other words, i e, that is, for instance; namely, specifcally), since
these forms mainly work on the constituent or phrase level. They only rarely
mark a two-place relation on the sentential or discourse level (for details, see
below, Chapter 11).
Synchronically, all semantic relations apart from enumeration/addition
form a closed class, and thus have a fxed number of adverbial connectors cod-
ing the respective category. New items cannot be coined instantaneously but
4 Almost identically, the Longman Grammar distinguishes the semantic categories
(1) Enumeration and Addition, (2) Summation, (3) Apposition, (4) Result/Inference,
(5) Contrast/Concession and (6) Transition (Biber et al. 1999: 875879).
Semantic categories of adverbial connectors 41
have to undergo the various common steps in language change in the case
of prepositional phrases, e. g., fusion, coalescence to become full, i. e. coded
members of the group.
5
The only exception is the category enumeration and
addition, which forms an open class, i. e. new items can be coined at all times
as nonce-formations from all kinds of lexical material (see Appendix B.1 for
forms such as, e. g., fourthly, sixthly or frst, frstly, in the frst place, frst of all).
Since they form an open class, they are not listed in Appendix A and are not
analysed in the quantitative sections of this study.
3.4. Pure and impure connectives
The Cambridge Grammar proceeds one step further and draws a distinction
between pure and impure connectives (Huddleston and Pullum 2002: 775).
In this view, pure connectives like moreover and also have no other function
than that of connecting their clause to the surrounding text, while impure con-
nectives such as nevertheless, then or consequently combine this connecting
function with a function of concession, condition or reason/result. This
distinction is also mirrored in the changes affecting the system of pure and
impure connectives over time. Changes in the ccc-relations (cause, conces-
sion/contrast) are indicative of more general patterns in language structure
and long-term developments than are the changes of the mono-dimensional ad-
verbial connectors expressing addition or transition. These do not primari-
ly testify to structural changes, but to changes induced by attempts at more
expressivity or novelty.
6
5 In the analyses, I follow the Invited Inferencing Theory of Semantic Change
(Traugott and Dasher 2002) to account for conventionalizing of pragmatic mean-
ings, i. e. implicatures or invited inferences, and their reanalysis as semantic mean-
ings (cf. Levinson 2000). Following Levinson, this dynamic theory of language
change builds on the distinction of three levels of meaning: coded meanings (con-
vention of a language at a given time), utterance-type meanings (generalized in-
vited inferences; GIINs) and utterance-token meanings (invited inferences which
have yet not been crystallized into commonly used implicatures; IINs). Histori-
cally, there is a path from coded meanings to utterance-token meanings (IINs) to
utterance-type, pragmatically polysemous meanings (GIINs), to new semantically
polysemous (coded) meanings. Only when certain adverbs or original phrases have
one of these semantic categories as their coded meaning, they will be analysed as
adverbial connector.
6 condition does not play a role here, since it is only coded by subordinators, not by
adverbial connectors; see above, Chapter 2.4.
42 The category adverb
3.5. Adverbial connectors in Present Day English: corpus fndings
3.5.1. The corpus of the Longman Grammar
As a basis for the following analyses of historical material, I will now sum-
marize the main corpus fndings for Present Day English as provided by the
comprehensive account of the Longman Grammar of Spoken and Written
English (Biber et al. 1999: 765776 and 880892). For the analyses of the
Longman Grammar, the team compiled a core corpus of four main registers
comprising ca. 5 million words each: conversation, fiction, newspaper
language and academic prose (see the summary and tables in Biber et al.
1999: 2428).
3.5.2. Different types of adverbials over the core registers
Although optional, adverbials in general are comparatively common a fact
which is, however, hardly surprising in view of the great variety of functions
and meanings they may code. While circumstance adverbials are by far the
most common class of adverbials in all four registers, linking adverbials are
most common in academic prose, but they still only account for less than 10
per cent of all adverbials in that register. While fiction makes great use of
many different circumstance adverbials for emphasis or description, conver-
sation has a slightly higher number of stance adverbials than the other regis-
ters to convey judgements and abilities or to mark exactly how speakers mean
their utterances to be understood.
Linking adverbials are more frequent in academic prose, which puts an
emphasis on conveying logical coherence. The most important situational
properties for academic prose, the text type primarily chosen for the dia-
chronic study here, are that it is (a) in a written mode, (b) does not show any
signs of interactiveness or online production and (c) has no shared immediate
situation between writer and reader. Its main communicative purpose is infor-
mation, argumentation and explanation for a specialist audience. Linking ad-
verbials are a very apt means to these ends, because they allow the writers to
mark the development of their arguments overtly by relating one proposition
to another and by explicitly showing contrasts, restatements and conclusions.
This is immensely important in this written, non-interactive mode which de-
mands precision.
Adverbial connectors in Present Day English: corpus fndings 43
3.5.3. Syntactic realizations of adverbials
As has been stated above, adverbials in Present Day English may not only be
realized by single adverbs or adverb phrases, but also by noun phrases, prepo-
sitional phrases, fnite and non-fnite clauses, such as participial clauses (Now
adding/added to that ), to-infnitive clauses (To conclude, ) or verbless
clauses. Apart from non-fnite and verbless clauses, these possibilities have
also been available in all of the earlier periods of English.
In terms of overall frequency (Biber et al. 1999: 767770; numbers per
one million words), prepositional phrases (ca. 50,000) and single adverbs (ca.
30,000) are the most common syntactic realizations, while fnite clauses (ca.
10,000), non-fnite clauses (ca. 5,000), noun phrases (ca. 5,000) and adverb
phrases (ca. 2,500) are relatively rare. Closer analysis shows, however, that
only circumstance adverbials exhibit a strong preference for prepositional
phrases. Stance and especially linking adverbials are realized more commonly
by single adverbs.
Linking adverbials display the strongest association with a single syntac-
tic form: almost 80 per cent of the linking adverbials are realized by single
adverbs (Biber et al. 1999: 768). The Longman Grammar explains these dif-
fering realizations by the fact that, unlike circumstance adverbials realized by
a prepositional phrase, many stance and linking adverbials have a more fxed
meaning. Individual adverbials of these types are therefore used in a much
higher token frequency a frequency which we would expect from linguistic
items forming a closed class (though the Longman Grammar does not argue
along those lines).
For the diachrony of adverbial connectors, it is also crucial that prepositional
phrases which are used as linking adverbials (e. g. as a result, for that cause) tend
to be less versatile than those used as circumstance adverbials. These frequent,
fxed collocations are a prerequisite for many of these prepositional phrases to
be fused and eventually univerbated and lexicalized (see below, Chapter 7).
3.5.4. Positions of adverbials
Adverbials are comparatively free as to their position in the sentence. While
subject, verb and object(s) are sequentially fxed in Present Day English, ad-
verbials can take three major positions: initial, medial, and fnal.
7
The overall
7 The Longman Grammar does not distinguish the position after the subject (Nach-
erstposition). In the present study, this position has, however, emerged to be very
44 The category adverb
frequencies show that all of the three positions distinguished in the Longman
Grammar are common (Biber et al. 1999: 770774). Each class of adverbial,
however, has a strong preference for a different position: while circumstance
adverbials are most frequently found in fnal position, stance adverbials are
commonly found medially. Linking adverbials favour initial position.
This distribution is obviously related to or even dependent on the different
meanings, and, in particular, scopes of these adverbials. Circumstance adver-
bials often complete the meaning of the verb and must thus follow the verb (and
therefore, in Present Day English, also the subject). Stance adverbials, which
typically have an extended scope over the proposition of the entire clause, may
be placed rather freely. Linking adverbials are most often used in initial posi-
tion, so that the connection between two clauses is clearly signalled as the
reader or hearer moves from the frst to the second connect. Thus the readers
or hearers processing of the discourse is facilitated by the initial linking ad-
verbials explicitly marking the logical relationship between the connects. The
quantitative fndings, however, also show another interesting feature, namely
that linking adverbials may also be placed at the end of a sentence (for example,
PDE causal then, concessive however and though, though predominantly in the
spoken medium). My diachronic analysis has shown this to be a relatively re-
cent phenomenon. Final positioning of linking adverbials seems to run counter
to the arguments relating to information processing given above, because one
wonders why clausal connection should be marked retrospectively at the end
of a sentence, i. e. not between the frst and the second connect, but after the
second connect. Since this position is specifc to English adverbial connectors
and is only rarely allowed with German connectors (Nachsatzposition; see
Pasch et al. 2003: 572), it will be discussed in more detail below (Chapter 10.5).
3.5.5. Linking adverbials: distribution of semantic categories
Linking adverbials are, as has been pointed out above, considerably more com-
mon in conversation and academic prose than in fiction and news. For
academic prose, this has already been related to register-specifc properties.
It is a very important purpose of academic prose to present and support argu-
ments in a precise and unambiguous way. The Longman Grammar relates this
higher frequency of linking adverbials to these communicative needs and, fur-
thermore, points out that it is the characteristic choice of this register to mark,
as the arguments are developed, the links between ideas overtly.
different from other medial positions with respect to focussing strategies (for
details, see below, Chapter 5.4.3).
Adverbial connectors in Present Day English: corpus fndings 45
In the other registers, in particular news and fiction, it is much more common
to leave the relationships apart from contrast/concession (see above) be-
tween ideas implicit. In news, for example, sequences of events are reported with
respect to their relationship in time, and relations of cause and result are pre-
sented so as to be inferred from a chronological sequence or common knowledge
and experience. In news and fiction, an overuse of linking adverbials might,
according to the Longman Grammar, sound too academic and patronizing.
This is also refected by the fact that texts of academic prose in particu-
lar use the pure linking adverbials expressing the relations enumeration,
addition and summation more commonly than the other registers. Unlike
concession, for instance, which highlights a contrasting information and is
therefore most commonly marked in all registers, these pure semantic rela-
tions in particular do not have to be obligatorily marked, but are explicitly
coded in academic prose, as this genre specifcally tries to mark all logical
relations overtly. These connectors help to structure the information in often
rather complicated academic prose and give the readers clear signposts of
where they are in the text.
It may at frst glance be surprising that linking adverbials are more common
in conversation than fiction or news. Yet, this higher frequency is mainly
due to the high frequency of two adverbial connectors, namely so and then as
resultive/inferential linking adverbials, and to an increasingly frequent use of
anyway and though.
3.5.6. Summary: Present Day English corpus fndings
These corpus fndings for adverbial connectors in Present Day English frst
of all suggested restricting the historical analysis to single adverbs, since they
constitute 80 per cent of the realizations of linking adverbials in Present Day
English. In view of the origin of many of the items, such as, e. g., nonethe
less or indeed, in (prepositional) phrases, however, the inclusion of lexicalized
prepositional phrases such as in fact, after all proved to be necessary (for the
problems of distinguishing lexicalized phrases in earlier stages of English, see
Chapter 7).
The positional frequencies further indicate that the sentence-fnal position of
linking adverbs awaits a closer analysis, since it is a recent phenomenon run-
ning counter to the expectations of information processing.
The predominance of adverbial connectors in the register academic prose
then explains the choice of corpus texts (see Appendix C.2). Fiction, for which
we would have had a substantial amount of material from all periods of English
46 The category adverb
in the form of saints lives or other narrative prose, is excluded here because
the Present Day corpora attest only very few numbers of adverbial connectors
for fctional texts (see also the examples, above, Chapter 1.7). News also shows
very few instances of adverbial connectors, and is furthermore a register which
may only be analysed with the emergence of the frst newspapers, i. e. from the
middle or end of the eighteenth century at the earliest.
For conversation, the register which shows the second highest score for
adverbial connectors, we have too little evidence from earlier periods of the
language to allow substantial results. In order to get at least an impression of
the spoken mode, all relevant texts from the Helsinki Corpus, such as Trials or
Plays, were analysed. In contrast to Present Day English, however, adverbial
connectors are not very frequent in these interactive, conceptionally oral texts.
What we fnd are a high number of interrogatives used as interjections, such as
why, how (see Appendix B.5.4); these, however, are excluded from the present
investigation of adverbial connectors because they do not mark a two-place re-
lation, but are rather to be seen as pragmatic idioms, constituting independent
(pragmatic) messages (Fraser 1999: 943; see below, 3.6). The analysis of these
texts has, nonetheless, proved to be highly benefcial because they record the
frst instances of sentence-fnal adverbial connectors (see below, Chapter 10.5).
The focus on academic prose is, however, not only due to the high frequen-
cy of adverbial connectors in this register, but also to the fact that the Longman
Grammar suggests that it is the characteristic stylistic choice of this register to
mark the links between ideas overtly.
3.6. Adverbial connectors discourse markers
In various research contexts, some of the items discussed in the present study
have not been classifed as adverbs, but as discourse markers (OE a in
Enkvist and Wrvik 1987 and Kim 1992; OE solice, OE witodlice in Lenker
2000; EModE indeed, in fact in Traugott and Dasher 2002; PDE after all in
Traugott 1997; for a discussion, see Fischer 2007: 280281). In the present
study, they are categorized as transitional. This label highlights that they
do not encode any kind of propositional meaning, but only work on the in-
terpersonal and textual levels, signalling the connection of two segments of
discourse, i. e. their status as connects.
The term discourse marker is notoriously diffcult to defne and has seen
manifold defnitions in the last years, when the study of discourse markers
has turned into a growth industry in linguistics (Fraser 1999: 932). Most of
the research on discourse particles for English is, sometimes critically, based
Adverbial connectors discourse markers 47
on the frst systematic and detailed study in the feld by Schiffrin (1987).
8
In
spite of all the ensuing work on the feld, however, it has not been possible to
defne the term in a generally accepted way, let alone to establish a fxed group
of linguistic items which would essentially be separated from other classes by
being discourse markers. For the issues discussed in the present study, it is
interesting to see that there are very wide defnitions of the concept discourse
marker. Fraser (1999), for example, arrives at the following defnition. Dis-
course markers are
lexical expressions drawn primarily from the syntactic classes of conjunctions, ad-
verbs, and prepositional phrases. With certain exceptions, they signal a relationship
between the segment they introduce, S2, and the prior segment, S1. They have a
core meaning which is procedural, not conceptual (Fraser 1999: 950).
The properties proposed in this defnition are basically the same as those pos-
ited for connectives in the present study (following the Handbuch der deut
schen Konnektoren, Pasch et al. 2003; see above, Chapter 2.2). Their central
property is that they are procedural rather than conceptual: They relate dis-
course segments, i. e. mark a two-place relation (Fraser 1999: 931) and do not
contribute to the propositional meaning of either segment. To capture these
properties, the present study uses the term connector. It excludes in line
with Fraser modal particles and style as well as content disjuncts, since
these do not signal a two-place relation (see the exclusion of peradventure,
perhaps, perchance; Chapter 12.2). It also excludes interjections such as oh,
yes/yea, or no/nay (see Appendix B.5.5), which do not mark two-place rela-
tions either, but are pragmatic idioms in that they constitute entire, separate
messages (Fraser 1999: 943).
Yet some of the connectives, in particular those coined from the source do-
main truth (see Chapter 8.4), show a development which has been labelled
Adverbial Cline in a semantic-pragmatic approach to grammaticalization
(Traugott 1997, 1999). This cline from clause-internal adverbial to sentence ad-
verbial to discourse particle is established as a regularity in semantic change
(Traugott and Dasher 2002) in order to distinguish it from grammaticalization
8 Schiffrin (1987: 328) gives the following tentative suggestions as to what specifc
conditions allow an expression to be used as a marker. A discourse marker has to
be syntactically detachable from a sentence, it has to be commonly used in initial
position of an utterance, it has to have a range of prosodic contours (e. g. tonic stress
and followed by a pause, phonological reduction), it has to be able to operate at both
local and global levels of discourse this means that it either has to have no mean-
ing, a vague meaning () . Cf. Kroon (1995: 757) and Brinton (1996: 2965) for
the literature on discourse markers and its terminology.
48 The category adverb
processes of the morpho-syntactic kind which involve, for example, increased
bonding and a loss of syntactic scope. The diachrony of many epistemic modal
adverbials expressing the speakers commitment to the truth of the proposition
(OE solice, ME forsoothe, EModE truly) shows that they at some late stage
of their development do no longer primarily signal a two-place relation, but
carry an independent pragmatic meaning, i. e. are discourse markers in a
more restricted sense. On the global level of textual organization, OE solice
and ME soothly, for example, often serve as episode boundary markers (see
below, Chapter 8.4), a function which may at the textual level be considered
connective. Furthermore, these items also exhibit other patterns of recur-
rent change than the other adverbial connectors (see below, Chapter 12). The
discrete, peripheral status of these connectors is indicated by their label tran-
sitional (which implies their borderline status). Yet, these considerations do
not affect the general relevance of these items for the present study: Before
they develop these primarily pragmatic functions, they unambiguously signal
two-place relations, mainly on the local level of discourse (Traugotts stage
sentential adverb; for details, see Chapter 8.4). Furthermore, their status as
connectives is also corroborated by the information gathered from meta-lin-
guistic texts of the respective periods (see the following Chapter 4).
4. Adverbs and conjunctions in earlier
metalinguistic thought
4.1. Introductory remarks
One source from which to collect relevant linguistic material which allows
for a better understanding of the development of adverbial connectors and
their relation to conjunction are metalinguistic texts. While it is not diffcult
to gather data from Present Day English sources such as grammars or dic-
tionaries, contemporaneous metalinguistic sources are unfortunately of rather
limited value before the middle or end of the Early Modern English period:
the sources before Campbell ([1776] 1963) do not distinguish adverbial con-
nectors as a separate, distinct category. The items are usually listed among
the adverbs, which are commonly sub-grouped semantically (see the quotes
from Locke in Chapter 1.1), but are not further distinguished into functional
subclasses. For a long time, the elements in question were included in an even
larger group of indeclinable items and are classifed together with prepositions
and conjunctions as particles. The increasing sophistication in the classifca-
tion of these linguistic elements is not only interesting because it illustrates the
developments in English language scholarship, but also because it mirrors the
growing importance of these elements, both in frequency, diversity and stabil-
ity of use in and after the Early Modern English period.
4.2. The Greek and Latin tradition
For a close analysis of adverbial connectors and conjunctions in earlier periods
of English, it is obviously also crucial to look at how they were conceptualized
by medieval grammars of Latin and the vernacular. Following the categoriza-
tion of Dionysius Thrax, there are eight parts of speech that turn up again and
again in the Latin grammatical tradition, namely (1) noun, (2) pronoun, (3) ar-
ticle, (4) verb, (5) participle, (6) adverb, (7) conjunction, and (8) preposition (see
Vorlat 1975: 43; Law 1997: 264269). Medieval grammarians take on Thraxs
classifcation, which they were acquainted with via transmission by Latin au-
thors such as Donatus, with only two adjustments: articles are omitted because
they do not exist in Latin, and the interjection is detached from the adverb and
becomes a separate word class (Vorlat 1975: 43). The standard sequence in me-
dieval grammars thus is nomen, pronomen, uerbum, aduerbium, participium,
coniunctio, praepositio, and interjectio.
50 Adverbs and conjunctions in earlier metalinguistic thought
The word classes which are important for our study adverb, interjec-
tion, and conjunction together with the preposition form the minor parts of
speech, which generally receive little attention from the grammarians (Vorlat
1975: 366419). These minor parts of speech are separated from the other word
classes by their indeclinable character and their dependency on other word
classes (particles).
The classifcation of adverbs into the syntactically determined subgroups
outlined above is a fairly modern one: in traditional grammatical treatises,
adverbs are not usually categorized according to their syntactic properties. The
most infuential early grammarians (Thrax, Varro, Donatus, and Priscian) take
the term adverbium in the strict sense of a word added to a verb whose meaning
it modifes or explains, and the authors indulge in elaborate, if not to say te-
dious, semantic classifcations of adverbs.
1
This exclusively semantic approach
is taken on by most of the English grammarians also those following the tra-
ditions of Ramus or the Port-Royal Grammar (see Michael 1970: 7275).
2
The
subcategories we now call stance adverbials (disjuncts) and linking adverbials
(conjuncts) are not mentioned in any of the grammars of the Medieval and Ear-
ly Modern Period (Vorlat 1975: 366387). This is not really surprising because
these categories are, as shown above, mainly based on syntactical properties.
In the early grammars, however, syntax is mostly restricted to questions of
concord and government (cf. Gneuss 1996: 15). Yet, the early grammars do not
completely lack value for the issue in question. Although they do not explicitly
refer to specifc syntactical properties, the grammarians repeatedly point out
that the distinction between the various word classes of the group of indeclin-
ables is not always clear-cut.
3
1 Donatus, for example, has adverbia fnita et infnita, and furthermore the catego-
ries adverbia loci, temporis, numeri, negandi, adfrmandi, demonstrandi, optandi,
hortandi, ordinis, interrogandi, similitudinis, qualitatis, quantitatis, dubitandi,
personalia, vocandi, respondendi, separandi, iurandi, eligendi, congregandi, pro
hibendi, eventus, comparandi (Ars Minor, De Adverbio, II, XIII). Priscian has
temporalia (with further subclassifcation), locorum, dehortativa, abnegativa, con
frmativa, iurativa, optativa, hortativa, remissiva, qualitatis, quantitatis, dubita
tiva, congregativa, discretiva, similitudinis, ordinativa, intentiva, comparativa,
superlativa, diminutiva (XV, 2837; cf. Vorlat 1975: 367).
2 The Rameian tradition, for example, contrasts adverbs as word connectors with
conjunctions as clause and sentence connectors (see Vorlat 1975: 368).
3 Various grammarians further remark that a number of prepositions in Latin as well
as in English (cf. up) may also function as adverbs (Vorlat 1975: 376). Adverbs are
furthermore often considered to be of an abbreviating character, because they may
be paraphrased by a prepositional phrase consisting of a preposition and a noun.
Thus sapienter is seen as short for cum sapientia or English wisely short for with
lfrics Old English Grammar 51
What is even more important for the present study in particular the situa-
tion in Old and Middle English is that already Priscian and Donatus point out
that they see no clear distinction between adverbs and conjunctions. In a dis-
cussion of the adverb, Donatus, for instance, explains that there is sometimes
no way of determining whether a specifc item is an adverb, a conjunction or a
preposition, unless its function in the sentence is taken into account:
Sunt etiam dictiones, quas incertum est, utrum coniunctiones, an praepositiones, an
adverbia nominemus [], quae tamen omnes sensu facile dinoscuntur, [] horum
quaedam accentu discernimus, quaedam sensu (II, XV; cf. Vorlat 1975: 367).
English grammarians adopt this remark time and again, i. e. that word class is
determined by function (Vorlat 1975: 367), which means that the grammar-
ians do indeed see the disparity in the scope of different adverbs, but do not
use this criterion as a property for classifcation. For the Early Modern gram-
marians, it does not seem to be an important issue; Newton, for example, in an
imitation of Wallis, states: I shall reckon some of these Words as Adverbs, and
some of the Adverbs as Conjunctions, they being often used in both Senses,
there will be no great harm done (Vorlat 1975: 376).
4.3. lfrics Old English Grammar
With respect to metalinguistic thought on English, the earliest extensive gram-
matical account is lfrics Latin-Old English Grammar (Zupitza [1880] 2001).
It is frst of all important to note that lfrics account, which is based on a
Latin excerpt of Priscian (see Porter 2002), basically follows the Latin tradition
in that he distinguishes adverbs and conjunctions in principle (see the two sec-
tions entitled Incipit Aduerbium and De Coniunctione). This distinction is,
however, overridden by functional considerations.
This may be exemplifed by the case of the adverb solice truly, verily,
which is morphologically clearly marked as an adverb by the adverbial suffx
{-e} added to its base, the adjective solic true (for details, see Lenker 2000,
2003 and below, Chapter 8.4.4). lfric, however, does not list it in his section
on the adverbs (Incipit Aduerbium; Zupitza 2001: 222242), but explicitly re-
fers to this adverb (!) in his section De Coniunctione (Zupitza 2001: 257266,
at 261). Conjunctions are defned as follows:
wisdom (see the examples of the tradition of the Port-Royal Grammar given in Vorlat
1975: 374, 380). For a recent grammar which revives this classifcation of adverbs in
one group with preposition, see Huddleston and Pullum (2002: 264).
52 Adverbs and conjunctions in earlier metalinguistic thought
Coniunctio est pars orationis indeclinabilis adnectens ordinansque sententiam.
swaswa lim gefstna fel to sumum brede, swa getig seo coniunctio a word tog-
dere. es dl gefstna and and gefrtwa ledensprce and hwilon tosct and
hwilon geendebyrt. nf es dl nane mihte ne nan angit, gif he ana stent, ac
on endebyrdnysse ledensprce he gelima a word (Zupitza 2001: 257258).
A Conjunction is an indeclinable part of speech connecting and ordering the sense
; in the same way as lime fxes a hide to any surface, so conjunction draws the
words together. These parts of speech fx and adorn Latin and sometimes they
divide and sometimes they order. This part of speech does not have any power
and no meaning, if it stands on its own, but in the order of Latin it links the words.
By discussing adverbs like solice truly in the section De Coniunctione,
lfric clearly shows that he considers the scope of the adverb to be not the verb
phrase but the sentence: he thus implicitly classifes solice as an adverbial
connector, a conjunct.
lfric even further distinguishes between two uses of solice as a conjunc-
tion. First, solice (and also witodlice and gewislice) are found as translations
of Latin items, which in our terminology are not conjunctions, but adverbial
connectors, namely autem, enim, uero and nam etc., which belong to the group
of the Expletivae or Completivae a gefylla and gefgeria a ledensprce,
and, eah e hig forltene beon, ne by swa eah re sprce andgit forlten
which fll and adorn the Latin, and, even if they are left out, the sense of the
utterance will not be lost (Zupitza 2001: 261):
[H]er synd a: autem, enim, uero, quidem, equidem, quoque, nam, namque, uide
licet. tu autem, domine, miserere mei et resuscita me. u, solice, drihten, miltsa
me and arr me. ego enim sum dominus, deus tuus ic, solice, eom drihten in
god. [] doctus sum nam legi ic eom gelred; solice, ic rdde. erat namque in
sermone uerax he ws, solice, on sprce sofst [].
lfric explicitly refers to a loss of the propositional meaning (ne by swa
eah re sprce andgit forlten), which is essential for the use of solice
truly as a sentential adverb with text-organizing function, i. e. as a pure,
transitional connector.
4
This is most obvious in the last example cited above
(erat namque in sermone uerax he ws, solice, on sprce sofst, He
was, therefore, true-speaking/reliable in his speech): lfric would certainly
not have chosen solice to render namque (but one of the alternatives such as
gewislice or witodlice) if its propositional meaning truly had still been princi-
pal to him, since this ambiguity could have obstructed the understanding of the
4 For the loss of propositional meaning (bleaching) and pragmatic strengthening as
prototypical features of grammaticalization and similar kinds of language change,
see Hopper and Traugott (1993: 8793).
Grammatical treatises of the Middle English period 53
proposition of the sentence, which has to do with true speaking (Lat. uerax,
OE sofst true, sincere).
The adverbs used in the function of a conjunction are, however, not only
stylistically important for the adornment of a text. lfric also lists witodlice
certainly, truly among the Rationales, as sind for sumon gesceade gesette
on endebyrdnysse ledensprce which are set for an (understanding of the)
argument in the text organisation of Latin (Zupitza 2001: 263). lfric thus
stresses that their function, translating Latin items such as ergo, igitur, ita,
itaque and utique, is to add reasoning to the discourse.
In sum, lfrics Grammar does indeed prove to be helpful for an analysis of
the syntactic scope and functions of Old English adverbs. For lfric, the defn-
ing element for a coniunctio is not the word class (neither for Latin cf. enim,
nam, etc. nor for English cf. solice), but their scope over the sentence and
their function as organizers of textual structure. He thus uses the same criteria
which are today chosen for the categorization of an item as a conjunct or
linking adverbial.
4.4. Grammatical treatises of the Middle English period
With respect to Middle English, grammatical treatises dealing with the ver-
nacular only survive from the end of the fourteenth century onwards, and their
production reaches its peak as late as 14601480. These Middle English gram-
matical texts, which are now conveniently collected in Thomson (1984), were
not designed as abstract grammars but as working tools for learning Latin
(Thomson 1984: xiv) and show an immense degree of interdependence. Be-
cause of their repetitive character, it suffces to quote exemplary samples from
one of the sources here (Aberystwyth, N. L. W., MS Peniarth 356B, fols. 54v
57v and 48r; Text A in Thomson 1984). The adverb, in accordance with the
Latin school tradition, is seen as modifying the verb only:
How knos u aduerbe? A party of speech at ys vndeclynyt, e wych ys cast to a
verbe to declare and fulfyll e sygnifc(ac)ion of e verbe [] (Thomson 1984: 6).
The accounts because they are grammars of Latin and not English only pro-
vide information on the different patterns of Latin adverb formation from the
different classes of adjectives, but virtually no English examples. This method
is also applied in the descriptions of the conjunctions:
How knos a coniunccion? A party of speech at ys vndeclynet and ionys oer par-
tys of spechys togedyr. How mony thyngus longon to a coniunccion? III. Wech iij?
Powere, fygur and ordyr. <How mon>y powers of coniunccion byn er? V. Wech v?
54 Adverbs and conjunctions in earlier metalinguistic thought
Sum be copulatyuis, sum byn <disiunctiuis, sum> byn explatyuis, sum casuels and
sum racionels (Thomson 1985: 7).
The only account of the conjunction which is not taken from Priscian or Dona-
tus and which supplies some information on the vernacular is found in Thom-
sons Accedence Text D (Thomson 1984: 3243; Cambridge, Trinity College,
O. 5. 4, fols. 4v6v). This source provides an exposition of the various catego-
ries of conjunctions which is unusually full; more importantly, it gives sample
sentences and their English translations (cf. Thomson 1984: 244):
How knowest a coniunccion? A party of reson that is not declynyd []. How know-
est a coniunccion expletyf? That at fulfylleth the sentence of a reson that is folw-
yng, as I forsothe haue souped, thu forsothe not, Ego quidem cenaui tu vero non
(Thomson 1984: 42).
Interestingly, the adverb forsothe truly is here like in lfrics Grammar
considered a conjunction of the subgroup Expletivae.
4.5. The prologue to the revision of the Wycliffte Bible
Apart from these very scarce remarks in grammatical treatises, there is some
information to be gained from the General Prologue to the revision of the
Wycliffte Bible (1388), which discusses several translation problems. Among
them are two passages which overtly deal with the translation of Latin adver-
bial connectors, such as autem or enim.
And whanne oo word is oonis set in a reesoun, it mai be set forth as ofte as it is
vndurstonden, either as ofte as reesoun and nede axen; and this word autem, either
vero, mai stonde for forsothe, either for but, and thus I vse comounli; and sumtyme
it mai stonde for and, as elde gramariens seyn (Forshall and Madden 1850: 57).
The author here explicitly asks for a repetition of certain conjunctions to make
the reasoning of the text, the textual organisation, transparent. He includes the
adverb forsothe among the words rendering Latin autem and vero and marks
their similiarity, but also difference, to unambiguous conjunctions such as ad-
versative but and connective and.
In a second instance, the reviser deals more generally with the problems of
translating items belonging to the minor parts of speech, such as aduerbis,
coniuncciouns, and prepositions, in a section which shows a high degree of
awareness of the problem of textual organization:
[A] translatour hath greet need to studie wel the sentence, both bifore and aftir, and
loke that suche equiuok wordis acorde with the sentence []. Also this word ex
signifeth sumtyme of, and sumtyme it signifeth bi, as Jerom seith; and this word
The prologue to the revision of the Wycliffte Bible 55
enim signifeth comynli forsothe, and, as Jerom seith, it signifeth cause thus, for
whi; []. Manie such aduerbis, coniuncciouns, and preposiciouns ben set ofte oon
for a nother, and at fre choice of autouris sumtyme; and now tho shulen be taken as
it acordith best to the sentence [] (Forshall and Madden 1850: 5960).
Here then, ME forsothe is given as a translation of enim and thus classifed as
a causal connector. Yet, both passages agree on the description of forsothe as a
rendering of Latin conjunctions and thus testify to the analysis of forsothe as an
adverbial connector with scope over the whole sentence.
The revision of the Bible translation itself, however, points towards a change of
attitude on the side of the reviser(s). In the Old Testament, autem, vero and enim
are almost always translated by forsothe and sothely in the earlier version, and also
very frequently in the second version (the alternative being but). In the New Testa-
ment, the earlier text again renders them by forsothe or sothely. The revised text,
however, almost exclusively uses the conjunctions but or and, or leaves autem,
vero etc. unrendered (Forshall and Madden 1850: xxiii, note a). This becomes
clear through a comparison of some verses of the Gospel according to Matthew,
chapter XXVI (The Plot against Jesus in the Earlier and Later Version):
Table 41: Translations of Lat. autem and enim in the Wycliffe Bible
(Matthew XXVI)
Verse Latin Earlier Version Later Version
XXVI, 8 [uidentes] autem sothely and
XXVI, 9 [potuit] enim forsothe for
XXVI, 10 [sciens] autem sothely But
XXVI, 12 [mittens] enim forsothe
XXVI, 17 [prima] autem [die] forsothe And
XXVI, 24 [flius] quidem forsothe Forsothe
XXVI, 26 [cenantibus] autem forsothe And
XXVI, 29 [dico] autem [uobis] forsothe And
XXVI, 32 [postquam] autem forsothe But
XXVI, 33 [respondens] autem sothely
XXVI, 41 quidem [promptus est] forsothe for
XXVI, 43 [erant] enim forsothe for
This comparison shows that the reviser, although he argues strongly in favour
of forsothe in the General Prologue where he refers to it as the common
translation of enim, has almost completely abandoned the word by the time it
comes to translating the New Testament. This frst of all allows us to infer that
contemporary writers and grammarians regarded forsothe as a conjunct with
56 Adverbs and conjunctions in earlier metalinguistic thought
text-organizing function which could and should be employed in prose texts to
translate Latin conjunctions, but it also seems to indicate that it was more and
more avoided in genuine Middle English prose.
In the slightly later translation of the Latin Technical Phlebotomy
( CLPHLEBO; ca. 14001420), for example, there is a striking difference be-
tween the Middle English translation of the Latin text which regularly uses
forsoe in post-frst-position, i. e., the sentence position of enim or autem, as in:
(38) Alexandire forsoe commandi a pacient leucofamcie to be feuboto-
myed; euel humourus forsoe if ey be in veynes ar competenly brougt
out be febotomye. Aftir, electuary frigidum confortatiue of e lyuere;
e lyuere forsoe it is moche wonte to be febled. Laboryng of ictercian
citrina, i. e., gelow iaundeyes, & also of agriacape gasiliontes, i. e., grene
jawnes, mynusche of bacilica of e rygt arme (CMPHLEBO, p. 51).
In a later, originally Middle English part, forsoe is not used at all. Instead,
items such as sentence-initial wherefore, for, also or for hwi are chosen as con-
nectors, as in:
(39) Wherefor som vnwise men after e consayle of oer onwise men will
ann mynusche hem, & of at folwi a werse errore ann e frste.
For e materie at was witoute e weynes be febotomye is drawen to
wiinne e veynes; ann falli & continewe febre & so oftyn time de.
Also som men when ei have a febre interpolat where e materie is
wiout e veynes, maki hem to be mynusch & falli into continell
febre; for whi: as (CLPHLEB, p.53).
4.6. Early Modern English dictionaries and grammars
For the Early Modern English period until Locke, i. e. from 1560 to the mid-
dle of the seventeenth century, our main relevant metalinguistic sources are
dictionaries (conveniently accessible in The Early Modern English Dictionar
ies Database, ed. Ian Lancashire EMEDD, now part of the database Lexi
cons of Early Modern English (LEME)), which includes 16 works from 1530
to 1657: six bilingual dictionaries,
5
fve English hard-word dictionaries,
6
the
5 John Palsgrave (1530; English French), William Thomas (1550; Italian English),
Thomas Thomas (1587; Latin English), John Florio (1598; Italian English), John
Minsheu (1599; Spanish English), and Randle Cotgrave (1611; French English).
6 Edmund Coote (1596), Robert Cawdrey (1604, based on the transcription by Ray-
mond Siemens; and 1617), John Bullokar (1616), and Henry Cockeram (1623).
Early Modern English dictionaries and grammars 57
frst full English-only dictionary by Thomas Blount (1656), three specialized
lexicons,
7
and the frst full English word-list by Richard Mulcaster in his The
frst part of the Elementarie (1582). Although these dictionaries in line with
the grammars for their time do not separate adverbial connectors from other
kinds of adverbs or conjunctions, they are a valuable source because they do
not only give lexical, but also functional synonyms (for examples, see below,
Chapter 11).
As far as grammars are concerned, there was no change in the treatment of
adverbs and conjunctions. To cite just one example, Lowth ([1762] 1969) has
the same limited defnition of adverbs as the Middle English grammatical trea-
tises, restricting them basically to modifers of verbs and adjectives:
Adverbs are added to Verbs and Adjectives to denote some modifcation or circum-
stance of an action or quality: as, the manner, order, time, place, distance, quality
(Lowth [1762] 1969: 90).
A conjunction is said to connect or join together Sentences, but the example
Lowth gives deals with conjunctions on the phrasal level:
so as out of two to make one Sentence. Thus, You, and I, and Peter rode to
London, is one Sentence made up of three by the conjunction and twice employed
(Lowth 1969: 9293).
This brief summary shows that Lockes intervention for the Particles in his
Essay Concerning Human Understanding ([1690] 1975) quoted at the very
beginning of this study did not have any major impact on the grammarians of
that period. It had, however, an impact on the very infuential school of rheto-
ricians, which promoted a re-orientation of rhetoric in line with the ideas
of the Enlightenment, notably the so-called Scottish Rhetoricians advocat-
ing a New Rhetoric. In Chapter 13, the ideas promoted by one of these
rhetoricians, George Campbell, in his The Philosophy of Rhetoric, will be
shown to have been decisive for the increasing sentence-medial positioning
of adverbial connectors in the written mode. The increasing sophistication in
linguistic terminology thus nicely mirrors the development of the linguistic
category itself.
7 Bartholomew Traherons translation of Vigon (1543), William Turner on herbal
names (1548), and John Garfeld on scientifc terms in J. Renous Dispensatory
(1657).
5. Connectors in Old English
5.1. Semantic and syntactic polyfunctionality
To understand the history of adverbial connectors and, more generally, clause
linkage in English over the centuries, it is necessary to frst describe the situ-
ation in Old English. In Old English, there are only very few linguistic items
which function exclusively as adverbial connectors (Group A),
1
namely the
complex formations nayls, swaeah and swaeahhwere nevertheless,
all of which mark the semantic relation contrast/concession.
All of the other connectors are semantically and/or syntactically polyfunc-
tional (Groups B to E). In essence, we can differentiate four major groups: (B)
circumstance or stance adverbs with a connective force, (C) so-called am-
biguous adverbs/conjunctions, (D) pronominal connectors, and (E) case forms
of the demonstrative.
(B) Circumstance and stance adverbs with connector force
(a) Circumstance adverbs adverbial connectors addition/ transition
rest frst
eac also
eft also
eftsona also
eftsones also
elles else
her here
nu now
(r)toeacan in addition, besides
a then
onne (mainly in post-frst-position) then (transition cause/result)
(b) Epistemic adverbs (stance adverbials) adverbial connectors transition:
eornostlice earnestly
huru indeed, truly
1 Mitchell (1985: 34303433) shows that an adverbial reading is possible or prefer-
able for all instances of OE hwere, swaeah and swaeahhwere for which
conjunction uses have been suggested.
Circumstance adverbs with connective force (Group Ba) 59
huruinga at least
solice truly
gewislice certainly
witodlice truly
(C) Ambiguous adverbs/conjunctions (symbol ):
2
eall-swa adv. also, moreover conj. as, so
nu adv. now, transition conj. now that
swa adv. so conj. so that ; as
eac (swylce)/ swylce eac adv. also, moreover conj. all such
a adv. then, transition conj. then when
eah adv. nevertheless conj. although
onne adv. then, transition/cause conj. then when
(D) Pronominal connectors:
form (foron, fory) adv. therefore conj. because
nayls nevertheless
(E) Case forms of the demonstrative adverbial connector cause/result:
(a) genitive s therefore
(b) instrumental y (ambiguous adverb/conjunction)
adv. therefore conj. because/since
5.2. Circumstance adverbs with connective force (Group B-a):
temporal and spatial adverbs
For some temporal and spatial circumstance adverbs, a conjunctive force is
widely established: frequently used connectors in Old English are OE eac, eac
swylce/swylce eac, (r)toeacan also (addition), eft again (addition), her
here (transition), nu now (transition), and a then (transition). All
of these thus code the semantic relations Addition or Transition.
While some of them are exclusively adverbs and are only semantically poly-
functional, a case here exemplifed by eft (see also eac, eac swylce/swylce eac,
2 For this term see Mitchell (1985: 2536). The adverbial meaning is given frst; this
is followed by the conjunction meaning after the dash. In each case, only one the
central meaning is given.
60 Connectors in Old English
(r)toeacan and her), nu and a may also be used in correlative construc-
tions and on their own as subordinators (here exemplifed by nu). They
are semantically and syntactically polyfunctional, and are therefore termed
ambiguous adverbs/conjunctions.
5.2.1. OE eft again
In origin, adverbial connectors of group B follow the path from circumstance
adverbial modifying a verb of communication on the phrase level, as sketched
above in Chapter 3.2.3. Eft again, for example, can be used as a regular cir-
cumstance adverb of time again, another time, once more (DOE, s. v. eft, 1.),
as in:
(40) . and se feh on Ianuario and r eft geenda.
and it begins in January and ends there again (ByrM, 1.1.229)
3
As such a circumstance adverb, it can modify a verbum dicendi (DOE, s. v. eft
4.a.iii), such as cyan speak; announce in
(41) Nu we wylla eft fullum mode her gecyan hu oft se Easterlica mon
ongin on Martio.
Now (time) we will again (time) make known here (space) with
thorough thought how often the Easter month begins in March (ByrM,
3.2.96)
Already in Old English, however, the co- and context in (41) can also be de-
leted. In contexts such as (42), eft functions as an additive/reinforcing adverbial
connector again, as another point of fact (indicating sequence or transition in
discourse) (DOE, s. v. eft, 4.). In these functions, eft is used quite typically in a
3 To illustrate the situation in Old English, examples are primarily chosen from texts
which are comparatively independent from Latin originals. Whenever possible,
they are taken from one of the Prefaces (COPREFCP, COPREFSO), the rather free
adaption of Boethius De Consolatione Philosophiae from King Alfreds Circle
(COBOETH) or Byrhtferths Enchiridion (ed. Baker and Lapidge 1991 = ByrM;
not all of the examples given here are in the extract chosen for the Helsinki Corpus,
i. e. COBYRHTF). Quotations from dictionaries or grammars are only used if there
was no instance in one of my selected texts because it is hard to judge from the
single sentences in dictionaries how large the piece of discourse (clause, paragraph,
text) the elements connect is, i. e. whether they work on the local or global level.
Circumstance adverbs with connective force (Group Ba) 61
collocation with another connector, the coordinating conjunction and (for these
collocations, see below Chapter 13).
(42) <Psile>: And eft, gif he by ahwar gesett r he standan ne mg,
onne sceal man hine us genyerian
<Psile>: And moreover, if it is written anywhere it cannot be, then
one must remove it thus: (ByrM, 3.3.184)
The same distribution is also attested similarly for all the other items listed
above. In example (41), for instance, we also fnd spatial, text-deictic her here
(for its transitional use see Clemoes 1985; Fries 1994: 530535) and the tempo-
ral, text-deictic nu now.
5.2.2. OE nu now (Group B-a, Group C)
5221 Temporal nu
Nu now, of course, may be used as a circumstance adverb of time referring
to the actual, present date in the world (now, at this time; BT, s. v. I.). Thus
nu in (43c) and (43d) defnitely refers to an actual date in February in the year
Byrhtferth is writing his Enchiridion, as shown by the obligatory circumstance
adverb in the relative clause e nu is which is now in example (43c) and the
collocation nu todg, onne ic is write now today, that I am writing this in
(43d). It is also a temporal adverb, but with an extended metaphorical meaning
in (43a) and (43b).
5222 Textdeictic nu
In (43a) and (43b), by contrast, nu does not refer to a fxed date, but to the now
of reading/or writing of the text, which may be at various dates in the future,
since the deictic centre here is not the real world, but the text itself (for a col-
lection of Old English examples, see Fries 1994). Nu has become text deictic, a
function which is often highlighted by verbs of communication which refer to
the foregoing (cf. handlian in 43b) or ensuing text production (cf. gecyan in
41 and tywan in 43a):
(43) Nu [43a] wylle ic bysne tywan ymbe a ing e we nu [43b] handle-
don and fste ymbe wron. (\Kl Febris\) , e nu [43c] is, eode on tun
62 Connectors in Old English
on unresdg, & nu to dg, onne ic is write [43d], ys se ffta dg
(ByrM, 1.2.234).
Now [43a] I will give an example of the things we have just now [43b]
treated and were intent upon. The month of February, which is now
[43c], came to town on Thursday, and now today, when I write this
[43d], is the ffth day.
5223 Adverbial connector nu TRANSITION/RESULT
In (44), nu functions solely as an adverbial connector see the collocation with
the subordinating conjunction gif if; here, it indicates a semantic relation of
transition verging on the relation result. This use derives from the text-
deictic uses sketched above, namely a full phrase such as if we now proceeded
to argue that two goods existed where nu again refers to the text as the
deictic centre:
(44) Gif nu tu good wren e ne meahton tsomne bion, & wren eah buto
goode, hu ne wre hit onne genog sweotol t hiora nre nauer t
oer? (COBOETH, 34.85.12)
If, moreover
?
/therefore
?
, two goods existed, which might not be togeth-
er, and were nevertheless both good, would it not then be suffciently
evident that neither of them was the other?
5224 Adverb/conjunction nu in correlative constructions
As an adverbial connector, nu can also just like a then and onne then
be used in correlative constructions, in which connectors are placed in each
of the connects, one functioning as a subordinator, the other as an adverbial
connector.
In the correlative construction in (45), the frst nu (45a) functions as a causal
subordinator introducing a pre-posed topic-forming subordinate clause (BT,
s. v. II Conj. now; since; when; for which Present Day English prefers unam-
biguous now that), while the second (45b) may be classifed as a circumstance
adverb of time, a text-deictic nu (now it seems ftting ):
(45) Nu [45a] we habba sceortlice amearcod ra hiwa gefeg e boceras
gyma, nu [45b] ing hyt us gedafenlic t we heom gecyon t we
r geheton (ByrM 3.3.234)
Now that we have briefy written the series of fgures the writers use,
now it seems ftting to us to tell them what we promised before .
Circumstance adverbs with connective force (Group Ba) 63
This example also shows that in contrast to V2 vs. V-fnal word order in
Present Day German constituent order cannot be used as a solid criterion for
differentiating Old English main and subordinating clauses: both clauses em-
ploy NuV (NuSVAO in the pre-posed subordinate clause; nuVSC in the main
clause).
In the Old English translation of Boethius, for example, there are many in-
stances of correlative constructions with identical constituent order in both the
subordinate and main clause (in 46: Nu SVO):
(46) a cw he: Nu [46a] u onne wast hwt a leasan gesla sint &
hwt a soan gesla sint, nu [46b] ic wolde t u leornodest hu u
(COBOETH, 33.78.27).
Then he said: Now that [46a; subordinator] you know (then) what the
false goods are, and what the true goods are, now [46b; circumstance
adverb] I wish that thou should learn how you .
5225 Conjunction (subordinator) nu
The use of nu as a subordinator is, however, not restricted to correlative con-
structions. It may also be found on its own; pre-posed, topic-forming (47a) as
well as in regular post-posed subordinate clauses (48):
(47) Nu [47a] we ealles heron habba gefangen, hyt gerist cyrtenlice t
we ne wandion naor Vton awendan nu [47b] ure gesetnysse to am
rihtingum e rimcrftige preostas cwea lunares (ByrM, 1.2.251257)
Now that [47a; subordinator] we have fully begun in this matter, it is
splendidly suitable that we not desist either Let us now [47b; circum-
stance adverb + verb of communication] turn our tract to those regulars
which priests skilled in computus call lunar.
(48) Se ilca God is, swa swa we r sdon, t hehste good & a selestan
gesla, nu hit is openlice cu t a selestan gesla on nanum orum
gesceaftum ne sint buton on Gode. (COBOETH, 34.84.3)
The same God is, as we before said, the highest good, and the best hap-
piness since it is evidently known that the best felicities are in no other
things but in God.
This account of nu can be taken as representative for most of the ambiguous
adverbs/conjunctions in Old English (for form, see below, Chapter 9; for the
exception of eah triggering the subjunctive in subordinate clauses, see below
64 Connectors in Old English
Chapter 10.5.2). Word order is only distinctive when verb-fnal order is em-
ployed: it marks a clause as subordinate. In sum, this means that
when standing frst in a sentence, any word which in its written form may be
either an adverb or a conjunction, (i) is in prose invariably a conjunction when fol-
lowed by a conjunctive order, (ii) is ambiguous, i. e. may be either part of speech,
both in prose and verse, when followed by the common order (Andrew 1940: 34).
These exemplary analyses of eft and nu also show that it is not possible to
undertake the kind of investigation envisaged in the present study by a ma-
chine-triggered count of items in the respective language corpora. The se-
mantic and syntactic polyfunctionality of the items requires a close reading
of the respective texts and the classifcation of highly frequent linguistic items
in view of their specifc co- and contexts. This analysis obviously also had to
be restricted to selected corpus texts, since most of the items in question are
high-frequency items in their respective periods: for Old English, the DOE
lists, for instance, 6,800 occurrences for eac, 5,500 for eft and 15,500 for the
various forms of form.
5.3. Ambiguous adverbs/conjunctions (Group C)
5.3.1. OE a
The most important group of connectors in Old English are those high-fre-
quency items that are, for lack of a better term, called ambiguous adverb/
conjunctions (see Mitchell 1985: 3010). Old English a then, for example,
is notorious for presenting translation problems (see BT, s. v. a, Mitchell and
Robinson 2001: chapter 4), because it may function as
a circumstance adverb of time then,
a circumstance adverb of space there,
an adverbial connector expressing transition,
a discourse marker whats more,
a simple subordinating conjunction indicating a temporal relation when
or may be employed as one (or both of) the items in a correlative construction
a a when then; then when (for examples, see (49) and (50)).
4
4 Obviously, there are still other homonyms, such as the Plural Nom./Acc. of the per-
sonal pronoun in the plural. This, however, is not a case of polyfunctionality, but of
homonymy.
Ambiguous adverbs/conjunctions (Group C) 65
In Old English narrative texts, a is often employed, as has been extensively
argued, as a pragmatic marker denoting foregrounded action, narrative seg-
mentation, or discourse-level shifts (see e. g. Enkvist and Wrvik 1987; Kim
1992 and the research summary in Brinton 2006: 314315).
In some narrative texts, a indeed seems to be a rather elusive marker of the
narrative sequence, i. e. a loose connector solely indicating that the narrative
is going on. This recurrent use of a is a shibboleth of Old English narrative
prose style, in particular in the sequence a followed by a verb, which is, by
contrast, virtually absent in Old English poetry (Andrew 1940: 318). See,
for example, the following passages from the Vision of Leofric (COLEOFRI,
which is classifed as a religious treatise in the Helsinki Corpus, but is clearly
not argumentative but narrative), in which every single sentence starts with a
(all of them with aV word order):
(49) a [49a] geseah he ofer a rode ane hand swylce heo bletsode; a [49b]
wende he rost t sum man hine bletsode, for am seo cyrce ws
eall folces afylled; a [49c] ns t na swa. a [49d] beheold he hit a
[49e] gyt geornor, a [49f] geseah he ealle a rode swa swutole, swylce
r nan ing beforan nre, & ws seo bletsiende hand styriende &
wendende upward. a [49g] forhtode he & tweonode him hweer hit
swa wre, swa him uhte. a [49h] mid s modes tweonunge a [49i;
adverb] teowde heo him swa swutole (COLEOFR, p.84)
Then (When?) [49a; adverb then or subordinator when, as soon as
in correlative construction] he saw over the cross a hand as if it blessed;
(then) [49b; adverb then (?second connector in the correlative con-
struction 49ab] he immediately went in order to fnd someone to bless
him, because the church was full with people; then/there [49c; adverb]
was none there. Then, when [49d; adverb then or subordinator when
in correlative construction] he studied it [the cross] then [49e; tempo-
ral adverb] even more intensely, he then [49f; adverb then (?second
connector in the correlative construction 49de] saw the complete cross
so clearly as if there had never been anything before it, and the bless-
ing hand was directing and going upwards. Then [49g; adverb] he was
frightened and he doubted whether it was so as it seemed to him. Then
[49h; adverb] against the doubt of the mind then [49i; adverb] it [the
cross] appeared so clearly .
Similar text passages are found in almost all pieces of narrative prose in
Late Old English (for the implications of this fnding for quantitative studies,
see below), just as in typologically Old English texts which were copied in
66 Connectors in Old English
the Early Middle English period (ca. 1175), such as the History of the Holy
Rood Tree. Again, all of the instances have sentence-initial a with aV word
order:
(50) a andswarede him dauid. & cw a cw e engel to him
a dyde he swa e engel him bed. a andswrode him moyses. a
andswarede him dauid & cw (CMROOD, p.6)
Then David answered him and said then the angel said to him
then he did as the angel had commanded; Then Moses answered him
Then David answered him and said .
The collection of corpus instances in Appendix B.5.1 (s. v. a) shows that
transi tional a is frequent in a wide variety of texts in Old English, but is only
attested until the frst half of the Middle English period (ME1 and ME2; in
ME2 only in one text, the Kentish Sermons, ca. 1275). This means that the
most obvious, recurrent pattern of Old English narrative structure more or less
suddenly disappeared in the middle of the thirteenth century.
5.3.2. OE eah, swaeah, (swa)eahhwere
contrast/concession
While there are no reliable criteria, such as, for instance, constituent order, for
differentiating the adverbial and the subordinator use in the cases of additive,
transitional and causal adverbial connectors or subordinators (for details on
the causals, see Chapter 9), this is different for connectors marking the rela-
tion contrast/concession. Here, the conjunction eah categorically governs
the subjunctive mood, whereas in the adverbial use of eah the verb takes the
indicative (see Quirk 1954; Mitchell 1985: 33863429). This differentiation
remains stable until the eventual loss of the subjective mood in Early Modern
English (see Molencki 1997b and Moessner 2006). Mitchell (1985: 3399) says
that there is rarely any diffculty.
Example (51) illustrates the uses of eah in the Old English adaptation
of Boethius Consolatio in King Alfreds Circle. (51a) and (51b) form a cor-
relative construction where the subordinator eah (51a; subjunctive ongite)
introduces the topic-forming pre-posed clause, and the adverbial connector
eah (51b; indicative wolde) strengthens the concessive interpretation in the
second connect. In (51c), the subordinator eah (subjunctive todlen) intro-
duces a post-posed concessive clause (for further examples, see below, Chap-
ter 10.5.2):
Position of adverbial connectors 67
(51) a andsworede se Wisdom & cw : Hwer u nu fullice ongite for-
hwy hit onne swa sie? a andswarede ic & cw: eah [51a] ic his nu
hwthwugu ongite [1 Prs. Sg. Subj.], ic wolde hit eah [51b] fullicor &
openlicor of e ongitan. a andsworode se Wisdom & cw: Genog
sweotol hit is tte God is anfeald & untodlendlic, eah [51c] hine
dysige men on mnig todlen [3 Prs. Pl. Subj.], onne hi dwoliende se-
ca t hehste god on a smran gesceafta (COBOETH, 33.74.2675.1)
Wisdom then answered and said: Do you now fully understand why it
is so? Then I answered and said: Although [51a] I understand it now in
some measure, I would nevertheless [51b] learn it more fully, and more
distinctly from you. Then Wisdom answered and said: It is suffciently
clear that God is single and inseparable, though [51c] foolish men divide
it into many, when/if they, erring, seek the highest good in the worse
creatures.
As is common for connectors expressing the semantic relation contrast/con-
cession, the adverbial connector eah expresses a contrastive rather than a
purely concessive meaning of counter-expectancy. This is induced by the equal
weight of the propositions of the two connects in such a paratactic construction
(see below, Chapter 10).
5.4. Position of adverbial connectors
5.4.1. Nacherstposition post-frst-position
Examples such as (51) show that so far we have not discussed one essential
distinguishing criterion of adverbial connectors and conjunctions, namely their
position in the sentence. One of the defning topological properties of conjunc-
tions in contrast to adverbial connectors is that they are obligatorily placed at
the beginning of their respective connects (though sometimes in collocations
following coordinating conjunctions, such as and (g)if or and though); adver-
bial connectors, on the other hand, may be placed in a variety of positions
(see above, Chapter 2.3). This criterion establishes eah in (51b) clearly as an
adverbial connector.
With respect to adverbials in general (i. e. not only adverbial connectors,
but also circumstance and linking adverbials), there has not been much
change in the history of English. Old English grammars generally state that
the free variation available to lfric in the position of the adverbs is avail-
able today likewise. So much depends on the writers purpose (Mitchell
68 Connectors in Old English
1985: 15921593). This means that the distribution of adverbials found for
Present Day English (see Chapter 3.5.4) is similar in Old English, at least as
far as the position of circumstance adverbials is concerned. There have, how-
ever, been drastic changes in the position of linking adverbials, since these
are from the Late Modern English period onwards increasingly found in
connect-medial and connect-fnal positions (for the details, see below, Chap-
ters 10.5 and 13).
To understand these changes, it is necessary to comment on one specifc
clause-internal position of adverbial connectors, namely the so-called Nach-
erstposition (Pasch et al. 2003: 71). For lack of a better expression, the present
study uses the loan-translation post-frst-position for this position of an item
after a sentence constituent which may fll the pre-feld on its own.
5
This position of adverbial connectors has, as far as I am aware, not been
discussed in the literature on English (neither historically nor synchronically)
in any detail until recently. In 2005 and 2006, van Kemenade has (with vari-
ous collaborators) presented a number of papers on one aspect of this posi-
tion, namely the use of the adverbs a/onne as what these linguists call fo-
cus particles (the only publication available so far is van Kemenade and Los
2006b). While I agree with the central claim for Old English, namely that in
its clause-internal use a and onne (but not only these!) are morpho-syntactic
markers which separate the topic domain (i. e. the given information) of the
clause from the focus domain (i. e. the new information), my fndings differ
with respect to the implications for the later periods of English, in which this
possibility of position has certainly not, as van Kemenade claims, disappeared,
but is still very much alive in the same functions as in Old English (it is only
OE a which is no longer used after the middle of the thirteenth century; see
above, Chapter 5.3.1). Because of these different views, I will frst of all intro-
duce the discourse function of this position of adverbial connectors with Pres-
ent Day German examples.
5.4.2. Present Day German adverbial connectors in post-frst-position
While most of the adverbial connectors in German may be placed rather
freely between the two connects (Nullposition) or in the pre-feld and
5 See also the refned defnition in grammis (s. v. Nacherstposition): Als Nacherst-
position bezeichnen wir eine topologische Position im Vorfeld nach der ersten
Konstituente und vor der linken Satzklammer, die von bestimmten Adverbkonnek-
toren besetzt werden kann.
Position of adverbial connectors 69
middle feld, their use in post-frst-position is highly dependent on the con-
text. In grammis (s. v. Zur Nacherstposition bei nicht positionsbeschrnk-
ten Adverbkonnektoren), the discourse restrictions are illustrated by the
following examples:
(52) Rehmann will sich nicht lnger den Schwarzen Peter zuschieben lassen.
Er packte also aus. [Mannheimer Morgen, 30.5.1995]
Rehmann does no longer want to be blamed. So / therefore he blew the
whistle.
(52a) Rehmann will sich nicht lnger den Schwarzen Peter zuschieben
lassen. Also packte er aus. [Nullposition; zero position]
(52b) Rehmann will sich nicht lnger den Schwarzen Peter zuschieben
lassen. *Er also packte aus. [post-frst-position impossible]
(52c) Rehmann war der einzige, der zu einer Aussage bereit war. Er
also packte aus [post-frst-position possible].
The comparison of examples (52b) and (52c) shows that a post-frst-position
of the adverbial connector predominantly encodes discourse relations: it is
only possible if the element in the frst position is in the (contrastive) focus
(52c). In co-texts such as (52b), where the pronoun er is purely anaphor-
ic, but not in the contrastive focus, the post-frst-position of the adverbial
connector is not allowed. This contrastive function of adverbial connec-
tors in post-frst-position also explains why most of the adverbial connec-
tors which may take this position express the semantic relations contrast
(e. g. PDG aber, allerdings, dagegen, hingegen) or result (e. g. PDG also,
schlielich).
(53) Die weiblichen Selbstoffenbarungen, so mutmat Kotthoff, frdern den
Gemeinsinn. Mnner hingegen meiden solche Entblungen.
Female self-relevations, Kotthoff surmises, endorse the public spirit.
Men, on the other hand, avoid such exposures.
(54) Allein im Groraum Hannover hat sich die Zahl der Szene-Projekte in-
nerhalb von 3 Jahren verdoppelt. In der Nrnberger Region allerdings
konnte sie nicht zulegen.
Only in the Greater Hanover area the number of scene-projects has
doubled in three years. In the Nuremberg area, on the other hand, there
has been no increase.
70 Connectors in Old English
5.4.3. Post-frst-position of adverbial connectors in Old English
As has been mentioned above, the functions and implications of adverbial con-
nectors in post-frst-position have only rarely been discussed in the literature.
In his chapter on r- and onne-sentences, Andrew, for instance, says:
Here we also simplify our problem by excluding sentences in which onne
(adv.) is not the head word (Andrew 1940: 21).
In some more detail, Mitchell (1985) comments on the diverse uses of a and
onne as conjunctions and adverbs:
It is a syntactical commonplace that, whereas a conjunction is the equivalent of
Modern German als, being used only with the preterite indicative of a completed
act in the past or of a series now regarded as a single act, onne is the equivalent of
wenn, being used frequentatively in the past, present and future, and of a single act
yet to be completed at some indefnite time (Mitchell 1985: 2562).
In his chapters on the adverbs, he then differentiates between the purely tem-
poral and the resultive use of onne in particular.
. The distinction between the conjunctions a and onne set out in 2562 seems
to hold generally for the adverbs when they are used in a purely temporal sense
When used in other than purely temporal senses, onne adv. can be accompanied
by a past tense . where it is not only correlative with frequentative onne, but
also marks a stage in the narrative where it implies a qualifcation or contrast
(Mitchell 1985: 11161117).
Mitchell (1985) does not comment on the constituent order of the respective
uses, but it becomes clear from my corpus texts that there is a high correlation
of the use of resultive onne in post-frst-position (not necessarily with the verb
in the past tense) and its inferential use:
(55) Ac t is openlice cu t sio godcunde foretiohhung is anfeald & un-
andwendlic, & welt lces inges endebyrdlice, Sumu ing onne on
isse weorulde sint underied re wyrde, sume hire nanwuht under-
ied ne sint (COBOETH, 39.129.712)
But it is evidently known that the divine predestination is simple, and
unchangeable, and governs everything according to order Some
things, therefore, in this world are subject to fate, others are not at all
subject to it.
Often, onne is found with imperatives, such as in
(56) Sumra wyrta oe sumes wuda eard bi on dunum, sumra on merscum,
sumra on morum, sumra on cludum, sumra on barum sondum. Nim
Position of adverbial connectors 71
onne swa wuda swa wyrt, swa hwer swa u wille, of re stowe e
his eard & elo bi (COBOETH, 34.91.1619).
Of some herbs, or some wood, the native soil is on hills, of some in
marshes, of some on moors, of some on rocks, of some on bare sands.
Take, therefore, tree or herb, whichever you will, from the place which
is its native soil .
This use of onne is not only found in translations from Latin, where it may
have been triggered by Latin enim and nam, which are commonly used in this
position, but also in more independent texts such as the Blickling Homilies:
(57) a halgan r Cristes cyme on hine gelyfdon, & hine lufodan, & mid
his rowunga hie wurdan alesde of helle wite, & mid his riste ge-
hlde. We onne synt e r fter fylgea; (COBLICK, Homily 2,
285288).
The saints believed in him before he had come and loved him and
through his sufferings they were saved from hell and (he) healed them
by his resurrections. We, then, are those who will follow afterwards
.
In Old English, this position is widely attested for a number of adverbial con-
nectors (fourteen out of thirty-two adverbial connectors; see the sub-entries
post-frst-position in Appendix B for the respective connectors). As in Pres-
ent Day German, it is predominantly found with resultive (see onne) or con-
trastive adverbial connectors, such as swa eah in (58) or eah in (59b):
(58) tforan lcum cwyde we setton a swutelunge on leden. mg swa
eah se e wile a capitulas fter re foresprce geendebyrdian
(COAEPREF, R 2.41)
In front of each homily we have put the argument in Latin; everyone
who wants to, however, may order the rubrics after the preface.
(59) & eah [59a; Conj.] t cild for geogoe sprecan ne mage onne hit
man fulla, his freonda foresprc forstent him eal t ylce e hit sylf
sprce. Is eah [59b] ma manna onne earf wre e ises behates
gescad ne cunnan (COWULF3, p.227228).
And although [59a] the child cannot speak because of its youth when
it is baptized, his friends intercession is equivalent to all that it himself
speaks. There are, however [59b], more men than would be necessary
who do not know the meaning of this vow.
72 Connectors in Old English
There can be no doubt that this post-frst-position of adverbial connectors is
triggered by discourse considerations. As has been mentioned above, however,
van Kemenade (2006b) goes one step further and argues the cue status of ad-
verbs such as a and onne in clause structure (see the summary in van Ke-
menade and Los 2006b: 231). Her main observation is that subject personal
pronouns appear on the left of a or onne (object pronouns do so optionally),
whereas nominal elements, including subjects, occur on the right. Other re-
strictions are that defnite subjects are used on the left of a or onne, and that
indefnite subjects are quite rare there (van Kemenade and Los 2006b: 232). In
Old English, the crucial clues are thus defniteness/anaphoricity, contrastive
topic status and prominence of discourse referent.
In van Kemenades view, the transition to Middle English marks a sharp con-
trast because the use of a and onne in clause-internal position becomes a good
deal less frequent in absolute numbers and this, according to van Kemenade
and Los (2006b), indicates that a/onne were in the process of losing their
discourse-marking properties, in particular because defniteness/anaphoricity
and contrastive topic status are thought to no longer provide the crucial clue for
appearance in this area of discourse. As the central implication, they fnd that
the syntactic organization of the clause, at least in Old English, is interwoven with
discourse organization much more closely than has been thought so far, and that the
transition to Middle English is one that results in a more strictly syntactic organiza-
tion of the clause (van Kemenade and Los 2006b: 224).
In view of the fndings of the present study, however, we see that this pattern
did not die out at the end of the Old English period (see, as a frst indicator, its
existence in the translations of the example sentences above): post-frst posi-
tions have been common in all periods of English in this specifc, focussing
discourse function, in particular for resultive then (see Appendix B.3.2; see, for
example, also the uses of therefore, which was only coined after the Old Eng-
lish period). Considering all kinds of adverbial connectors, major changes in
frequency and variability have, by contrast, occurred with respect to the medial
and fnal position of adverbial connectors from the Late Modern English period
onwards (see below, Chapter 13).
5.5. Pronominal connectors
5.5.1. Pronominal connectors in Old English
The connectors discussed so far have mainly been single lexical items which
are used as explicit markers of sentence or discourse connection, i. e. are situ-
Pronominal connectors 73
ated on levels III (Explizit verknpfte Hauptstze Explicitly linked main
clauses) or IV (Verknpfung durch subordinierende Konjunktionen Link-
age by subordinating conjunctions) of Raibles typological continuum of con-
nectors (Raible 1992). In Raibles continuum, which is based on compara-
tive synchronic data of the Present Day Romance languages but also takes
the diachronic perspective from Latin to todays French-based creoles into
account, there is a phase II situated between asyndesis (I) and the explicit
paratactic connection (III) which is important for the history of English con-
nectors. This phase is called Junktion durch Wiederaufnahme (eines Teils)
der vorhergehenden Stze and singles out anaphoric connectors which com-
prise an explicitly text deictic, pronominal element, such as OE dative m in
form therefore because and instrumental on or y in foron or fory
therefore because (cf. Present Day German forms with genitive des of the
article in des-wegen therefore; lit. this-for, des-halb therefore; lit. this-for or
dative dem of the article in dem-nach therefore; lit. this-after).
These Old English pronominal connectors (Group D), like the use of case
forms of the demonstratives in resultive s and y therefore, will be dis-
cussed in more detail below in Chapter 9. For a better understanding of the loss
of polyfunctionality and the diachrony of adverbial connectors in general, it is,
however, helpful to follow the developments in the Romance languages, from
Latin to French-based creoles.
5.5.2. Pronominal connectors in the history of the Romance languages
While in Present Day French these anaphoric connectors mostly comprise a
form of the article or demonstrative in a full nominal phrase as, for instance,
cause de cela because of this, cest pourquoi this is why or pour cette
raison for this reason Raible shows in his diachronic chapter (Chapter IV;
Raible 1992: 154190) that many originally anaphoric forms have developed
into opaque conjunctions (2. Von der pronominalen Reprise oder Vorweg-
nahme und vom relativen Anschluss zu Konjunktionen (Latein, romanische
Sprachen); Raible 1992: 160170).
In Latin, these forms are manifold and allow anaphoric as well as cataphoric
relations. For the comparison with the development of the causal connectors
in English, I will here illustrate the system by causal pronominal connectors
in Latin (the respective pronominal element is underlined): for the semantic
relation cause result, there are anaphoric adverbial connectors such as
propterea, eapropter, propter id/hoc, quapropter, ob id/hoc, pro eo, or hac
gratia therefore (for the full list, see Raible 1992: 163164). These forms can
74 Connectors in Old English
also be used cataphorically in pre-posed clauses in collocations with quod,
i. e. propter ea quod, eapropter quod because. The forms comprising quod
are then differentiated and yield the widely used forms of conjunctions with a
second element que in the Romance languages, such as French puisque, parce
que, avant que or Spanish pues que, porque; there, que functions as a general
subordinating particle similar to PDG dass, OE e or ME and PDE that (cf.
now vs. now that).
Two subtypes are to be distinguished: (a) correlative constructions such as
French par ce que (with demonstrative ce originally in the frst connect), and
(b) pour que without a correlative (for a detailed analysis see Stempel 1964,
Krefeld 1989). With loss of transparency, conjunctions may move from type (a)
to type (b) when the originally anaphoric element is fused, as in Rhaeto-Ro-
manic cura cha as which has developed out of qua hora cha the hour that.
5.5.3. Connectors in Louisiana French and French-based creoles
For a better understanding of the processes described in the following chapters
for the history of English adverbial connectors, it is illuminating to follow the
path of these transparent conjunctions, which are structurally identical to, for
example, OE for m e (cf. French pour ce que), in franais cadien (Louisi-
ana French) and in French-based creoles of the Caribbean.
With respect to the subordinating particle, French has a tendency to oblig-
atorily use the subordinating particle que, whereas Louisiana French shows
exactly the diverging tendency: the subordinating particle que is optionally
deleted so that parce que appears as parce (Raible 1992: 167169, 200202).
This has drastic consequences for the hierarchy of connectors.
Da es im franais cadien nicht, wie im Deutschen, eine besondere Nebensatz-S-
tellung gibt, und auch der Modus Konjunktiv als Signal der Integration fast
ganz verschwunden ist, bedeutet dies im Grunde, auf die kontinental-franz sischen
(schriftsprachlichen) Integrations-Techniken bezogen, ein Zusammenfallen der
Techniken III und IV zu einer einzigen (Raible 1992: 200).
Since there is, in contrast to German, no specifc word order for subordinate claus-
es in franais cadien and since also the subjunctive mood as a signal of inte-
gration has been lost almost completely, this basically means, when compared to
the integration patterns of (written) continental French, that levels III and IV have
merged into a single level.
This tendency is taken even further in some Caribbean French-based creoles. In
contrast to Louisiana French, where the deletion is optional, puisque has there
developed into non-transparent pis; similarly, parce que appears as non-trans-
Pronominal connectors 75
parent pas (Raible 1992: 201). This means that parce que has lost both of its
deictic elements ce and que, a process identical to the development of form
e into for in the history of English. Yet this does not mean that these creoles
have only simplifed their inventory of connectors, since they at the same time
coin new connectors from lexical material or from recurrent syntactic phrases,
though only rarely as pronominal connectors (Raible 1992: 201202).
6. Adverbial connectors in the history of English
6.1. Introductory remarks
Based on this summary of connectors in Old English and on the corpus fnd-
ings for contemporary English (Chapter 3.5), the history of the adverbial con-
nectors in general will now be analysed from various perspectives, starting
with a survey of the connectors attested in the periods from Old English on-
wards. In all of these sections, the fndings will be compared with the relations
to and the developments of the other sentence and discourse connectors, i. e. of
coordinating and subordinating conjunctions.
6.2. The diachrony of coordinating and subordinating conjunctions:
general tendencies
In contrast to subordinators and adverbial connectors, the number and also the
functions of the coordinating conjunctions relevant for the present study, i. e.
those working on the sentence or discourse level, have remained fairly stable
in the history of English. With the exception of the emergence and disappear-
ance of recursive for as a rather loose additive connector (discussed below in
Chapter 9), the only major change is the replacement of lexical material from
the Early Middle English to the Late Middle period in the cases of ac by but
and oe by the reduced form or:
1
additive relation (Appendix B.1.5): OE to PDE: and
ME to LModE: for
contrastive relation (Appendix B.4.4): OE to ME2: ac; from ME3: but
reformulatory relation (Appendix B.4.4) OE oe; from ME3 reduced
form or
The scenario is rather different for subordinators and adverbial connectors. The
diachrony of English subordinators has been analyzed in a grammar-cum-dic-
tionary-method (Kortmann 1997: 53) by Kortmann in the wider context of the
typologies of adverbial subordinators (Kortmann 1997: 291335). This study
has yielded the following general tendencies, which are used as a contrastive
1 The OED sees er, the full form of reduced or, as a variant of oe, oa; see
OED, s. v. other.
Introductory remarks 77
plane for the present investigation of the development of adverbial connectors.
Over the periods, Kortmann fnds
an increase in the number of subordinators,
a decrease in their syntactic polyfunctionality,
a decrease in their semantic polyfunctionality and
an increase in their morphological complexity.
With regard to the individual semantic relations he fnds that
subordinators for the relation contrast/concession are coined much later
than for other relations and
that the inventory of concessive subordinators is least stable over time.
As for the respective periods (Kortmann distinguishes between OE, ME,
EModE and PDE only, i. e. does not have a separate category for Late Modern
English), Middle English emerges as the crucial period for shaping the inven-
tory of Present Day English subordinators. A huge number of subordinators
were also coined in the Early Modern English period, but most of these innova-
tions were ephemeral. Since most of these Early Modern English coinages have
not survived into Present Day English, this period despite its high number of
new subordinators has had only a weak impact on Present Day English (for
the details, see Kortmann 1997: 334335 and below, Figure 6.3).
6.3. The diachrony of adverbial connectors: general tendencies
In a frst step, these general tendencies found by Kortmann (1997) for subordi-
nators will now be tested for the adverbial connectors. In spite of the fuzziness
of the category of adverb, the polyfunctionality or heterosemy of lexemes and
the gradient character of adverbial subclasses, it is still possible to compile a
rather full list of adverbial connectors for the respective sub-periods in the his-
tory of English (with, of course, the exception of the enumerative adverbs, such
as fourthly, ffthly, etc.; for this open class, see above, Chapter 3.3).
This inventory of English adverbial connectors is summarized in the synop-
sis of the present study Appendix A which lists all single adverbs as well as
lexicalized verbal and prepositional phrases which have been used as adverbial
connectors in the various periods of English (i. e. only fully transparent phrases
such as for that reason or on the contrary are excluded).
2
2 For the principles of exclusion, i. e. the properties a prepositional phrase has to fulfl
in order to be included, see below, Chapter 7.4.
78 Adverbial connectors in the history of English
The present qualitative and quantitative analysis starts with a very general
description and evaluation of the developments in the feld of English adverbial
connectors over the periods, which allows a frst comparison of the similarities
and divergences in the development of adverbial connectors and subordina-
tors. In the following chapters, these fndings are then examined from various
perspectives in order to demonstrate and explain the systematic changes in the
morphological make-up and the use of cognitive source domains for the coin-
age of new adverbial connectors (Chapters 7 and 8).
The differing developments in the individual semantic relations then show
that changes in the impure connectors the so-called ccc-relations refect
typological changes of English (Chapter 10 on causal and resultive connec-
tors, Chapter 11 on concessive and contrastive connectors), while the changes
in the pure connectors various additive and transitional relations are
mainly triggered by rhetorical and pragmatic motivations (see Chapters 11
and 12 on additive and listing connectors and Chapter 8.4 on truth-intensifers
which are used as transitional connectors). Rhetorical and pragmatic con-
siderations also elicit changes in constituent order, i. e. the increasing use of
adverbial connectors in medial position in the written medium and in fnal
position in the spoken medium (discussed in an exemplary analysis of PDE
though in Chapter 10.5). These will in turn allow more detailed assessments
on the hows and whys of the diachrony of adverbial connectors in the history
of English.
6.4. New adverbial connectors in the sub-periods of English:
methodology
The diachrony of English adverbial connectors is summarized in Table 6.1,
which frst of all lists the repertoire of the Old English items which could
function as adverbial connectors. As has been shown above, most of these
could also be used as subordinating conjunctions, without any change in their
form or, very often, also without any change in the form of their connects
(for constituent order or the use of the subjunctive mood as distinguishing
criteria, see above, Chapter 5.3). Following Mitchell (1985: 2536, 3010),
these items are called ambiguous adverb/conjunction and are marked by
the symbol .
In chronological order, Table 6.1 then presents all the items which have
emerged as new adverbial connectors in the respective sub-periods of English.
In the left column, only those innovations are listed which have survived until
Present Day English. The adverbial connectors in the right column died out
New adverbial connectors in the subperiods of English: methodology 79
in some period before Present Day English (e. g. herefore, forwhi, overmore)
or have lost their capacity to function as adverbial connectors (e. g. albeit or
wherefore).
3
The sections after the Old English period thus only list the new adverbial
connectors, i. e. those which emerged in the respective period in this func-
tion. The period ME1/2 (11501350) shall serve to illustrate this methodo-
logy: The innovations attested in this period are certain, certainly, certes, for,
forsoothe, herefore, for that, forhwi, more, neverthelatter, on oer half, oer
side, overall, therefore, yet and wherefore. Some of these are already recorded
for Old English (e. g. giet yet, forhwi forwhy), but had not been used as
adverbial connectors before. OE giet yet, for example, was predominantly
used as a temporal adverb in Old English (still, yet; see BT, s.vv. git, gita, and
CH, s. v. get), a use which has survived into Present Day English (implying
continuance from a previous time up to and at the present time: see OED, s. v.
yet, 2.5).
4
From Early Middle English onwards (see the date 1205 for its frst
occurrence in this function in Appendix A), it has by a conventionalization
of conversational implicatures common for temporal adverbs expressing con-
comitance (cf. PDE still and see below, Chapter 10.5.3.2) acquired contras-
tive/concessive meanings and is used as an adverbial connector nevertheless.
Since it could also be used as a subordinator yet (that) notwithstanding that,
although during the Middle English period (OED; s. v. yet, III.9b), it is clas-
sifed as an ambiguous adverb/conjunction (indicated by the symbol ). Yet
has survived in its adverbial connector function until the Present Day English
period, as is attested by the wide variety of texts and collocations with other
concessive/contrastive connectors such as but, nevertheless, or still (see in Ap-
pendix B.4.1).
As a second illustration, compare OE for-hwy, which was only used as an
interrogative adverb introducing a direct question: for what reason?, why?,
or as a conjunction introducing a clause in apposition to or exemplifying the
statement in the principal clause: (as to) why , in that, in the fact that in Old
English (DOE, s. v. for-hwm, for-hwon, for-hwy, A.1, B.1). In Early Middle
English, it began to be used as a resultive adverbial connector therefore (see
the date 1225 in Appendix A.1 and the frst occurrence in the corpus texts in
3 For the exact periods, see Appendix A.1 and the Index of Adverbial Connectors
(Appendix A.2), and also the discussion of the respective semantic relations.
4 Only in the collocation a giet furthermore is it attested as a reinforcing adverbial
connector. This meaning in addition, or in continuation; besides, also; further, fur-
thermore, moreover; with a numeral or the like = more is present in, for example,
PDE yet another (see OED, s. v. yet, 1.).
80 Adverbial connectors in the history of English
Appendix B.3.1 in CMHALI). Appendices A.1 and B also provide evidence
that it was used in this adverbial connector function only during the Middle
English period, i. e. it is not attested after ME4 (in CMHILTON).
The lexicalization of collocations and prepositional phrases in a specifc pe-
riod can be exemplifed by ME on oer half and ME oer side. While the nouns
healf half and side side, part are found in Old English, they are not recorded in
the lexicalized phrases on oer half or oer side on the other hand as adverbial
connectors marking the contrastive/antithetic relation (see Appendix B.4.2).
5
For another kind of change in function see ME forsooth, which is employed as
a transitional adverbial connector in Middle English (in ME3 and ME4; see Ap-
pendix B.1.5), but is only sporadically and jocularly used as an interjection in con-
temporary English (for a full analysis, see Lenker 2003 and below, Chapter 8.4.5).
6.5. The inventory of Present Day English connectors: donor periods
The data collected in Table 6.1 allow a survey of the long-time developments of
adverbial connection in English and thus a comparison with Kortmanns data
for subordinators (Kortmann 1997).
At frst glance, two fndings are fairly obvious: First, very few (seven out of
thirty-two) of the Old English items have survived. Among them are the still
very frequent additive/reinforcing also and the resultive/transitional so and
then, but also the fairly marginal else (reformulatory), here and now (both tran-
sitional). There is no adverbial connector with an uninterrupted history from
Old English for the ccc-relations cause/result and contrast/concession (for
the meandering path of though, which is not attested from Middle English to
Late Modern English, see Appendix B.4.1 and below, Chapter 10.5). This means
that the system of clause linkage by adverbial connectors was almost completely
re-organized from the Middle English period onwards.
Secondly, new adverbial connectors have been coined in all of the periods of
English until Present Day English and so as a general tendency the absolute
number of types has increased over the periods. Table 6.2, which lists the num-
ber of adverbial connectors per sub-period, shows that the number of types has
increased from thirty-two adverbial connectors in Old English (including the
ambiguous adverbs/conjunctions) to ffty-nine in Present Day English.
5 On oer half is only attested in two texts from the earliest Middle English period
(ME1; all of them are texts from the so-called Katherine Group found in Oxford,
Bodleian Library, Bodley 34); oer side re-appears in Early Modern English in the
full phrase on the other side.
The inventory of Present Day English connectors: donor periods 81
T
a
b
l
e

6

:

A
d
v
e
r
b
i
a
l

c
o
n
n
e
c
t
o
r
s

i
n

t
h
e

h
i
s
t
o
r
y

o
f

E
n
g
l
i
s
h
S
u
r
v
i
v
o
r
s

t
o

P
r
e
s
e
n
t

D
a
y

E
n
g
l
i
s
h
L
o
s
t

o
n

t
h
e
i
r

w
a
y

t
o

P
r
e
s
e
n
t

D
a
y

E
n
g
l
i
s
h
O
E
O
E

e
a
l
l
-
s
w
a

a
l
s
o

,

e
l
l
e
s

e
l
s
e

,

h
e
r

h
e
r
e

n
u

n
o
w

s
w
a

s
o

e
a
h

t
h
o
u
g
h

o
n
n
e

t
h
e
n

r
e
s
t

f
r
s
t

,

e
a
c

a
l
s
o

e
a
c

s
w
y
l
c
e
/
s
w
y
l
c
e

e
a
c

a
l
s
o

,

c
u

l
i
c
e

t
h
e
r
e
f
o
r
e

(
l
i
t
.

c
e
r
t
a
i
n
l
y

)
,

e
f
t

a
l
s
o

,

e
f
t
s
o
n
a
/
e
f
t
s
o
n
e
s

a
l
s
o

,

e
o
r
n
o
s
t
l
i
c
e

(
l
i
t
.

e
a
r
n
e
s
t
l
y

;

t
r
a
n
s
i
t
i
o
n
)
,

f
o
r

t
h
e
r
e
f
o
r
e

f
o
r

o
n

t
h
e
r
e
f
o
r
e

f
o
r

t
h
e
r
e
f
o
r
e

,

h
u
r
u

n
e
v
e
r
-
t
h
e
l
e
s
s

,

h
u
r
u

i
n
g
a

n
e
v
e
r
t
h
e
l
e
s
s

,

n
a

y
l

n
e
v
e
r
t
h
e
l
e
s
s

h
w

w
h
a
t

t
h
e
n

(
t
r
a
n
s
i
t
i
o
n
)
,

h
w

e
r
e

n
e
v
e
r
t
h
e
l
e
s
s

,

s
o

l
i
c
e

(
l
i
t
.

t
r
u
l
y

;

t
r
a
n
s
i
t
i
o
n
)
,

s
w
a

e
a
h

n
e
v
e
r
t
h
e
l
e
s
s

,

s
w
a

e
a
h
h
w

r
e

n
e
v
e
r
t
h
e
l
e
s
s

t
h
e
n

(
t
r
a
n
s
i
t
i
o
n
)
,

a

g
i
e
t

a
l
s
o

,

(

r
)
t
o
e
a
c
a
n

a
l
s
o

s

(
d
e
m
o
n
s
t
r
a
t
i
v
e
;

g
e
n
i
t
i
v
e
)

t
h
e
r
e
f
o
r
e

e
a
h
h
w

e
r
e

/

h
w

r
e

e
a
h

n
e
v
e
r
t
h
e
l
e
s
s

y

(
d
e
m
o
n
s
t
r
a
t
i
v
e
;

i
n
s
t
r
u
m
e
n
t
a
l
)

t
h
e
r
e
f
o
r
e

,

g
e
w
i
s
l
i
c
e

(
l
i
t
.

c
e
r
t
a
i
n
l
y

;

t
r
a
n
s
i
t
i
o
n
)
,

w
i
t
o
d
l
i
c
e

(
l
i
t
.

t
r
u
l
y

;

t
r
a
n
s
i
t
i
o
n
)
M
E

1
/
2
M
E

1
/
2
c
e
r
t
a
i
n
l
y
,

f
r
s
t
,

f
o
r
,

a
t

(
t
h
e
)

l
a
s
t
,

m
o
r
e
,

o
v
e
r

a
l
l
,

t
h
e
r
e
f
o
r
e
,

y
e
t
c
e
r
t
a
i
n
,

c
e
r
t
e
s
,

f
o
r
s
o
o
t
h
,

h
e
r
e
f
o
r
e
,

i
w
i
s
,

f
o
r

t
h
a
t
,

f
o
r
h
w
i
,

n
e
v
e
r
t
h
e
l
a
t
t
e
r
,

n
o
u
g
h
t
-
f
o
r
-
t
h
a
n

n
e
v
e
r
t
h
e
l
e
s
s

,

o
n

o

e
r

h
a
l
f
,

o

e
r

s
i
d
e
,

w
h
e
r
e
f
o
r
e
M
E

3
M
E

3
f
u
r
t
h
e
r
,

f
a
r
t
h
e
r
m
o
r
e
/
f
u
r
t
h
e
r
m
o
r
e
,

m
o
r
e
o
v
e
r
,

n
e
v
e
r
t
h
e
l
e
s
s
,

o
t
h
e
r
w
i
s
e
,

t
h
u
s

a
l
g
a
t
e
s
,

e
r
g
o
,

f
o
r

w
h
i
c
h
,

f
a
r
t
h
e
r
o
v
e
r
/
f
u
r
t
h
e
r
o
v
e
r
,

i
t
e
m
,

n
o
u
g
h
t
-
f
o
r
-
t
h
a
t

n
e
v
e
r
t
h
e
l
e
s
s

n
o
u
g
h
t
-
f
o
r
-
t
h
i

n
e
v
e
r
t
h
e
l
e
s
s

,

o
v
e
r
m
o
r
e
,

o
v
e
r

t
h
i
s
/
t
h
a
t
,

o
t
h
e
r
w
a
y
s
,

s
e
k
i
r
l
y
,

s
u
i
n
g
l
y

c
o
n
s
e
q
u
e
n
t
l
y

,

t
r
u
l
y
,

v
e
r
i
l
y
M
E

4
M
E

4
c
o
n
t
r
a
r
i
w
i
s
e
,

f
n
a
l
l
y
a
f
t
e
r

m
o
r
e
o
v
e
r

h
o
w
b
e
i
t
,

n
o
t
w
i
t
h
s
t
a
n
d
i
n
g
E
M
o
d
E
1
E
M
o
d
E
1
a
g
a
i
n
,

c
o
n
s
e
q
u
e
n
t
l
y
,

c
o
n
t
r
a
r
i
l
y
,

l
i
k
e
w
i
s
e
,

n
o
n
e
t
h
e
l
e
s
s
,

s
u
r
e
l
y
,

w
e
l
l

a
l
b
e
i
t
,

i
n

f
n
e
,

h
o
w
s
o
m
e
v
e
r
,

l
i
k
e
w
a
y
s
,

o
v
e
r

a
n
d

b
e
s
i
d
e
s
,

s
e
m
b
l
a
b
l
y
E
M
o
d
E
2
E
M
o
d
E
2
a
b
o
v
e

a
l
l
,

b
e
s
i
d
e
s
,

h
o
w
e
v
e
r
,

i
n
d
e
e
d
,

a
t

l
e
a
s
t
,

(
i
n

t
h
e
)

m
e
a
n
t
i
m
e
,

(
i
n

t
h
e
)

m
e
a
n
w
h
i
l
e
,

i
n

s
u
m
,

s
u
r
e
b
e
s
i
d
e
,

c
o
n
t
r
a
r
y
,

c
o
n
t
r
a
r
i
w
a
y
s
,

h
o
w
s
o
e
v
e
r
,

l
a
s
t
l
y
,

o
v
e
r

a
n
d

a
b
o
v
e
E
M
o
d
E
3
E
M
o
d
E
3
a
f
t
e
r

a
l
l
,

o
n

t
h
e

o
t
h
e
r

h
a
n
d
,

h
e
n
c
e
,

s
t
i
l
l
t
h
e
n
c
e
,

v
i
d
e
l
i
c
e
t
,

w
h
e
n
c
e
L
M
o
d
E
1
L
M
o
d
E
1
a
c
c
o
r
d
i
n
g
l
y
,

a
t

t
h
e

s
a
m
e

t
i
m
e
,

t
o
o
i
n

t
r
u
t
h
L
M
o
d
E
2
a
l
t
o
g
e
t
h
e
r
,

a
n
y
h
o
w
,

c
o
n
v
e
r
s
e
l
y
,

o
f

c
o
u
r
s
e
,

a
t

a
l
l

e
v
e
n
t
s
,

i
n

f
a
c
t
L
M
o
d
E
3
a
l
l

t
h
e

s
a
m
e
,

i
n

a
n
y

c
a
s
e
,

a
n
y
w
a
y
,

a
t

a
n
y

r
a
t
e
,

i
n

t
h
e

s
a
m
e

w
a
y
P
D
E
a
l
l

i
n

a
l
l
,

c
o
r
r
e
s
p
o
n
d
i
n
g
l
y
,

i
n
c
i
d
e
n
t
a
l
l
y
,

i
n
s
t
e
a
d
,

p
l
u
s
,

r
a
t
h
e
r
,

s
i
m
i
l
a
r
l
y
82 Adverbial connectors in the history of English
Table 62: Absolute number of adverbial connectors per sub-period (types)
O
E
M
E
1
/
2
M
E
3
M
E
4
E
M
o
d
E
1
E
M
o
d
E
2
E
M
o
d
E
3
L
M
o
d
E
1
L
M
o
d
E
2
L
M
o
d
E
3
P
D
E
32 27 36 29 34 45 46 48 52 56 59
A steady increase of adverbial connectors is, however, only attested from the
Early Modern English period onwards. During the Middle English period, the
absolute number of adverbial connectors did not change considerably, when
we compare the beginning (M1/2 twenty-seven connectors) with the end of the
period (ME4 twenty-nine connectors). There was, however, a sharp increase in
the middle of the period (with thirty-six connectors in ME3).
This general tendency of a constant increase of adverbial connectors though
only after the Early Modern English Period is supported by the analysis of
the inventory of adverbial connectors in the selected corpus texts which were
chosen for the comparative quantitative analysis (see Appendix C.2).
Table 63: Absolute number of adverbial connectors as attested in the corpus texts
(types)
O
E
M
E
1
/
2
M
E
3
M
E
4
E
M
o
d
E
1
E
M
o
d
E
2
E
M
o
d
E
3
L
M
o
d
E
1
L
M
o
d
E
2
L
M
o
d
E
3
17 18 30 24 33 33 26 31 38 35
This table shows that the authors of the homilies and religious treatises which
are investigated as corpus texts use a wider variety of types of adverbial con-
nectors from the Early Modern English period onwards. Yet, the differences
between authors of ME3 (e. g. Chaucer, Purvey, the authors of the Wycliffte
Sermons) and LModE3 (e. g. Swift, Hume) are not as striking as we might
have expected (thirty vs. thirty-one different connectors). This refects the fact
that adverbial connectors in particular the so-called impure connectors
coding the ccc-relations belong to a closed class of items serving a function
at the grammatical-lexical interface. It may, however, also be indicative of the
fact that the authors of the respective periods though there is an increasing
variety of connectors to choose from just like todays authors prefer certain
adverbial connectors over others (in Present Day English, for instance, authors
of academic prose prefer either thus or hence for the causal relation; see
Biber et al. 1999: 889).
The inventory of Present Day English connectors: donor periods 83
Figure 61: Number of adverbial connectors per sub-period: absolute number of
types in sub-period and in corpus texts C.2.
The diagram in Figure 6.1, which summarizes the fndings sketched above,
suggests a continuous, steady path of innovations in the increase of adverbial
connectors. When we have a closer look at the donor periods of the inventory
of Present Day English connectors, we see that there is indeed no single period
which could be said to have supplied the bulk of the adverbial connectors of
todays inventory.
Figure 62: Donor periods of Present Day English adverbial connectors
These fndings are intriguing when we compare the development of adver-
bial connectors with the diachronies of coordinators and subordinators. As has
been pointed out above (Chapter 6.2), there has been apart from the replace-
ment of lemmas virtually no change in the inventory of coordinators.
84 Adverbial connectors in the history of English
With respect to the subordinators, my fndings for adverbial connectors and
those by Kortmann (1997) for subordinators are not fully comparable because
Kortmann does not distinguish a separate category for Late Modern English.
Yet a comparison of the percentages of the donor periods for the connectors
forming the Present Day English inventory nonetheless yields some interest-
ing results: Kortmann could show, as has been summarized above, that the
crucial period for subordinators is the Middle English period, contributing ca.
forty-seven per cent of the subordinators of Present Day English, as contrasted
with ca. twenty-fve per cent from Old English and ca. thirteen per cent from
Early Modern English (Kortmann 1997: 334335). For the adverbial connec-
tors, it is not possible to establish such a single crucial period. The infux of
new, surviving connectors seems to be much more stable over time (see the
graph in Figure 6.3).
Figure 63: Percentages of donor periods for Present Day English adverbial con-
nectors and subordinators (in per cent of PDE connectors; numbers for
subordinators according to Kortmann 1997)
The general line of development is, as was to be expected, parallel for adverbial
connectors and subordinators in the Old and Middle English periods, when the
originally ambiguous adverbs/conjunctions lost their syntactic polyfunction-
ality and a new set of the respective connectors was required. In particular,
there was a need for new kinds of subordinators as well as adverbial connectors
when the Old English pronominal connectors comprising a case form of the de-
monstrative (e. g., the dative m) had become opaque (cf., for example, forms
such as ME forthen; see MED, s. v. for-than). The changes in this period are,
therefore, mainly qualitative in nature and are thus triggered by the typological
Middle English: a period of experiment 85
changes English underwent after the Old English period (for details, see below,
Chapters 7 and 9).
Yet, after this time of parallel development lasting until the end of the Middle
English period, there is a stark difference in the stability of the innovations in
the category of subordinators on the one hand and adverbial connectors on
the other: the inventory for subordinators was in addition to the repertoire
of coordinating conjunctions (see above, 6.1) set to almost three quarters
(ca. seventy-three per cent) by the end of the Middle English period. For the
relations cause/result, for example, Kortmann states that [b]y and large,
then, the inventory of causal connectors in Present Day English resembles that
of (Late) Middle English most closely (Kortmann 1997: 331). The fndings
are strikingly different for adverbial connectors: more than sixty per cent of
todays adverbial connectors were only coined in the Early Modern, Late Mod-
ern and Present Day English period.
6.6. Middle English: a period of experiment
This picture of a rather steady path in the increase of adverbial connectors,
however, changes if we have a closer look at the number of new adverbial con-
nectors coined in the various sub-periods, in particular those which have sur-
vived into Modern English.
We then see that there is no steady increase of new coinages, but that there
are three notable peak periods before Present Day English: ME 1/2 (1150
1350), EModE2 (15701640) and LModE2 (17801850).
Figure 64: Innovations: surviving adverbial connectors per sub-period
(absolute number of innovations)
86 Adverbial connectors in the history of English
This fnding, i. e. that the history of adverbial connectors is not as steady as the
graphs for the donor periods suggested, is supported by a closer investigation of
the innovations (irrespective of whether they have survived into Modern English
or not) and disappearances per period as presented graphically in Figure 6.5:
6
Figure 65: Adverbial connectors: absolute number of new coinages vs. number of
last occurrences per sub-period
This shows that the inventory of adverbial connectors was extremely unstable
until the Early Modern English Period (EMod2; 15701640). Only from that
period onwards do we see a more constant development of new coinages and
losses in the respective periods. The Middle English period in particular stands
out as period of trial and error, so that even without a qualitative analysis
(which will be provided in the following chapters) we see a struggle for a new
system of connectors.
For a last diagram illustrating the instabilities in the Middle English, see
Figure 6.6, which compares the absolute number of new coinages per period
with those adverbial connectors which have not survived into Present Day Eng-
lish. This diagram again shows that the majority of those adverbial connectors
which were coined during the Middle English period have not survived.
In sum, we fnd that there are striking differences in the diachrony of the
three main types of connectors used to explicitly signal the textual relations of
sentences or discourse: While the inventory of coordinating conjunctions has,
in essence, remained stable, there have been drastic changes in the categories
of both subordinators and adverbial connectors.
6 The last two periods LModE3 (18501920) and PDE (1920 ) have been excluded
because it is impossible to make sure that items last attested in corpus texts of
LModE3 will not be used again.
Middle English: a period of experiment 87
Figure 66: Adverbial connectors: absolute number of innovations vs. non-survivors
As regards subordinators, Middle English emerges as the crucial period: the
inventory of subordinators has remained unchanged to a large extent (i. e. al-
most three quarters) since the end of the Middle English period. Most of the
subordinators coined in the Early Modern English period were ephemeral and
have not survived into Present Day English, so that this period emerges as a
period of experiment and transition, which eventually has not had any major
impact on the Present Day inventory of subordinators.
This scenario and time frame is different from that of adverbial connectors.
There is no single period which can be established as crucial for the Pres-
ent Day English inventory adverbial connectors. There has been a constant
coinage of new adverbial connectors in all periods of English, which fnally
resulted in the Present Day English inventory. A more detailed investigation of
new coinages and losses, however, shows that with respect to adverbial connec-
tors it is the Middle English period which emerges as a period of experiment
and transition.
The next chapters will investigate various factors which help to explain these
divergent histories. Chapter 7 will summarize the main developments in the
morphology of adverbial connectors from various perspectives: How are the
new linking adverbs coined? Which means of word formation are employed
and which language material Germanic or Romance is used? Is there an
underlying, recurrent pattern in these new elements?
7. Adverbial connectors: morphology
7.1. The expansion of the English lexicon
Since the diagrams showing the donor periods of Present Day English connec-
tors seem to suggest a more or less constant fow of innovations of adverbial con-
nectors over the periods, one might, at frst glance, attribute this to the expansion
of the English lexicon in general, especially since we are dealing with linguistic
elements which are said to be situated at the interface between the lexicon and
grammar. It is standard knowledge that with about 252,000 words or 415,000
words including the complex words (numbers according to the New Shorter
Oxford English Dictionary; see Leisi and Mair 1999: 46) the vocabulary of
Modern English is the largest of all European languages. English lost much of its
native Germanic vocabulary after the Old English period and borrowed an enor-
mous number of words from French in the Middle English period, in particular
in the fourteenth century (my period ME3; see Baugh and Cable 2003: 167187;
Scheler 1977: 52). It then further enriched its vocabulary signifcantly in the Ear-
ly Modern English period, mainly from Latin and Greek sources (Baugh and
Cable 2003: 214234). In one of the recently published studies on Late Modern
English, Beal shows that in this period with a peak at the beginning of the
nineteenth century (my period LModE2) new words were coined at a rate not
seen since the sixteenth century because of trade, exploration and colonization
and in particular technological and scientifc innovations (see the large number
of, for example, neo-classical compounds or eponyms; Beal 2004: 1429).
In view of these facts, the increase of linking adverbs after the Old Eng-
lish period seems to agree with what we expect from the development of the
English language. At frst glance, the innovations in the feld of adverbial con-
nectors could be thought to be another instance of just this development of the
enlargement of the English lexicon, with peaks in the Middle English, Early
Modern English and Late Modern English periods. A closer investigation of
the innovations, however, shows that this purely numerical parallelism does not
get to the heart of the matter. With the increase in vocabulary sketched above,
the English vocabulary became a mixed vocabulary (Leisi and Mair 1999:
Chapter 3); even in its basic vocabulary as documented in the General Service
List, only about half of the types are Germanic (Scheler 1977: 72).
The summary account of innovations in the feld of adverbial connectors in
Table 6.1 above, however, shows that there are in all of the sub-periods only
very few loans among the new connectors:
The expansion of the English lexicon 89
Table 71: New adverbial connectors: loans
ME 1/2 French: certain, certes
ME3 Latin: ergo, item
EModE3 Latin: videlicet
PDE Latin: plus
This means that the scenario sketched above has to be discarded: there are
almost no unambiguous loans among the new adverbial connectors. The use
of most of the Latin loans is furthermore limited to certain legal or religious
(liturgical) registers, as has been shown above for item furthermore. Item is
only used as a reinforcing connector in some handbooks or law texts of the
Late Middle and Early Modern English period (CMEQUATO, CMDOCU4,
CMREYNES, CMPRIV, CELAW1; cf. Chapter 1.7 and the list of occurrences
in Appendix B.1.2).
Similarly, additive/equative videlicet to wit is only attested in Early Mod-
ern English law texts. Moreover, it is often not clear if it is indeed employed on
the sentence level signalling a two-place relation or if it does not only provide
an amplifcation on the phrasal level:
1
(60) in the said frst recited Act mentioned shall alsoe before such Justice or
Justices of the Peace by whom such Oath is to bee given and administred
likewise take an Oath (which such Justice or Justices of the Peace is and
are hereby impowered to administer) to this Effect, videlicet I A. B. doe
sweare That on the Five and twentieth day of December in the Yeare
of our Lord One thousand six hundred ninety & fve I was actually a
Prisoner in the Custody of the Goaler of Keeper of the Prison of C. in
the County of D. (CELAW3, An Act for Relief of Poor Prisoners for
Debt or Damages, p.VII, 76).
Another reinforcing connector with Latin roots is plus, which was frst bor-
rowed as a preposition in Early Modern English and has only acquired a
connective function since the 1960s. It is still considered colloquial by the
1 Phrases such as for example, for instance, in other words, to wit have been excluded
from the quantitative analyses of the present study because they are predominantly
used as appositions on the phrase level (i. e. provide an amplifcation of one sen-
tence constituent) rather than as connectors on the sentence or discourse level (see
also below, Chapter 11).
90 Adverbial connectors: morphology
OED (s. v. plus C.2: colloq Introducing a clause: and furthermore; and
moreover).
Plus furthermore and ergo are not attested at all in the corpus texts. Ergo
is according to the OED used from ca. 1400 to 1780; the four quotes given
in the OED are also exclusively from legal and liturgical texts. This implies
that item, videlicet and ergo are predominantly used often in very formulaic
contexts by authors who would also compose similar kinds of texts in Latin
(OED, s. v. ergo).
With regard to loans from French, we only fnd transitional certain and
certes as unambiguous items borrowed in ME1/2. Both of them are, however,
predominantly found in texts translated from a French exemplar (see Appendix
B.5.2). Moreover, in most of the occurrences of certain, a connective reading is
also not fully unambiguous in examples such as (61), i. e. whether certeyn func-
tions as a content disjunct, modifying the propositional content of the connect,
or a non-propositional adverbial connector:
(61) The seconde the semeth is light ynowgh, for thou maist hurte no man
but with thy tonge. Certeyn the frst is as light as thys, for ther may no
matere be of yuel wille wher no couetise is (CMAELR4, p.17).
In addition to these, we fnd a number of hybrids, i. e. adverbs formed from
French or Latin bases by addition of the native suffx -ly, such as correspond
ingly, consequently, conversely, equally, fnally, incidentally, semblably or
similarly (see Table 7.2). All of these are commonly used as circumstance ad-
verbs and acquired their connector value in collocations with a verb of commu-
nication (see above, Chapter 3.2.3); with the exception of consequently (see Al-
tenberg 1984: 45)
2
and fnally (see Appendix B.1.1), most of these are used very
sparingly. Both semblably (CEEDUC1A) and correspondingly (FROWN-D),
for instance, are only used once (see Appendix B1.3).
The only hybrids which can be said to belong to the core group of adverbial
connectors are the contrastive, antithetic contrariwise and contraryways (frst
appearing in the phrase in contrary wise in ME4; for the lexicalization of the
suffx -wise/-ways, see OED, s. v. -ways and Lenker 2002). Contariwise is fre-
quent in Early Modern English (see Appendix B.4.2).
2 In the material collected from the LOB and LLC corpora, Altenberg (1984: 40)
fnds altogether nine instances of consequently and two of accordingly (in contrast
to 265 of so and sixty-two of therefore).
The expansion of the English lexicon 91
Table 72: New adverbial connectors: hybrids
ME3 verily
ME4 fnally, contrariwise
EModE1 consequently, semablably, surely
EModE2 contrariways
LModE1 accordingly
LModE2 conversely
PDE correspondingly, incidentally, similarly
In sum, we see that the borrowed and hybrid items certainly do not form the
core group of adverbial connectors of Present Day English. They also stand out
structurally, mainly because of their length (correspondingly, incidentally and
similarly).
Most of the new elements which emerge in a new function as adverbial con-
nectors in the Early and Late Modern English period are formed with Ger-
manic word material. The main categorial sources are prepositional or other
phrases (cf. lexicalized prepositional phrases such as above all or after all or
coinages such as all in all, anyhow or anyway; see Table 7.3). This means that
the increase in the number of items is not primarily part of the general expan-
sion of the English vocabulary in the Middle English, Early Modern English
or Late Modern English periods, which is predominantly attributable to loans
from French, Latin or Greek.
This is particularly intriguing when we compare these fndings to those for
the subordinators. In the feld of subordinators, borrowed items are much more
frequent from the Middle English period onwards, in particular in Early Mod-
ern English, the period of experiment for the subordinators (see Kortmann
1997: 300, 304305). The higher number of loans in Early Modern English
subordinators is mainly due to the use of present or past participles, often in
complex subordinators with the general subordinating particle that. Most of
these are borrowed from or at least modelled on the Romance patterns:
3
considering (that), conditioned (that), excepting (that), provided (that), providing
(that), supposing (that)
during (that), granted (that).
Although many of these subordinators are ephemeral and not very frequent
(Kortmann 1997: 309), it is still striking that the subordinators, which are
3 This pattern was also transferred to native material, such as the calques outtaken
(that), notwithstanding (that) or seeing (that).
92 Adverbial connectors: morphology
commonly classifed as grammatical elements, mirror the general tendencies
of the expansion of the English lexicon, i. e. the high number of borrowed
material from Latin in the Early Modern English period, to a larger extent
than the adverbial connectors, which are situated at the interface between
the lexicon and grammar. This is even more intriguing in view of the fact
that the inventory of Present Day English adverbial connectors was expanded
considerably in the Early and Late Modern English period (in contrast to the
repertoire of subordinators, which has remained unchanged to almost three
quarters since the end of Middle English) and that adverbial connectors are
particularly frequent in the very formal registers of academic prose (see
above, Chapter 3.5.5), where we would have expected a substantial impact of
the classical languages.
7.2. Major morphological changes
To gain a better understanding of these processes, the various source categories
of adverbial connectors are summarized in Tables 7.3 and 7.4, which list all
Old English adverbial connectors as well as the new coinages in the respective
periods. Table 7.3 provides a survey of the adverbial connectors according to
their patterns of word formation and distinguishes simple adverbs, compound
adverbs and derivations. Table 7.4 then lists the complex formations and dif-
ferentiates between pronominal connectors which may be realized as case
forms of a pronoun or may comprise such a case form (see above Chapter 5.5)
and lexicalized prepositional (or other) phrases.
Table 73: Adverbial connectors: source categories
Simple Adverbs Case Forms Compounds Derivations
OE
eac
eft
her
huru
nu
swa
a
eah
onne
Superlative:
rest
elles
eftsones
eftsona
huruinga
rto-eacan
culice
eornostlice
solice
gewislice
witodlice
Major morphological changes 93
Simple Adverbs Case Forms Compounds Derivations
ME 1/2
certain
certes
for
iwis
yet
Comparative:
more
certainly
ME 3
ergo
item
Comparative:
further
furthermore
furtherover
moreover
overmore
algates
sekirly
truly
verily
ME 4
Comparative: <after> fnally
EModE 1
<again>
well
?howsomever consequently
semblably
surely
likeways
likewise
EModE 2
<beside>
sure
however
?howsoever
besides
contraryways
EModE3
still
videlicet
hence
thence
whence
LModE1
too accordingly
LModE2
conversely
LModE3
PDE
plus
Comparative:
rather
correspondingly
incidentally
similarly
94 Adverbial connectors: morphology
7.3. Simple adverbs, compounds and derivations
7.3.1. Simple adverbs
With respect to the adverbs proper, the inventory of Old English and early
Middle English adverbial connectors comprises a fairly large number of mor-
phologically simple, i. e. so-called original adverbs. The items listed in the
left column of Table 7.3 show that this is not a recurrent pattern in the later peri-
ods of English: it is only attested in contrastive/concessive still and additive too
(which is, however, an exception because it cannot be placed sentence-initial-
ly). We might also add the univerbated additive/contrastive again (OE on-gean
from ongegn lit. on-straight; cf. PDG ent-gegen facing locally; see OED,
s. v. again) and beside (OE be sidan [dative singular] by the side). The origin
of these items in a prefxation and a prepositional phrase are, however, no lon-
ger fully transparent by the time they are frst used as adverbial connectors in
EModE1 and EModE2 respectively (signalled by < > in Table 7.3).
7.3.2. Compounds
Only a very limited use has been made of composition. In Old English, there
are only three compounds, two of which express the relation of reinforcement
furthermore: eftsona/es (cf. eft again, furthermore + sona immediately)
and rtoeacan (cf. rto thereto + eacan as well).
The only period which shows a marked impact of a specifc kind of com-
pounds is ME3, where we fnd the new coinages furthermore, furtherover, more
over and overmore (all of them additive/reinforcing connectors). These forms
are peculiar in that they use a double comparative: the comparatives more (cor-
responding to senses of the positive adjectives great, much, and many) or over
(local upper, higher in position) are added to bases which are already compara-
tive themselves. Farthermore and furthermore, for instance, are formed from
the spatial comparative further/farther, which may, however, also be used as
an adverbial connector by itself (see OED, s. v. further 1. more forward; to or
at a more advanced point. a. in space, or in a course of procedure or develop-
ment ). This in this case double pleonasm is particularly evident when
we compare the simple comparative more (as used from ME1/2) with the forms
moreover and overmore (both coined in ME3) which swap over their bases.
This pattern of forming adverbs and adjectives of place (rarely of time) in
the comparative degree with more is fully transparent but no longer productive
in later periods of English. The OED considers an analogy with or calque on
Simple adverbs, compounds and derivations 95
North Germanic formations probable, since the earliest instances in English
occur in thirteenth century sources and since similar formations consisting of
a comparative and meir occur in North Germanic (see OED, s. v. -more suffx).
The basic pattern, however, i. e. the use of temporal or spatial adverbs in the
comparative, is a native pattern of Old English, too (cf. after, rather), and is also
attested widely cross-linguistically (see Chapter 8.3 on cognitive source do-
mains). The high number of compound adverbs in the pure semantic relation
of addition is, however, certainly no coincidence, since they iconically by
their length highlight that the author wants to make the connection between
the connects explicit (see below, Chapter 11).
7.3.3. Derivations: adverbs in -lice, es and ways/wise
The other morphological forms of adverbs functioning as adverbial connectors
refect the increasing tendency of English to distinguish adverbs from adjec-
tives by various suffxes, in particular by the suffxes
-ly (from OE -lice) by reanalysis of adjectival -lic and adverbial -e (see
McIntonsh 1991, Nevalainen 1997; this reanalysis must have taken place in
the pre-literary period, since suffxation in -lice is already attested in very
early glossaries, such as pinalErfurt; see Sauer 2006: 265),
-es; originally a nominal or adjectival case form genitive singular mas-
culine/neuter which is already in Old English transferred to feminine
nouns, as in OE nihtes (cf. OE feminine niht; see also PDG nachts), and
then re-analysed as an adverbial suffx (in besides, hence, thence, whence),
-ways/-wise from original prepositional phrases such as in a contrary/like
wise yielding contrariwise/-ways or likewise/-ways (see OED, s. v. -ways
and Lenker 2002 on -wise).
In all periods, new adverbial connectors are coined from derived circumstance
adverbs in OE -lice (added to adjectives which do not end in -lic) and later -ly.
Yet in Old English the semantic range of adverbial connectors in -lice is very
restricted. Virtually all of the adverbs are truth-intensifying adverbs which
may be used in connective function for the expression of the marginal connec-
tor relation of transition (culice certainly, eornostlice earnestly, solice
truly, gewislice certainly and witodlice truly; see below, Chapter 8.5); only
culice may also code the resultive relation therefore (see OED, s. v.3). This
semantic restriction of adverbial connectors in -ly remained fairly unchanged
until the Early Modern English period (cf. certainly, sekirly, truly, verily, sure
ly). Only from that time onwards are original circumstance adverbs in -ly also
96 Adverbial connectors: morphology
used for the relations of result (e. g. consequently, accordingly) or contrast
(conversely). In sum, we can say that derived adverbs in -ly have never played a
central part as adverbial connectors. They are predominantly found in the feld
of pure (i. e. additive or transitional) connectors, but only rarely in the more
complex impure connectors marking the ccc-relations cause, concession
and contrast.
This is different with adverbs formed in -es, which are attested as a case
form from OE reformulatory elles else (genitive singular neuter). In the thir-
teenth century, the ending -es is by analogy appended to the ambiguous adverb/
conjunction algate however; nevertheless, a loan from Old Norse (ON alla
gtu; cf. OED, s. v. algate). Similarly, it is used in besides (cf. OE be sidan) and
the originally spatial/directional and then resultive hence (OE heonan/heonon
with the directional suffx -an hence; from there) as well as the parallel thence
and whence (which may also be used as adverbial connectors). Adverbs in -es
are thus used predominantly for the impure ccc-relations: they are more spe-
cifc because they are not derived from adjectives, but have their origins in,
originally spatial, adverbs or (prepositional) phrases (cf. besides; algates).
7.4. Pronominal connectors lexicalized phrases
Simple, compound and derived adverbs have thus been employed as adverbial
connectors in all periods of English. Yet, other formations such as pronominal
connectors and lexicalized phrases (listed in Table 7.4) will be shown to have
been much more important since they are more indicative of the structural
changes in the system of adverbial connectors and their causes.
The following section deals with complex expressions serving as adverbial
connectors. To be included, such complex expressions have to be lexicalized.
This means they have to be morphologically and semantically fxed (at least to
a certain extent) and must fulfll the following conditions:
(i) They must have lost at least some properties of the original phrase (e. g.
modifability of the head noun of a noun phrase).
(ii) They must have at least one adverbial reading which is not fully re-constru-
able from the meaning of their parts.
This, for instance, means that on the contrary is not included because its mean-
ing is fully re-construable from its parts, while on the other hand, which re-
quires an understanding of the metaphorical transfer of hand, is included. The
Pronominal connectors lexicalized phrases 97
section here only introduces the relevant terms and issues and sketches the
general lines of development which will then be discussed in more detail in the
chapters on the development in the causal/resultive and contrastive/concessive
relations (see Chapters 9 and 10).
Table 74: Pronominal connectors lexicalized phrases
Pronominal
Connectors:
Demonstratives
Pronominal
Connectors:
Other Pronouns
Lexicalized
Prepositional
Phrases
Lexicalized
Phrases
OE
case forms:
s
y
form
foron
fory
nayls
hwere
swaeah-hwere
eahhwre
hwt
hwt a
swa
eall-swa
swaeah
eac swylce
a giet
ME 1/2
for that
forhwi
?neverthelatter
nought-for-than
?therefore
wherefore
forsooth
on oer half
overall
herefore
oer side
ME 3
for which
?nevertheless
nought-for-that
nevertheless
nought-for-thi
nevertheless
over this
over that
otherways
otherwise
ME 4
contrariwise
howbeit
notwithstanding
EModE1
?nonetheless albeit
nonetheless
over and besides
98 Adverbial connectors: morphology
Pronominal
Connectors:
Demonstratives
Pronominal
Connectors:
Other Pronouns
Lexicalized
Prepositional
Phrases
Lexicalized
Phrases
EModE2
above all
indeed
in fne
at least
in the meantime
in the meanwhile
in sum
over and above
EModE3
after all
on the other hand
LModE1
at the same time
in truth
LModE2
anyhow of course
at all events
in fact
altogether
LModE3
anyway at any rate
in the same way
all the same
PDE
all in all
instead
7.5. Pronominal connectors
7.5.1. Pronominal connectors in Old English
The notion as well as the main properties of pronominal connectors have
been introduced in Chapter 5.5 above. In Raibles (1992) continuum, they are
situated between asyndesis (I) and the explicit paratactic connection (III). This
phase II, then, singles out connectors comprising a pronominal element, which
function as explicit signposts of discourse deixis.
In Old English, we fnd case forms of the demonstrative, which are used to
mark a resultive relation therefore, namely the genitive s (masculine/neu-
ter) and the instrumental/locative y. While s is very rare and not attested at
all in the corpus texts of the present study, y is regularly employed in indepen-
Pronominal connectors 99
dent and correlative constructions (for the sequence cause result therefore
as well as for the sequence result cause for). Consider, for example, the
following example from the beginning of Wulfstans Sermo Lupi ad Anglos:
(62) Leofan men, gecnawa t so is: eos worold is on ofste, & hit nealc
am ende, & y hit is on worolde aa swa leng swa wyrse (COWULF3, R2)
Beloved men, know what the truth is: this world is in haste, and it ap-
proaches the end; and therefore it is ever worse and worse in the world;
This pattern of pronominal connectors is the predominant one for the Old
English causal/resultive connectors form, foron and fory for; therefore,
which consist of the preposition for and forms of the simple demonstrative,
namely the dative m (masculine/neuter) or the instrumental on or y. It is
also attested in the contrastive/concessive connector nayls nevertheless.
In addition to demonstratives, the concessive/contrastive relation is marked
by connectors containing the originally interrogative pronoun hwer which
of the two?, which may also be used as an indefnite pronoun either of two
(see BT, s. v., and Campbell 1959: 716718).
Other cases of pronominal connectors are hwt (a) what (then), which uses
the neuter (nom./acc.) form of the interrogative pronoun hwa who (Campbell
1959: 716) and, in particular, all of the forms which employ the extremely
polyfunctional particle swa so, which may also on its own be used as a resul-
tive adverbial connector (see BT, s. v. swa IV.). For the polyfunctionality of swa,
see the following entry chart extracted from BT; s. v. swa (all of these uses are
discussed in detail in Schleburg 2002).
Table 75: OE swa
I. rel pron As, that
II. demonst pron;
III a. sw sw such as
IV. adv
(1) defned by that which precedes so, in this or that way, thus
(2) defned by that which follows, (3) used indefnitely, so and so
V. adverbial conjunction,
(1) with indic. as -
(2) with indic. or subjunct. expressing an actual or possible result, so that
(3) with subjunctive, as (if)
(4) with optative, so
(5) with a conditional force, provided that, if so be that, so :-
(6) marking a consequence, so, therefore, on that account :-
(7) local, where :-
100 Adverbial connectors: morphology
(8) temporal, as, when :--
(9) marking the grounds of action, as, since :-
(10) although, yet :-
VI. sw . . . sw,
(1) - where sw occurs once with a demonstrative, once with a relative force, so
as, so that, as as :--
(2) correlative, (a) either or, as well as :-- (b) whether or :--
(c) sw hwer sw . . . sw whether or :-
7.5.2. Pronominal connectors in the history of English
The morphological make-up and the functions of the central pronominal con-
nectors marking the relations cause/result (OE y, form, foron, fory
because; therefore and concession/contrast (OE swaeah, eahhwere,
swaeahhwere, etc.) will be discussed in detail in the chapters on their re-
spective semantic relations in Chapters 9 and 10. For a frst survey of the his-
tory of pronominal connectors, Table 7.4 above shows that this pattern is, as
far as forms of the demonstrative are concerned, only used until the Middle
English period (compare, in ME3, additive over this, over that and contrastive/
concessive nought-for-thi nevertheless, nought-for-that nevertheless).
Speakers and writers of Present Day English may, of course, freely use ad-
verbial connectors which include a pronominal reference to the unit which is
to be related, such as
in addition to this/that (i. e. what I have mentioned), for (all) this/that (i. e. what
I have said), besides this/that, as a result of that, because of this/that, despite/in
spite of this/that, instead of this/that, etc. (for a fuller list, see Quirk et al 1985:
632 and Fraser 1999: 939940).
In contrast to Old or Early Middle English, however, these phrases are neither
fully lexicalized nor are they very frequent. In the LLC and LOB corpora, for
example, the full lexical phrase for this reason is used only seven times (com-
pared to 265 instances of so, sixty-two of therefore and thirty-fve of thus).
Overall, Altenberg fnds 385 instances of adverbial connectors in contrast to
only 14 instances of pronominal connection (Altenberg 1984: 40). This is a
striking difference to the use of pronominal form, which is used in about 80
per cent of the cases of causal/resultive connection in Old English (see below,
Chapter 9, Table 9.6).
In contrast to German (see the bold pronominal elements in causal/resultive
PDG demnach, deswegen, infolgedessen or concessive PDG nichtsdestoweni
ger), English thus has not only given up this very productive pattern of forming
Pronominal connectors 101
Old and Middle English connectors. Intriguingly, Present Day English has not
even kept a single of these pronominal connectors which had been central in
its earliest periods.
4
In view of the general history of the English language, it is certainly no co-
incidence that Middle English is the last period which employs (or rather: tries
to employ) adverbial connectors of this pronominal kind in larger numbers,
since Early Middle English is the time when the nominal categories case and
gender were lost. Much more important for the present issue of pronominal
connectors, however, are the changes which affected both of the Old English
demonstratives (the proto-article se/t/seo and the emphatic demonstrative
es/is/eos). These forms, which were pronominal as much as they were ad-
jective- or determiner-like, are no longer infected for case and gender after the
Old English period (see Brunner 196062: II, 130133; Lass 1999: 112116).
At the very beginning of the Middle English period (around 1200), we see a
collapse of the old system of demonstratives, but at the same time also the
emergence of two distinct categories: the invariable article the, and the demon-
stratives this/that. While there is among the nineteen different forms of the
demonstratives nowhere a direct ancestor for the Middle and Present Day
English article the (for its etymology, see OED, s. v. the dem. a. (def. article)
and pron.), it is clear that the neuter nominative singular forms of the simple
(OE t) and the emphatic demonstratives (OE is) yielded the proximal/
distal deictics this and that.
5
The collapse of both of the paradigms of the demonstratives thus frst of all
triggered when the case forms OE m and y lost their transparency as da-
tive/instrumental forms of the demonstrative the coinage of a number of new
adverbial connectors in earlier sub-periods of Middle English (ME1 to ME3).
Most of these, such as the additive over this/that or the resultive for that, em-
ploy the Old English pattern of pronominal connectors formed by a preposition
and the demonstrative that (for similar patterns in the Romance languages,
especially Latin, see above, Chapter 5.5). These innovations, however, have
4 Nevertheless and nonetheless might be considered marginal cases because they
employ the form the, which is also found in other contexts preceding an adjective
or adverb in the comparative degree (cf. the sooner the better). For speakers of
Present Day English, the formation from the original instrumental of the demon-
strative, y, however, is not transparent. The OED, for example, does not give a
meaning but the paraphrase The radical meaning is in or by that, in or by so
much (OED, s. v. the adv.).
5 Loss of infections is, as with most infections, earliest in the East and in the North.
We fnd remnants of infected forms of the Old English demonstratives in some
southern and western texts as late as the thirteenth century.
102 Adverbial connectors: morphology
probably not survived because, at about the same time, that also developed into
the general subordinator (instead of OE e, which, in turn, was probably given
up because of its homonymy with the new invariable article the). From early
Middle English onwards, that may be used as a complementizer, a relativizer
and often pleonastically as an indicator of a subordinate clause. For the
conjunctions, Fischer summarizes the developments which are analysed in
detail in Kivimaa (1966) as follows:
The scenario therefore is not that e was replaced by that; rather, that, by now the
general subordinator, becomes added, as a kind of pleonastic element, to all kinds
of conjunctions. Thus we fnd in Middle English, now that, (g)if that, when that,
and after prepositions (prepositions not encountered in Old English conjunctive
phrases) before that, save that, in that, etc. (Fischer 1992: 295).
As will be argued in an analysis of causal/resultive connectors below (Chapter
9), this use of that as the general subordinating particle seems to have inhibited
a further extensive use of pronominal connectors such as for that. When placed
sentence-initially, for that is not only (as OE form had been) ambiguous as to
whether it is an adverb or a conjunction, but openly misleading, because hear-
ers or readers of Middle English expect a pre-posed subordinate clause with
(pleonastic) that and not a paratactic construction introduced by an adverbial
connector. Indeed, we notice that no items employing this pattern of pro-
nominal connectors have been coined after the Middle English period and
none of the ones coined in the Middle English period have survived or are
still transparent as pronominal connectors. It is this development which also
accounts for the fnding (sketched above) that most of the Middle English in-
novations did not survive the Middle English or Early Modern English period.
Middle English is a period of transition and experiment, mainly because the
pronominal connectors coined in its sub-periods were given up.
7.6. Lexicalized (prepositional) phrases
7.6.1. General tendencies
As has been pointed out above, pronominal connectors have the advantage
of being inherently deictic.
6
They function as signposts of discourse deixis,
because they explicitly signal the link between two sentences or two chunks of
6 I use deixis and deictic as the hyperonym covering all kinds of deixis, although
some linguists argue for a distinction between reference and discourse deixis; for
details, see below, Chapter 9.4.1.
Lexicalized (prepositional) phrases 103
discourse by using their deictic value to anaphorically (in correlative construc-
tions also cataphorically) take up one or more constituents of the proceeding
or ensuing discourse (cf. Raibles defnition Junktion durch Wiederaufnahme
(eines Teils) des vorhergehenden Satzes Linkage by resumption of (a part of)
the preceding sentence; Raible 1992). This overt marking of the anaphoric (or
cataphoric) relations is for reasons of information signalling and processing
much more important for adverbial connectors than for subordinators. Subor-
dinators usually work on a local level of discourse and the information given in
the subordinate clause is commonly not pursued in the following discourse (for
details, see above, Chapter 2.4). In paratactic constructions, however, the prop-
ositions of the two connects are of equal weight: both of them may be pursued
in the following discourse. It is therefore most important that the adverbial
connectors, which signal the semantic relation between the two connects, are
distinctive in their deixis. After pronominal connectors were given up, speak-
ers and writers of English have thus not given up using deictic material alto-
gether, but mainly choose time and space deictics to compensate for the loss of
pronominal connectors (for details, see below, Chapters 8.2 and 8.3). Most of
these are coded in lexicalized prepositional or other phrases.
7.6.2. Lexicalization
Table 7.4 nicely illustrates compare the two columns on the left with the
two columns on the right that only in Middle English the period of ex-
periment both the patterns of pronominal connectors and that of lexicalized
phrases are used to a considerable extent. Old English, by contrast, virtually
does not employ lexicalized phrases. From Early Middle English onwards, on
the other hand, speakers and writers ever more rarely coin new adverbial con-
nectors comprising pronominal material.
7
As has been pointed out above, prepositional and other phrases are only
included if they show a certain degree of lexicalization (for the criteria, see
above, 7.4). Often, the originally complex connectors show orthographical fu-
sion, such as fully fused indeed. For others, orthographical variants are at-
tested which support their inclusion in the present study. In Middle English,
for instance, the phrase at/on/of last fnally is attested as alast (see MED,
7 Anyhow and anyway (comprising the element any which may be used pronomi-
nally in Present Day English), for example, are not relevant for the present issue,
because they are generalizing rather than explicitly deictic. For this pattern, see
below, Chapter 10.
104 Adverbial connectors: morphology
s. v. a-last). This also applies to longer phrases such as over and above fur-
thermore, a popular reinforcing connector in Early Modern English, which is
also recorded as the single word hoverendebuv (see OED, s. v. over and above).
Similarly, in Present Day American English writers sometimes choose to sig-
nal the lexicalization of the phrase after all in its spelling afterall as one word
(see OED, s. v. after; 10.; quoted above, Chapter 1).
Some of the lexicalized phrases, however, also show subsequent reduc-
tion: this is most evident in PDE of course, which in particular in spoken
language is commonly reduced to course (and often hard to differentiate
from causal cause < because). Interestingly, the OED has a separate entry
for course, but lists the highly frequent of course among the entries of course
(see OED, s. v. course n. 37c, and Vandenbergen-Simon and Aijmer 2001 for
details on the high frequency and polyfunctionality of of course in Present
Day English). This shortened form has been attested since the beginning of
the twentieth century (see OED, s. v. course, course). Its frst attestation as an
adverbial connector is recorded for 1967:
1901 MERWIN & WEBSTER Calumet K i. 13 Have you tried to get any
of it here in Chicago? Course not. Its all ordered and cut out up to Led-
yard. 1904 G. S. PORTER Freckles xvi. 326 I bet you its a marked tree!
Course it is! cried the Angel. 1967 O. NORTON Now lying Dead iv. 58
Course, he might go off. He does sometimes (OED, s. v. course, course).
7.6.3. Verbal and nominal phrases
Most of the phrases which have been lexicalized as adverbial connectors are
prepositional phrases. There are only three originally verbal phrases, namely
the ambiguous adverbs/conjunctions albeit, howbeit and notwithstanding (that).
Albeit and notwithstanding are calques modelled on French (cf. French tout soit
il que and OF non obstant). All of these verbal phrases were frst used as sub-
ordinators (see Kortmann 1997: 333), but are also though less frequently
attested as contrastive/concessive adverbial connectors during the Middle and
Early Modern English periods (see Appendix B.4.1). As has been shown above,
the lexicalization of verbal phrases was a regular pattern for the coinage of new
subordinators, in particular from non-fnite constructions. This pattern is, by
contrast, not typically employed for the formation of adverbial connectors.
Most of the new adverbial connectors are lexicalized forms of prepositional
phrases, usually of the simplest kind, i. e. preposition plus noun (without any
other pre- or postmodifcation). This agrees with a more general tendency in
Lexicalized (prepositional) phrases 105
the structural make-up of the English lexicon, namely the substitution of single
words by multi-unit items (cf. the Chapter Der Wortverband als Wort in Leisi
and Mair 1999: 94111). Although the term Wortverband (referring to fxed
lexical bundles with a particular meaning) is as the authors admit a rather
vague term, it still manages to characterize phenomena such as the substitution
of simple, often borrowed verbs by phrasal and prepositional verbs, or by ver-
bal operator and gerund (cf. to have a look, to give a ring, to make a move, etc.).
The origin of these verbal forms (including phrasal and prepositional verbs)
has commonly been sought in everyday, colloquial language. Yet, in a corpus
study of phrasal verbs from 1500 to 1700, Thim (2006) shows that this claim
is unsubstantiated: he suggests that it is their specifc function rather than their
presumed informality which determines their use in Early Modern English
texts (Thim 2006: 305). Similarly, with the exception of transitional indeed and
in fact, none of the lexicalized phrases discussed here shows a strong affnity to
speech-based texts (CEPLAY, CETRI, CLPLAY). Furthermore, a colloquial
origin is not likely for the lexicalization of adverbial connectors, which are
mainly employed in written genres, more specifcally academic prose. By
contrast, a closer look at the prepositional phrases reveals that they show recur-
rent patterns in the material they use: the two basic source domains are time
and place/space. place/space in particular is, however, only explicable on the
basis of the spatialization of language in literacy.
8. Cognitive source domains
8.1. General tendencies
The analyses in Chapter 7 have shown that new adverbial connectors coined
after the collapse of the system of pronominal connectors in Old English are
only rarely simple transfers of purely lexical, propositional adverbs from cir-
cumstance to linking adverbials (see the few cases such as consequently or
similarly). Much more frequently, the innovations comprise material from a
very restricted range of cognitive source domains which help to provide the
necessary discourse deictic references for the explicit linkage of sentences or
chunks of discourse.
In his study of subordinators, Kortmann (1997) shows that interclausal rela-
tions differ markedly with regard to their cognitive basicness or centrality
for human reasoning and also their cognitive complexity or specifcness.
According to Kortmann, highly grammaticalized, preferably monomorphe-
mic or single-word connectors which are frequently used and are stable over
time indicate cognitive basicness (lexical primes code cognitive primes;
Kortmann 1997: 342). Most of the basic relations thus show a low degree of
complexity. Yet, cognitive basicness and cognitive complexity must neverthe-
less be kept separate, since concession, for instance, exhibits a high degree
of cognitive complexity, and yet clearly belongs to the core set of cognitively
basic relations. On this basis, Kortmann found a core of twelve basic rela-
tions (cause, condition, concession [ccc], result, purpose, simultaneity
overlap, simultaneity duration, anteriority, immediate anteriority,
terminus ad quem, place and similarity) and several layers of relations of
an increasingly peripheral nature.
These relations can be parcelled into four networks, i.e. temporal, locative,
modal and ccc-relations. Network-transcending changes are generally uni-
directional, so that we can distinguish between source (locative, modal) and
goal (ccc) domains. The cccc- or four-c relations (contrast is now com-
monly added as a fourth domain; see Couper-Kuhlen and Kortmann 2000)
constitute the prototypical goal network, i.e. endpoints of network-tran-
scending semantic changes. Cross-linguistic polysemy patterns (Kortmann
1997: 175211) also show that affnities are strongest between the temporal
and the cccc-networks, so that original temporal connectors often develop
cccc-readings.
107
Table 8.1 provides an overview of the three prototypical domains which can
be established as source domains for adverbial connectors. Among them are
time and space, which have long been known to create deictic reference frames
in language. motion is listed as a separate category here, but is essentially a
combination of space and time: the concept of motion implies the movement
in/over a certain space during a certain period of time. truth/fact is here
introduced as a third category because it plays an important role in innovations
in the feld of concessive and transitional connectors.
Table 81: Cognitive source domains: overview
place/space time motion truth/fact
OE
her here
rtoeacan also
eft again
eftsona again-soon
eftsones again-soon
nu now
a then
onne then
rest earliest > frst
a giet also
culice certainly
eornostlice earnestly
huru certainly, indeed
huruinga certainly,
indeed
solice truly
gewislice certainly
witodlice truly
ME
after moreover
algates (cf. gate)
further
farthermore
fartherover
furthermore
{frst; cf. fore in front}
furtherover
herefore
moreover
notwithstanding
on oer half
oer side
otherways
overall
overmore
over this/that
therefore
wherefore
at (the) last
fnally
nevertheless
neverthelatter
yet
certainly
certain
certes
forsooth
iwis
sekirly
truly
verily
108 Cognitive source domains
place/space time motion truth/fact
EModE
above all
again (cf. OE on-gean)
beside
after all
again
in fne
still
(i.e. non-
motion)
indeed
sure
surely
besides
hence
on the other hand
over and besides
over and above
thence
whence
however
howsoever
howsomever
lastly
(in the) meantime
(in the) meanwhile
LModE
anyway
in the same way
at the same time {of course} in fact
in truth
PDE
instead {rather
faster}
{} transparency of source category uncertain
time and space
8.2. Source domain time
Table 8.1 shows that the two central source domains time and space have been
used since the Old English period. In Old English, we predominantly with
regard to their token-frequency fnd simple temporal (OE nu now, OE onne
then), spatial (OE her here) or temporal/spatial (OE a then; there) circum-
stance adverbials which are used as connectors to signal the relation transi-
tion (for the occurrences, see Appendix B.5.1). This use of text-deictic nu,
her and resultive/transitional onne is illustrated by Old English examples in
Chapter 5.
Through the use of these temporal or spatial adverbs, the spoken or written
text is established as the frame of reference (instead of the real world). This, for
example, allows collocations such as now then, which would be incongruous
in a purely temporal use, but are attested in all periods of English (see the lists
in Appendix B.5.1):
(63) Nu enne on oer half. nim (CMSAWLES, p. 131).
Now, then, on the other hand, take .
Source domain time 109
In addition to these transitional connectors, Old English also employs temporal
and spatial adverbs for the relations of addition (OE eft, eftsona/-es again and
rtoeacan there-to-also) in a natural transfer of the temporal sequence on the
sequence of the argument. The source category time is thus preferred over space
in Old English, certainly no coincidence in a time when most of the language
production was oral (for the spatialization of language in literacy, see below, 8.3).
For a closer inspection of the transfers from source to goal domain, the fol-
lowing table lists the new adverbial connectors by the semantic relations they
code.
Table 82: Source domain time innovations per period
OE ME EModE LModE
addition
listing/
summma-
tive
eft again
eftsona again + soon
eftsones again +
soon
a giet also
rest earliest >frst
at (the) last
fnally
again
in fne
lastly
result onne then
contrast/
conces-
sion
nevertheless
neverthe
latter
yet
after all
(in the) meantime
(in the) meanwhile
still (i.e. non-motion)
however
howsoever
howsomever
of course
at the
same time
transition nu now
a then
onne then
The innovations in the semantic relation of addition are fairly straightfor-
ward: new adverbial connectors are coined in the sub-group of listing connec-
tors, which are used to signal the sequence of arguments by elements of time
deixis (e.g. at the last, fnally, etc.).
With respect to the ccc-relations, then is used to explicitly code not only the
transitional, but also the resultive relation, following the principle post hoc,
propter hoc (see below, Chapter 9). Table 8.2 shows, however, that most of
the innovations from the domain time are not found in the relation of cause/
result but in the relation contrast/concession: From Early Modern Eng-
lish onwards, the predominant pattern is the transfer of expressions marking
concomitance or simultaneity, as in at the same time, (in the) meantime or in
110 Cognitive source domains
the meanwhile and also still, which implies non-movement over the respective
time period (on still and yet, see Knig and Traugott 1982; on after all, see
Traugott 1997 and below, Chapter 10). Cross-linguistically, simultaneity and
concomitance expressions have been found to be one of the major new sources
of concessive connectors, subordinators (cf. PDE since, PDG whrend) as well
as adverbial connectors (cf. PDG dabei; PDBavarian derweil). In his study on
cross-linguistic patterns in the formation of concessives (Where do Conces-
sives Come from?, Knig 1985b: 269), Knig shows that simultaneity and con-
comitance are often augmented by concessive inferences, and that the coinage
of connectives is a process in which these conversational implicatures become
conventionalized. Consider, for example, the conversational implicatures Al
though he is poor and Although he is rich in sentences such as (64a) and (64b):
(64) (a) Poor as he is, he spends a lot of money on horses.
(b) Rich as he is, he spends a lot of money on horses.
8.3. Source domain place/space
With respect to the source domain place/space,
1
it has long been known that spa-
tial thinking provides humans with analogies and tools for understanding other
domains, so that many of todays metaphors of everyday language are spatial.
Spatial cognition probably plays this central role because it seems to be the evolu-
tionarily earliest domain of systematic cognition (Lakoff and Johnson 1980: 17;
Levinson 2003). In addition to this cognitive centrality, we must, however, also
consider another essential aspect: it has been suggested that the metaphoric ex-
tension of spatial terms into metalinguistic ones naming speech acts is a function
of literacy, since it is only with literacy that language is objectifed in visual space
(see Ong 1982). In one of the still pivotal studies on the subject, Ong states that
[v]isual presentation of verbalized material in space has its own particular econo-
my, its own laws of motion and structure. [] Texts assimilate utterance to the hu-
man body. They introduce a feeling for headings in accumulation of knowledge:
chapter derives from the Latin caput, meaning head (as from the human body).
Pages have not only heads, but also feet, for footnotes. References are given to
what is above and below in a text when what is meant is several pages back or
farther on. (Ong 1982: 100)
1 For the distinction between place and space, see Levinson (2003). This distinction
is only fundamental for universal accounts of place/space in language because
Levinson fnds that ideas of Western philosophy (from Aristotle onwards) are not
necessarily transferable to non-European ideas and languages.
Source domain place/space 111
Similar ideas have already been expressed by Kant, who argues that in a writ-
ten page we have frst to note the difference between front and back and to
distinguish the top from the bottom of the writing; only then can we proceed
to determine the position of the characters from left to right or conversely
(quoted from Levinson 2003: 9).
These considerations concerning the domain space and literacy clearly sug-
gest that literacy seems to favour the widespread use of spatial terms as mark-
ers of discourse structure. Table 8.1 supports this idea because it attests a large
number of new adverbial connectors employing the source category space in
the Middle English and Early Modern English period. In these periods, Eng-
lish developed into an Ausbausprache and more and more replaced Latin and
French in written documents and in registers which had, after the Norman
Conquest, been reserved to Latin and French (for the re-establishment of Eng-
lish from 1200 to 1500, see Baugh and Cable 2003: Chapter 6).
With respect to the close interdependency of Latin and literacy in the ver-
naculars, it is, of course, very hard to decide whether some of these formations
are indeed native formations as opposed to calques on Latin (or, for Middle
English, French) models. Latin uses basically the same source domains for
adverbial connectors as English does: the spatial adverb super above, on top,
over, for instance, can also be used as an adverbial connector for the semantic
relation addition moreover. Similar to English then, spatial Latin exinde
thence; after that, next in order, thereafter, then can be employed for the se-
mantic relations addition furthermore or result by that cause. Temporal
interim meanwhile, in the meantime; at the same time is, just like English
expressions for simultaneity, used to code concession however, nevertheless.
Similar parallels are found in French, as, for instance, concessive/transitional
aprs tout which is parallel to after all. Latin or French infuence can thus
never be ruled out with certainty, just as it cannot be claimed with any cer-
tainty, at least with respect to the high-frequency adverbial connectors (an
exception is the morphologically foreign notwithstanding, comprising a
non-fnite verbal form; see above, Chapter 7.6.3). Since we are dealing with
very basic cognitive source categories and their transfer to texts, these trans-
fers are of course similar in each of the individual languages and are found
in all of the European languages which have an extended tradition of literacy
(see the Euroversals in Kortmanns cross-linguistic analysis of subordina-
tors; Kortmann 1997).
In line with the spatialization of language in literacy as described above,
many of the new coinages from Middle English onwards use the spatial exten-
sion of the text on a page as the frame of reference, mainly to signal the relation
addition. Here, the spatial sequence is used to express the sequence in the line
112 Cognitive source domains
of argument. The innovations show various degrees of transparency: while
the early pronominal connectors over this/that and also the double compara-
tives such as furthermore or moreover (discussed above, Chapter 7.3.2) have
lost some of their transparency, the later coinages such as beside(s), over and
besides are fully transparent in their spatial origin.
Table 83: Source domain place/space innovations per period
OE ME EModE LModE
addition rtoeacan
also
a giet
also
after moreover
further
farthermore
fartherover
furthermore
furtherover
moreover
overall
overmore
over this/that
frst (cf. fore in
front)
above all
again
beside
besides
over and besides
over and above
in the
same way
result herefore
therefore
wherefore
hence
thence
whence
contrast/
concession
algates (cf. gate)
notwithstanding
anyway
contrast/
reformulatory
on oer half
oer side
otherways
again (OE
on-gean)
on the other hand
instead
transition her here
The same differences may be seen in those connectors which signal the resul-
tive expression. The early coinages (ME1/2) from the spatial domain, herefore,
therefore and wherefore, employ spatial terms, which do, however, not carry
the full discourse deictic load themselves, but are formations similar to pro-
nominal connectors. Instead of the demonstrative, a proximal (here) or distal
(there) element is employed, which takes over the deictic (anaphoric or cata-
phoric) function of the demonstrative. This is a regular pattern for adverbial
innovations at the beginning of the Middle English period, as attested by the
many adverbs and conjunctions with deictic there, here and where (often in
ambiguous adverbs/conjunctions; symbol ) as their frst element. The follow-
Source domain place/space 113
ing lists are compiled from the OED (on the history of the there-compounds,
see sterman 1997):
(a) thereabout, thereabouts, thereabove, thereafter, thereafterward, thereagain,
thereagainst, thereamong, thereas (conj.), thereat, thereatour, thereaway, there
aways, therebefore, therebeside, thereby, theredown, thereforth, therefro, there
from, theregain, therehence, therein, thereinne, thereintill, thereinto, there-mid,
ther-mid, ther(e)-mide, -mydde, there-nigh, thereof
(b) hereafterward, here-again, here-against, hereat, hereaway, hereaways, herebe
fore, hereby, herefrom, hereft, here-hence, herein, hereof, hereon, hereout, here-right,
hereto, heretofore, heretoforetime, hereunder, hereunto, hereupon, herewith
(c) whereabout (interrog. and rel. adv.), whereabouts, whereafter (rel. adv.), whereas
(rel. adv., conj.), whereat, whereby, wherefro, wherefrom, where-hence, wherein,
whereinne, whereinsoever, whereintill, whereinto, wheremid, whereof, whereon,
whereout, whereso, wheresoever, wheresome, wheresomever, wherethorough,
wherethrough, wheretill, whereto, whereunder, whereuntil, whereunto, whereup,
whereupon, whereever, wherewith, wherewithal
With respect to the adverbial connectors, only the highly frequent therefore has
survived. In Present Day English, the pronominal sense for that (thing, act,
etc.); for that, for it is labelled by the OED as archaic or colloquial (OED,
s.v. therefore, 1.). It is distinguished from the adverbial connectors by a differ-
ent stress pattern (stress on the second syllable vs. initial stress for the adverbial
connector) and thus it is to be doubted whether the spatial origin of therefore is
recognized by speakers of Present Day English.
While therefore, herefore and wherefore thus exhibit a mixed character,
hence, thence and whence, which begin to be used as adverbial connectors in the
Early Modern English period, are purely spatial, and take the spatial extension
of the text as their sole frame of reference. They develop a resultive meaning
modelled on the cognitively central post hoc, propter hoc. Interestingly, here
the proximal hence has survived (in contrast to the distal there in therefore).
2
These differences in the employment of spatial concepts for resultive con-
nectors thus again mirror the general tendencies in the coinage of new adver-
bial connectors in the history of English. While speakers and writers of Middle
English employ essentially pronominal elements, speakers and writers of Early
Modern English refer to the spatial reference only.
2 In many of the occurrences of hence as a linking adverbial, this spatial meaning
is emphasized by a verb which also comprises a spatial element, such as proceed:
For hence it proceedeth that Princes fnd a solitude, in regard of able men to serue
them in causes of estate, because there is no education collegiate (CEEDUC2B,
Bacon, Sample 2, p. 3V).
114 Cognitive source domains
8.4. Source domain truth/fact
8.4.1. General tendencies
The present study distinguishes the domain truth/fact as a third important
source domain. This domain is, however, restricted to the semantic relations
concession/contrast (as a conversational implicature) and, in conventional-
ized form, transition (which is used as the general term for labelling the
semantic relation marked by truth-intensifers, which have not been discussed
in earlier literature).
Table 84: Source domain truth
OE ME EModE LModE
contrast/
concession
{huru certainly, indeed}
{huruinga certainly, indeed}
3
certain
certainly
sure
surely
in truth
transition eornostlice earnestly
solice truly
gewislice certainly
witodlice truly
certainly
certain
certes
forsooth
iwis
sekirly
truly
verily
indeed
sure
surely
in fact
in truth
8.4.2. Conversational implicatures: concession/contrast
In interactive communication, the semantic relation of concession commonly
involves the acknowledgment of a claim. Discourse-functional approaches (see
the various studies by Couper-Kuhlen, Barth-Weingarten and Thompson) thus
see concessivity as a relation of claim acknowledgement of claim coun-
terclaim, which prototypically holds between two turns at a talk. They estab-
lish a cardinal concessive pattern, prototypically represented as
A: x (claim)
B: X (acknowledgement of claim)
Y: (counterclaim)
3 OE huru is of unclear etymology (see Holthausen 1934, s.v. huru); it has survived in
the Swedish subordinator ehuru although.
Source domain truth/fact 115
It is present when a speaker (A) makes a point or claim in one turn and when
a second speaker (B) acknowledges the validity of this claim in the next turn
but proceeds stating that a potentially incompatible point (Y) also holds. In
discourse, but (usually classifed as adversative) is by far the most frequent con-
nective signalling this relation (96 per cent of all cases with an explicit marker;
see Barth-Weingarten 2003: 80).
This acknowledgement of the claim may be signalled (or emphasized) by
epistemic adverbs, which in their propositional meanings denote the concepts
truth or certainty. In German, the conversational implicatures of such ad-
verbs have been conventionalized in the concessive connector PDG zwar (< ze
ware PDG in Wahrheit, in truth; Kluge and Seebold 2002, s.v. zwar).
Yet for an unambiguous expression of concession/contrast, both zwar (ob-
ligatorily) and the not yet fully conventionalized PDG gewiss need a correlative
aber or jedoch but, on the other hand in the second connect. This corresponds
to the situation in all periods of English: In certain contexts, epistemic adverbs
such as sure, surely or of course may have conversational implicatures of con-
cession/contrast. This conceding move expressed by it is true but is
most evident in full sentences such as it is true used in parenthesis, as in (65):
(65) This instinct, it is true, arises from past observation and experience; but
can any one give the ultimate reason, why past experience and observa-
tion produces such an effect, any more than why nature alone should
produce it? (CLHUM1, Part III).
(66) It is true there has been all along in the world a notion of rewards and
punishments in another life, but it seems to have rather served as an en-
tertainment to poets or as a terror of children than a settled principle by
which men pretended to govern any of their actions (CLSERM1A, Swift
On the Wisdom of this World).
The full phrase it is true is very common in this function in Late Modern
English (periods LModE1 and 2; see Appendix B.5.2), both sentence-medially
and sentence-initially. The concessive/contrastive function is, however, also in-
herent over all periods in the adverbs certain(ly) and sure(ly), from ME1 and
EModE1 respectively Example (67) shows that correlatives such as surely
but are, in particular, used as explicit clues for highlighting the relation and
linkage of larger chunks of discourse.
(67) And surely we may pronounce upon it, in the words of St. James, that
This wisdom descended not from above, but was earthly and sensual.
116 Cognitive source domains
What if I had produced their absurd notions about God and the soul? It
would then have completed the character given it by that Apostle, and
appeared to have been devilish too. But it is easy to observe from the
nature of these few particulars that their defects in morals were purely
the fagging and fainting of the mind for want of a support by revelation
from God (CLSERM1A, Swift On the Wisdom of this World).
8.4.3. transition
The adverbs solice (truly; from so truth, true, solic true) and witod
lice certainly
4
are employed in various functions in Old English: on the
phrase level, they may be used as circumstance adverbials, mainly in direct
speech with a frst-person subject (e.g. in phrases such as ic secge solice
I tell [you] truly), or they are employed as emphasizers. Yet, as I have ar-
gued in some preliminary studies for this book (Lenker 2000, 2003, 2007b),
they also have another function which is not commonly noted in dictionar-
ies in Old English prose: as sentence adverbials and eventually adverbial
connectors or discourse markers, they lose much of their original meaning,
extend their scope from the phrase level to at least the sentence level. At the
same time they develop a metatextual function they are used as discourse
markers
5
demarcating episode boundaries on the global level of discourse
and as highlighting devices on the local level of discourse (cf. Lenker 2000).
For lack of a better category, I have classifed these adverbials as transitional
adverbial connectors, since they, similar to PDE now, mainly work on the
textual level in their connector function. In most of their occurrences, they
in line with the properties of adverbial connectors signal a two-place rela-
tion and do not alter the propositional meaning of any of the connects (i.e.
they are non-propositional).
In a recent study, I could show that this development from circumstance
adverbial to connective is not specifc to solice and witodlice, but that ety-
mologically similar adverbs, i.e. those with a basic meaning truly, show the
same functional diversity and development (Lenker 2007b). The fndings of
these studies will be briefy summarized in the following sections (using in
4 Witod is the past participle of the verb witian to order, to decide, which is cognate
with the verb witan to know.
5 I here follow a broad defnition of discourse markers, similar to that of Fraser
(1999), which includes all kinds of connectives which express a two-place relation
(see above, Chapter, 3.6). It excludes, however, items without such a connective
function, i.e. disjuncts such as frankly or briefy or interjections such as oh or yes.
Source domain truth/fact 117
addition to the corpus of the present study also the corpus material of these
earlier studies), and will be related to one of the clines Traugott (1995, 1999;
Traugott and Dasher 2002) has established as a regularity in language change,
6

the so-called adverbial cline from
clause-internal adverbial > sentence adverbial > discourse particle.
In a last step, I will show that this onomasiological approach suggests that
the adverbial cline from referential to connective function is inherent in the
semantics of the lexemes investigated, if we take a pragmatic perspective and
consider communicative principles, in particular Grices Maxims of Quality
(Do not say what you believe to be false; Do not say that for which you lack
adequate evidence; Grice 1989: 2527).
8.4.4. OE solice ME soothly
These adverbs with the basic meaning truly are, to be linguistically more pre-
cise, epistemic modal adverbials expressing the speakers commitment to the
truth of the proposition. Accordingly, they may be in very restricted contexts
be used as circumstance adverbials, most typically with a frst-person subject
and a verb of communication (OE secgan, ME seien, tellen)
7
in the present
indicative form, such as
(68) Nacode he scrydde, and swa ic solice secge, ealle nyd-behfnysse he
ws dlende am e s behofodon (Passion of St Eustace, 9; ed.
Skeat 1900: 190).
6 This cline has been the subject of various papers by Traugott and is summarized in
Traugott and Dasher (2002: 152189). In the last decade, it was considered to be a
subtype of grammaticalization in an approach which might be labelled a seman-
tic-pragmatic approach to grammaticalization, and it was argued that this ad-
verbial cline should be added to the inventories of clines in grammaticalization;
see, for example, the title of Traugott (1995 [1997]). Grammaticalization processes
proper and the cline sketched above indeed agree in several important character-
istics, in particular semantic bleaching, increased subjectifcation and pragmatic
strengthening. These structural similarities are also stressed in a more recent pub-
lication (Traugott and Dasher 2002), but the cline is now no longer regarded as a
subtype of grammaticalization, but as a regularity in semantic change.
7 Of the altogether twenty manner adjunct/emphasizer uses of sothli in Chaucer,
fourteen modify sei(en) or tell(en). In Gowers Confessio Amantis, four of the fve
direct-speech instances are with the communicative verbs say and tell, one with
lieve to believe.
118 Cognitive source domains
The naked he clothed; and, as I truly tell, he distributed to every neces-
sity of them that had need thereof.
(69) And thus I mai you sothli telle, , I am in Tristesce al amidde
(Gower, Confessio Amantis, 4.34963499; ed. Macaulay 19001901).
And so I can truly tell you, , I am completely in the middle of Despair
.
Solice or soothly with a verb of communication can either be used in a paren-
thesis, as in (68), or as a matrix clause, as in (69). These uses provide the basis
for all the later uses of solice as a stance adverbial (style disjunct) and connec-
tor on the textual level: speakers want to stress the assertion of another clause
by explicitly pointing to its truth value.
First, this underlying phrase gives rise to the adverbs use as emphasizers
which may be defned as enhancing the truth value or force of a sentence
(Quirk et al. 1985: 485).
(70) Apolloni, ic oncnawe solice t u eart on eallum ingum wel gelred
(Apollonius 16, 24; ed. Goolden 1958).
Apollonius, I know truly [indeed] that you are well taught in all things.
(71) and if that he / ne hadde soothly knowen therbifoore / that parftly hir
children loved she (Chaucer, Clerks Tale, (4) 688690; ed. Benson
1987).
and if he / had not truly [indeed] known before / that she loved her
children perfectly .
While examples (70) and (71) might be considered ambiguous cases (I know
in a true (certain) way or I [intensifer] know .), an emphasizer use is
clearly suggested by the stative verbs in (72) and (73) and further by the seman-
tic contrast soothly vs. as I gesse in (73):
(72) Ic eom solice romanisc. and ic on hftnyd hider geld ws (Passion
of St Eustace, 344; ed. Skeat 1900: 210).
I am truly a Roman, and I was brought hither in captivity.
(73) But Venus is it soothly, as I gesse (Chaucer, Knights Tale, (1) 1102; ed.
Benson 1987).
But it is indeed Venus, as I guess.
Source domain truth/fact 119
In non-interactional, narrative contexts (i.e. not in direct speech) and thus in
the majority of their occurrences,
8
however, the scope of these adverbs is not
restricted to the phrase level, but extends to the whole sentence.
(74) Ws he solice on rihtwisnysse weorcum swie gefrtwod (Passion
of St Eustace, 4; ed. Skeat 1900: 190).
[transition marker], he was greatly adorned with works of right-
eousness.
(75) Soothly, the goode werkes that he dide biforn that he fl in synne been
al mortefed and astoned and dulled by the ofte synnyng (Chaucer, Par
sons Tale, (10) 232; ed. Benson 1987).
[transition marker], the good works he did before falling into sin are
all nullifed, rendered null and made void by his frequent sinning.
The adverbs are here originally employed as stance adverbials and replace an in-
dependent sentence, a matrix clause, with a different proposition and a specif-
cally frst-person subject solice ic eow secge t (style disjunct) or so
is t ic secge t (content disjunct). This change to sentence adverbial in-
volves increased scope and syntactic freedom:
9
the adverb is no longer part of
the core syntactic structure, becomes syntactically optional and is often found
sentence-initially.
As circumstance adverbials or emphasizers, solice/soothly are primarily
found in direct speech with a frst-person (singular) subject because their propo-
sitional meaning truly, truthfully demands a human agent with high trustwor-
thiness, most likely the speakers themselves. In their sentential use, on the other
hand, there is no constraint on either the subject of the sentence, which can be third
person or even inanimate, or the verb, which can be in all tenses and moods. The
adverbs do, however, retain their epistemic meaning truly in that they introduce
the voice of the narrator or author without directly mentioning her or him and al-
low the speakers to mark their opinion, their attitude or even their faith or trust in
the veracity or importance of the proposition. The adverbs thus show increased
subjectifcation, if this term is understood with Traugott as a process which fo-
cuses on the subject of a discourse and emphasizes subjective valuations.
10
8 In Chaucer, eighty-six of the 114 instances of soothly are found in pieces of narra-
tive prose.
9 This is one of the main reasons why it is diffcult to regard this development as a
case of grammaticalization in the morpho-syntactic sense.
10 For the terms subjectifcation and intersubjectifcation, see Traugott and Dasher
(2002: 1924, 8999). For a completely different conception of subjectifcation, see
Langacker (1999).
120 Cognitive source domains
The fundamental difference between the various uses of solice/soothly and
its use as an adverbial connector lies in the fact that the adverb loses almost all
of its propositional meaning truth as a connector. Solice and soothly can no
longer be considered style or content disjuncts because a paraphrase ic secge
solice or so is t ic secge is impossible in these contexts.
In (76), you fnd the instances which initially made me think about a textual
function of solice when working on the rubrics of the Old English West Saxon
Gospels (Lenker 1997). There, the beginnings of the gospel lections read dur-
ing the Mass, i.e. the beginnings of the episodes, are marked by sentence-initial
solice
(76) (Luke I.26) Solice on am syxtan mone ws asend gabriel se engel
fram drihtne on galilea ceastre
RUBRIC
(Luke I.39) Solice on am dagum aras maria and ferde on muntland
mid ofste. on iudeisce ceastre
RUBRIC
(Luke I.56) Solice maria wunude mid hyre swylce ry monas. and
gewende a to hyre huse
RUBRIC
(Luke II.1) Solice on am dagum ws geworden gebod fram am
casere augusto. t eall ymbehwyrft wre tomearcod (Old English
Version of the Gospels; ed. Liuzza 1994)
Luke I.26 [transition marker] in the sixth month was sent Gabriel the
angel by the Lord to a Galilean town .
Luke I.39 [transition marker] in these days Mary got ready and went
to the hill-country with haste to a Iudeaen town .
Luke I.56 [transition marker] Mary lived with her such three months
and went then (back) to her house. .
Luke II.1 [transition marker] in these days an order was given by the
Emperor Augustus that all the world should be described .
Solice does not show any of its propositional or emphasizer meanings in these
examples the notions of in a true way (circumstance adverbial) or of enhancing
the truth value or force of a sentence as an emphasizer (implied I tell you truly)
are no longer important. Instead, the focus is on the organization of discourse
in which solice functions as a marker of the global text structure: it signals the
beginning of a new episode. It is not employed conceptually, but procedurally.
This procedural function is, however, not only found on the global level of
discourse but also on the more local levels, a textual function especially attest-
Source domain truth/fact 121
ed for ME soothly. In prose texts, soothly or for soothly often follow quotations,
as may be seen from the two examples taken from Chaucers Parsons Tale, an
argumentative piece of prose on the seven deadly sins.
(77) Homycide is eek by bakbitynge, of whiche bakbiteres seith Salomon that
they han two swerdes with whiche they sleen hire neighebores. For
soothly, as wikke is to bynyme his good name as his lyf. Homycide is
eek in yevynge of wikked conseil by fraude For whiche the wise man
seith Fedeth hym that almost dyeth for honger; for soothly, but if thow
feede hym, thou sleest hym (Chaucer, Parsons Tale, (10) 565570; ed.
Benson 1987).
Homicide is also by backbiting, of which backbiters Salomon says
that they have two swords with which they slay their neighbours. For
[transition marker], it is as wicked to take away a mans good name as
his life. Homicide is also in giving wicked counsel by fraud, This is
why the wise man says Feed him who is almost dead from hunger; for
[transition marker], if you do not feed him, you kill him.
(For) soothly here signals the continuation of the original argument and marks
or highlights the authors personal opinion in respect to the quotation.
Other frequent collocations with conjunctions, such as but soothly, and
soothly or now soothly, also testify to its similarity to conjunctions (for the
re levance of these collocations, see below, Chapter 13).
11
Yet, solice/soothly
also reveal this procedural function when on their own. ME soothly is simi-
lar to listing or summative connectors in Present Day English employed to
introduce or to summarize an argument:
(78) as God seith be the prophete Ezechiel, Ye shal remembre yow of
youre weyes, and they shuln displese yow. Soothly synnes ben the
weyes that leden folk to helle (Chaucer, Parsons Tale, (10) 141; ed. Ben-
son 1987).
as God says through the prophet Ezechiel, You shall remember
your ways, and you will not like them. [transition marker], sins are
the ways that lead people to hell.
11 Among the prose instances of soothly, the numbers in the corpus passages chosen
for the present chapter are as follows: for soothly (31), but soothly (16), and soothly
(7) and now soothly (2); soothly alone is used thirty-one times. For the explanatory
force of collocations, see Traugott and Dasher (2002: 168).
122 Cognitive source domains
In sum, OE solice and ME soothly thus follow exactly the path Traugott
(Traugott 1995; Traugott and Dasher 2002) has suggested as the so-called ad-
verbial cline: original manner adverbs come to be used as sentence adverbs
and fnally as adverbial connectors working on the interpersonal or textual
level, signalling the continuation of an argument or highlighting an episode or
a part of the argument. The adverbs which originally expressed the speakers
commitment to the truth of the proposition have become non-propositional.
8.4.5. ME forsooth(e)
A similar pragmatic and semantic history can be seen in ME forsooth(e), which
enters the semantic feld at the beginning of the Middle English period, even
though at that time soothly is still extensively employed. The new item is like
solice/soothly overtly related to the concept of truth (noun and adjective OE
so, ME sooth), but it is not certain whether it should be analysed as a univerba-
tion of the prepositional phrase for soe (see OED, s.v. forsooth) in truth, for
a truth or as a prefxed form of the adjective so modifed by the intensifying
prefx for- very true (cf. OE formicel very great, forwel very well; for a
more detailed discussion see Lenker 2003: 276278 and in particular Lenker
2008).
The collocation is only attested seven times in Old English, but becomes
very frequent from the thirteenth century onwards, mainly as an emphasizer
or intensifer (cf. MED, s.v. forsooth, and Lenker 2003: 274278). In contrast
to the adverbs mentioned above, however, it is also abundantly used in an em-
phasizer function in both positive and negative answers to questions (see also
MED, s.v. forsoth, sense 2b), as in:
(79) Ye/Yis, forsothe, quod I (Chaucer, Boece 3, prosa 11.203; prose 12.57;
ed. Benson 1987)
(80) Nay/No forsothe, quod I (Chaucer, Boece 3, prosa 12.150; 4, prose
2.100; ed. Benson 1987).
Todays negative connotations of forsooth, which is now only employed par-
enthetically with an ironical or derisive statement (OED, s.v. forsooth), are
probably a consequence of the overuse of these formulae. In Early Modern
English, forsooth even came to be considered a superfuous interjection mark-
ing the speech of the lower classes in London (Lenker 2003: 283286) and
is, accordingly, used for such a depiction of characters in Early Modern English
Source domain truth/fact 123
plays. Consider, for example, the overuse of different emphasizers in the fol-
lowing passages from Middletons A Chaste Maid in Cheapside:
(81) (Port.) Yes by my faith Mistris, theres no true construction in that, I
haue tooke a great deale of paines, and come from the Bell sweating.
Let me come tote, for I was a Schollar forty yeers ago, tis thus I war-
rant you: (\Matri\) , it makes no matter: (\ambobus parentibus\), for a
paire of Boots: (\patri\) , pay the Porter: (\amantissimis charissimis\),
hes the Carryers Man, and his name is Sims, and there he sayes true,
forsooth my name is Sims indeed, I haue not forgot all my learning.
A Money matter, I thought I should hit ont (Middleton, Chaste Maid,
CEPLAY2B, 2).
(82) (All.) Verily you are an Asse forsooth, I must ft all these times, or heres
no Musicke, (Middleton, Chaste Maid, CEPLAY2B, 27).
In Middle English, however, the scope of forsoothe could also be wider and
extend over the whole sentence. Accordingly, it is like solice/soothly found
in collocations with conjunctions, such as and and but and is also employed in
the procedural, textual functions sketched for solice/soothly above In (83),
for instance, and forsothe has a summative meaning and indicates the end of
the episode:
(83) And forsothe this foreseide woman bar smale bokis in hir right hand,
and in hir left hand sche bar a ceptre.
And whan she saughe thise poetical muses (Chaucer, Boece 1, prose
1.41; ed. Benson 1987).
And [transition marker] this abovementioned woman bore small
books in her right hand, and in her left hand she bore a sceptre. [New
paragraph] And when she saw the Muses of poetry .
The polysemous character and the various functions of ME forsothe on differ-
ent levels may also be inferred from a contemporary metalinguistic scource,
the Catholicon Anglicum, dated 1483. This English-Latin dictionary provides
the following list of Latin translations for ME forsothe:
Amen, Autem, certe, enim, enion, eciam, equidem, nempe, nimirum, profecte,
quippe, reuera, siquidem, utique, vero, vere, quidem, quoque, porro, veraciter,
quin, quineciam, quinimmo, quinin, veruntamen (ed. Herrtage 1881: 138).
Only a small number of the Latin equivalents (certe, vero, vere, veraciter, ver
untamen) exhibit like forsoothe a connection to the concept of truth.
124 Cognitive source domains
Only some of them may be employed as emphasizers, while most of them
(autem, enim, quidem ) are procedural, i.e. they belong to the group of Latin
adverbs for which Kroon (1995) has established a discourse function. While
vero truly, for instance, is procedural in Latin in that it functions on the inter-
personal level of communicative acts and moves, autem is a connective on the
level of textual organization (Kroon 1995: 371375). This allows us to infer that
a contemporary lexicographer indeed regarded forsoothe as an adverb with a
text-organising function.
8.4.6. OE treowlice ME/EModE/PDE truly
PDE truly (OE treowlice) also follows this adverbial cline, but had to develop
an epistemic meaning in a frst step and is thus more similar to the items em-
ploying the cognitive source domain fact, such as indeed or in fact.
In Old English, the adverb treowlice (cf. treowe adj. faithful) could only be
used as a circumstance adverbial, meaning faithfully, loyally, with steadfast
allegiance (see OED, s.v. truly, 12; MED, s.v. treuli, 15). In Middle English,
we still fnd a number of examples for this conceptual use, as in
(84) He was also a lerned man, a clerk, / That Cristes gospel trewely wolde
preche (Chaucer, General Prologue, (1) 480481; ed. Benson 1987).
He was also a learned man, a clerk / who would loyally preach Christs
gospel .
(85) This preyere moste be trewely seyd, and in verray feith, and that men
preye to God ordinatly and discreetly and devoutly (Chaucer, Parsons
Tale, (10) 1044; ed. Benson 1987).
This prayer must be spoken reliably, and in true faith, so that men pray
to God properly and discreetly and devoutly.
In the General Prologue, Chaucer describes the manner in which the clerk
would preach the gospel as loyally, with steadfast allegiance, and in (85) the
manner in which the prayer is to be said is prescribed by the adverb trewely, but
also by the propositional phrase in verray faith (cf. also the following circum-
stance adverbials ordinatly, discreetly and devoutly).
In a more abstract sense, trewely can also mean accurately (cf. MED, s.v.
treuli, 9), as is evident from the rubric of Chaucers Astrolabe which requires
the instruments to be set correctly:
Source domain truth/fact 125
(86) To knowe the degrees of longitudes of fxe sterres after that they be de-
termynat in thin Astrelabye, yf so be that thei be trewly sette. (Chaucer,
Astrolabe 2.18. Rub.; ed. Benson 1987).
To know the degrees of longitudes of fxed stars after they have been
placed in your Astrolabe, if it is the case that they are accurately set.
Only by the thirteenth century does trewely become semantically epistemic
(cf. OED, s.v. truly, senses 3 and 5; MED, s.v. treuli, sense 11). Its epistemic
force is evident in (87) from the Reeves Tale, since trewely does not primari-
ly designate that the students are going to pay faithfully, but indicates their
subjective perspective (they are not going to pay!) and marks the speech act to
be an (unfulflled) promise.
(87) But specially I pray thee, hooste deere, / get us som mete and drynke,
and make us cheere, / and we will payen trewely atte fulle (Chaucer,
Reeves Tale, (1) 41314133; ed. Benson 1987).
But we ask you in particular, dear host / bring us some food and drink,
and make us merry / and well fully pay for it, thats for sure.
The newly acquired epistemic meaning and its functional truth value is most
obvious in the translation of the Latin Amen, (amen) dico vobis in the New Tes-
tament. These formulaic expressions are used when Jesus reinterprets the Old
Testament by virtue of his authority as the Son of God and they thereby require
an epistemic value. While the Old English texts translate these by so ic (eow)
secge or solice/witodlice ic (eow) secge (West-Saxon Gospels), the Earlier and
Later Wycliffte versions use treuli I seie (see Matthew XXVI.13.21.34; Mark
XIV.9.18.25; Luke XXII.59; John XIII.16.20.21).
At this stage, the functional variability and polysemy of trewely is most
clearly seen in phrases such as loven trewely, in which the adverb may function
as a manner adjunct, emphasizer or sentential adverb In the examples taken
from Gowers Confessio Amantis, only the word order and collocation with the
conjunction for distinguish the uses which in modern editions are then high-
lighted by punctuation, i.e. the comma in (89).
(88) Sithe I have trewly loved on (Gower, Confessio Amantis, 3.66; ed. Ma-
caulay 19001901).
Since I have indeed loved one.
(89) For trewli, fader, I love oon (Gower, Confessio Amantis, 5.2536; ed.
Macaulay 19001901).
For indeed, father, I love one.
126 Cognitive source domains
Finally, trewely is also found in the same collocations as soothly namely with
the conjunctions and, for and but trewely and is also similar in its function
of highlighting the connective function of the coordinators for and and which
it collocates with:
(90) And ye shul understonde that looke, by any wey, whan any man shal
chastise another, that he be war from chidynge or reprevynge. For
trewely, but he be war, he may ful lightly quyken the fr of angre and of
wratthe, which that he sholde quenche (Chaucer, Parsons Tale, (10)
628; ed. Benson 1987).
Understand, in any case, that when a man has to chastise another, he
should beware lest he chide or reproach him. For [transition marker],
unless he be wary, he may very easily kindle the fre of anger and wrath,
which he should quench.
(91) the which a long thyng were to devyse. / And trewely, as to my jugge-
ment / me thynketh it a thyng impertinent / save that he wole convey-
en his mateere; / But this his tale, which that ye may heere (Chaucer,
Clerks Prologue, (4) 5356; ed. Benson 1987).
it would take a long time to describe this thing to you. / And [transi-
tion marker], in my mind / it seems to be an irrelevant thing / save that
he wanted to introduce his material; / But here is his tale which you may
hear (now).
The adverb became very popular in the course of the Middle English period
(Swan 1988: 255), as can be seen in many examples from London English, such as
(92) & I wene ther wer a-boute a xxx craftes, & in Chepe they sholden
haue sembled to go to a newe eleccion, &, truly, had noght the alder-
men kome to trete, & maked that John Norhampton bad the poeple gon
hoom, they wolde haue go to a Newe eleccion (CMDOCU3, The Ap
peal of Thomas Usk, 28).
and I guess that there were about 30 guilds, and in Cheapside they
should have assembled to go to a new election, and, [transition mark-
er], had not the aldermen come to make terms and caused that John
Norhampton bade the people to go home, they would have gone to a new
election.
(93) & that al strange vitaillers sholden with thair vitailles frelich kome
to the Cite, to selle thair vitailles as wel be retaile as in other wyse,
Source domain truth/fact 127
hauyng no reward to the Franchise. And, truly, the ful entent was that
al the ordinances that wer ordeyned in hys tyme (CMDOCU3, The
Appeal of Thomas Usk, 24).
and that all foreign traders in food-stuffs should come with their
victuals freely to the City [London], to sell their victuals by retail-trad-
ing as well as in other ways, having no regard of the freedom from tax
in London. And, [transition marker], the full intention was that all the
commissions that were commissioned in his time .
OE treowlice thus shows an identical line of development as the adverbs dis-
cussed above. The major difference, however, is that treowlice frst had to
change its propositional meaning. It is important to note, however, that with
all of these adverbs the various meanings do not replace each other imme-
diately language change necessarily involves polysemy, and loss of original
meaning is relatively rare (Traugott and Dasher 2002: 280281). Thus the old
and new meanings may coexist in the same text, a property commonly referred
to as layering.
8.4.7. PDE indeed and in fact
As has been mentioned above, the adverbs analysed here are semantically and
functionally similar to PDE indeed and in fact.
12
These lexemes share the ex-
pression of factuality in their original propositional meaning and only acquired
epistemic meaning (veracity) in the course of their history. The development
of indeed from a bare prepositional phrase with propositional meaning (in the
deed/act) to its procedural, textual use whats more has been repeatedly dis-
cussed by Traugott and may be summarized as follows (cf. Traugott and Dasher
2002: 165). In stage II, indeed develops an epistemic meaning and becomes
semantically similar to soothly, forsoothe and truly:
Stage I: PP in action / in practice 13001850 in the act > in truth
Stage II: epistemic uses 1450- in truth > in addition
Stage III: procedural, textual uses 1600- whats more
The adverbial cline is also identifable in the history of in fact, originally a
prepositional phrase with the head fact (borrowed from Latin in the sixteenth
12 I here exclude actually (included in Lenker 2007b) because its functions as a text-
ual connective are not as clear-cut (see also Lenk 1998: 155188).
128 Cognitive source domains
century), which develops procedural uses, frst as an epistemic adversative
and, fnally, as a transitional connector (cf. Traugott 1999; Traugott and Dasher
2002: 165169).
Stage I: adverbial of respect in which (1670)
Stage II: epistemic, adversative uses (1680)
Stage III: procedural, textual uses (1815)
8.4.8. Regularities in change
Apart from the shifts from one semantic domain to another summarized above,
all the lexemes analysed here share the pragmatic-semantic tendencies which
Traugott and Dasher (2002) have established for adverbs with a procedural
function as a regularity in semantic change (Traugott and Dasher 2002: 281).
Table 85: Regularities in semantic change (after Traugott and Dasher 2002: 281)
non-subjective > subjective > intersubjective
content > content-procedural > procedural
scope within proposition > scope over proposition > scope over discourse
truth-conditional > > non-truth-conditional
8.4.9. Truth, facts and communicative principles
In their conclusions, Traugott and Dasher (2002) state that adverbs of an origin-
al meaning in fact, deed, action which refer to factuality seem to provide the
appropriate semantics of a lexeme to undergo the recurrent changes of the
adverbial cline (Traugott and Dasher 2002: 281). I have argued elsewhere,
however, that I would like to go one step further and suggest that the semantics
of these adverbs are not only appropriate but that the original propositional
meaning of these lexemes almost inevitably leads to pragmatic strengthen-
ing, to their use as an emphasizer and, eventually, as a connector (see Lenker
2007b). I rest this view on one of the seminal concepts in pragmatics, Grices
Co-operative Principle and in particular his Maxims of Quality (Grice 1989:
2627):
13
13 For a survey of the principle, the maxims, revisions, applications and problems see
Levinson (1983: 100166; 2000: 1221) and Mey (2001: 6888).
Source domain truth/fact 129
1 Do not say what you believe to be false.
2 Do not say that for which you lack adequate evidence.
Obviously, these maxims are directly connected to the adverbs analysed in this
chapter: while Quality (1) is related to the adverbs meaning truly, Quality
(2) clearly encompasses those with an original meaning of in fact, in deed, in
action.
If we accept the most basic of Gricean assumptions, namely that all partners
in a communicative act want to be cooperative, then the Maxim of Quality
implies that we should only talk about the veracity or factuality of a propos-
ition if speakers/writers or listeners/readers doubt it, i.e. in highly emotional or
negative contexts. Speakers/writers may thus want to mark their speech acts
as performative (see the promise in (87)) or explicitly express their commit-
ment to the truth of the proposition because they fear the non-acceptance of the
interlocutors (cf. the many instances of solice, soothly and verily with verbs of
communication). This gives rise to the conversational implicatures of conces-
sion as shown above in Chapter 8.4.2. Further, the speakers reference to truth
could be cooperative in negative contexts or when they want to mark a hedge,
such as in Im not sure whether it is true but or as far as I know .
In many of the instances quoted above apart from some manner adjunct
and emphasizer uses none of these contexts are given. Stating the veracity
or factuality of a proposition by means of the adverbs in question would thus
seem superfuous and therefore un-cooperative. The recurrent use of these ad-
verbs would hence also impair the Maxim of Quantity (Do not make your
contribution more informative than is required) and the Maxim of Relation
(Make your contribution relevant).
14
Yet, according to Grices hypothesis,
when a conversation does not proceed according to the maxims, hearers/read-
ers assume that, contrary to appearances, the principles are nevertheless being
adhered to at some deeper level (cf. Levinson 1983: 102).
In cases of this sort, inferences arise to preserve the assumption of co-oper-
ation. Relying on the Maxim of Quality, listeners/readers will principally as-
sume that speakers are telling the truth. If speakers/writers choose to assert the
veracity or factuality explicitly by using a sentential adverbial such as solice,
forsoothe, trewely, indeed or in fact, the invited inference of the hearer/listener
is that the meaning of these items stating the truth or factuality has to be found
on another than the propositional or lexical level.
14 This Maxim of Relevance (Be relevant) is considered the only important one in
the cognitively oriented approach by Sperber and Wilson (1996). For a survey and
criticism of the concept, see Levinson (2000: 5464).
130 Cognitive source domains
The frst level which is to be expected for these truth-intensifying adverbs
are their concessive/correlative uses (in the correlative patterns surely but)
or their epistemic use implying increased subjectifcation, when speakers and
hearers agree in their conceptualization of the adverbs as emphasizers: the
speaker wants to stress the assertion of another phrase or clause by explicitly
pointing to its truth value. In case this level does not provide the expected
meaning in accordance with the Co-operative Principle because the propos-
ition and the grammatical context (inanimate subject, subjunctive mood) will
not allow a meaning of truth or factuality, hearers will fully understand
that the speakers commitment to truth does not relate to the proposition but
again to a different, deeper level, i.e. to the organization of discourse itself.
The adverbs are employed and will be understood as connectors, because oth-
erwise the interlocutors would violate the Maxim of Quality.
15
Once they are
predominantly interpreted on this level of textual organization, however, they
have lost much of their force as truth-intensifers on the manner-adverb level
and have to be replaced by new lexemes, which may in due time follow the
same cline. In sum, I would like to suggest that in addition to the source do-
mains time and space truth/fact should be added as a third domain from
which adverbial connectors are recurrently coined.
16
15 Levinson (1983: 100101) deduces the notion of cooperation from general consid-
erations of rationality which might consequentially mean that we are dealing with
a universal here; cf., for example, Latin concessive verumtamen nevertheless.
16 For the differences which distinguish these transitional connectors from those cod-
ing semantic relations, see below, Chapter 12.
9. Shifting deictics in the history of English causal
connection
In the following Chapters 9 to 11, a closer inspection of the development of ad-
verbial connectors of the different semantic relations will show that adverbials
for the different relations were affected in different ways. First of all, we have
to draw a distinction between what the Cambridge Grammar calls pure vs.
impure connectors, i. e. those connectors which have no other function than
that of connecting their clause to the surrounding text in contrast to those which
combine this connecting function with a function of concession, condition
or reason/result (Huddleston and Pullum 2002: 775). This distinction cor-
responds to the group of relations which has been established as a typical goal
domain above (see Chapter 8) and which is singled out as the set of cccc-rela-
tions in cross-linguistic, typological, and grammaticalization studies (cause,
condition, concession, contrast). For the adverbial connectors, however,
only cause/result (present chapter) and concession/contrast (Chapter 11)
are relevant.
For reasons of information structure and hierarchy (see above, Chapter 2.4),
the relation of condition is also cross-linguistically not signalled by adver-
bial connectors, which express an equal weight of both connects (and not the sub-
ordinate-superordinate hierarchy inherent to if then relations). The semantic
relation of addition (Chapter 11) and the marginal one of transition (Chapter
12), on the other hand, are for the same causes not expressed by subordination,
since they as pure coordinators merely signal the additive relation of sentences
or chunks of discourse explicitly. Additional connectors thus only highlight a re-
lation already present and are not obligatory for an understanding of the texts.
They are signposts facilitating the reading process in registers such as academic
prose, specifcally trying to mark all logical relations overtly. Adverbial connec-
tors signalling the relations cause/result and contrast/concession, on the
other hand, are much less genre-dependent (see above, Chapter 3.5).
These basic differences are mirrored in the developments of the various ad-
verbial connectors for the specifc relations: While the changes in additional
and transitional connectors mainly concern replacement of lexical elements and
refect changing stylistic or rhetorical ideas (see below, Chapters 11 to 13), the
developments in the ccc-relations are much more deeply rooted and refect the
typological and structural changes pointed out above. This will frst be shown
by a closer inspection of the developments in the semantic register of cause/
result, which prototypically shows all of the tendencies sketched so far: the
132 Shifting deictics in the history of English causal connection
path from one central pronominal connector, which is an ambiguous adverb/
conjunction (OE form), to a differentiated system of conjunctions and adver-
bial connectors from Early Modern English onwards. It also refects the shift of
the expression of discourse deixis from pronominal to, predominantly, spatial
frames of reference.
9.1. Earlier research on causals
The following exemplary investigation provides an analysis of the most fre-
quent connective relation in texts, the relation of cause/result/inference
(for the numbers in Present Day English, see Biber et al. 1999: 880; Figure
10.25).
1
The present study aims to illustrate that a detailed investigation of the
instabilities in the systems of Middle and Early Modern English causal con-
nectors allows for a better understanding of two peculiarities of the system of
Present Day English causal connection and its history, namely the remarkable
lack of an adverbial connector marking the relation result

cause (cf. Latin

nam, enim or German nmlich, deswegen) and, secondly, the late triumph of
the now central causal connector because after 1750. The focus will not be
on single connectors (as because in Higashiizumi 2006), or on morphologi-
cal (as in Liggins 1955; Kivimaa 1966; Rissanen 1997) or genre aspects (as in
Claridge and Walker 2002), but on systematic changes in the forms of clausal
connectors after the Old and Early Middle English period which, in specifc
contexts, triggered the coinage of new causal connectors to avoid ambiguities
and facilitate information processing.
The focus of the chapter is therefore on a closer inspection why some of the
forms emerging in early Middle English, such as the adverbial connectors for
whi (ME1 to ME3) or for that (ME1 to ME3) or the subordinator for as much as
(ME2 to EModE2) were rejected while others, such as hence or because, have
survived.
2
The changes specifc to English become particularly evident when
we compare the various systems of causal connectors in the history of English
with Present Day German and consider the implications of Modern German
1 In contrast to most other relations in clause linkage, expressions for causality have
been examined quantitatively in a number of synchronic (see Altenberg 1984, Biber
et al. 1999) and diachronic corpus studies (see Liggins 1955 for Old English; Ris-
sanen 1997 for the history of causal connectors in English; Markus 2000 for Middle
English and Claridge and Walker 2002 for Early Modern English).
2 This chapter is a slightly revised and updated version of Lenker (2007a). I
would like to thank John Benjamins for granting permission to print this chapter in
revised form.
Causal connectors 133
and Old English but not Old English and Modern English being typologi-
cally close to one another.
Such an approach requires the close examination of causal connectors and
their textual functions in comparable texts. In addition to the material for ad-
verbial connectors collected from the corpus texts in Appendix B.3, this chap-
ter thus investigates all causal connectors i. e. not only adverbial connectors,
but also coordinating and subordinating conjunctions in the singular argu-
mentative text extant for all periods of English, the adaptations and transla-
tions of Boethius De Consolatione Philosophiae in King Alfreds Circle (OE;
ed. Sedgefeld 1899), by Chaucer (ME; ed. Benson 1987), Colville (EModE;
ed. Bax [1556] 1897), Queen Elizabeth I (EModE; ed. Pemberton 1899) and
Preston (1695) (EModE) and in Present Day English and German translations
(Watts 1969 and Neitzke in Grasmck 1997), in the representative prose se-
lections of the Helsinki Corpus (HC). These fndings are cross-checked and
supplemented by an analysis of causal connection in Byrhtferths Old English
Enchiridion (ByrM; ed. Baker and Lapidge 1995), a text relatively independent
from Latin, as well as by information elicited from dictionaries (DOE, MED)
and machine-readable corpora, in particular the Dictionary of Old English
Corpus (DOEC) and the Corpus of Middle English Prose and Verse (CME).
9.2. Causal connectors
9.2.1. The relation cause: cause result vs. result cause
The category of cause can be further split into causal relation on the one
hand (PDE because, since) and the relation of result (PDE therefore, so
that) on the other hand. First, it is important to note that all connectors form
complex propositions. Thus a sequence of two propositions (94) proposition
A: John is ill and proposition B: John wont come tonight becomes a com-
plex proposition if a sequence of two propositions A and B expresses a new
thought on a level other than that of the isolated propositions (Rudolph 1989:
176). In contrast to asyndetic sequences which need not, but may express a
causal relation, such as (94a) John wont come tonight He is ill or (94b) John is
ill He wont come tonight, causal connectors commonly have a two-fold func-
tion: they a) connect two states of affairs, and at the same time, b) convey the
speakers opinion on the confguration of these state of affairs, as in
(95) John wont come tonight because he is ill. cause
(96) John is ill so that he wont come tonight. result
134 Shifting deictics in the history of English causal connection
While the frst complex sentence (95) marks a causal relation, (96) reverses
the sequence of information and relates the same state of affairs by marking
the relation as result. The same relation of result can also be expressed by
the employment of an adverbial connector, as in
(97) John is ill. Therefore he wont come tonight. result
In Present Day English (and also in Old English, see below, Tables 9.6 and 9.7),
there is a (slight) preference for the linear sequence cause result by both
speakers and writers (55 per cent and 52 per cent; see Altenberg 1984: 52).
9.2.2. Present Day English causal connectors: corpus fndings
In Present Day English, because (with its reduced forms cos/coz) is by far the
most common causal connector, marking over a third of all relations of cause/
result (see Altenberg 1984: 4045 on the basis of the LOB and the LLC, and
also Biber et al. 1999: 836, 887). With 45 per cent of all instances in the LLC
(Altenberg 1984: 45), because is by far the most frequent connector in the
spoken medium.
In contrast to this stereotyped coding in todays spoken English, the relation
cause is marked in a lexically and grammatically much more varied form in
the written medium, ranging from asyndesis (Sue wont come tonight She is
ill) to syndesis by a variety of explicit linkers. In addition to conjunctions (PDE
because, for, since or as), we fnd adverbial links in the form of prepositional
phrases (PDE for that reason) or adverbial connectors (PDE therefore, hence)
as well as a number of clause-integrated expressions such as PDE the reason
is or the result is. Of the altogether ninety-eight lexical, grammatical and po-
sitional subtypes recorded in LOB and LLC (Altenberg 1984: 39), only subor-
dination by conjunctions (53 per cent) and adverbial connectors (31 per cent)
are highly frequent (in contrast to 8 per cent each for prepositional phrases
and clause-integration; Altenberg 1984: 4044; see also Biber et al. 1999: 842,
887). For this reason, the present chapter will concentrate on the two central
ways of causal connection, conjunctions and adverbial connectors, and will not
further consider lexical means or non-univerbated prepositional phrases.
9.2.3. Causal connectors: word classes and topology
The major means of expressing the relation cause result (examples (98),
(99), (101a), (102a), (103a)) and result cause (examples (100), (101b), (102b),
Causal connectors 135
(103b)) are illustrated in Table 9.1 not only by Present Day English but also by
Present Day German examples, because German is typologically much more
similar to Old English than Present Day English and, more importantly, be-
cause German employs means which are no longer used or common in Present
Day English (see also above, Chapter 2, Table 2.2).
German uses adverbial connectors coding the relation result (R) cause
(C) (PDG nmlich; see (100)) and correlative constructions (see (103a), (103b)),
which are not generally common on the sentential level in Present Day English
(*Because we are stronger, we will/are going to therefore win; *We will/are
going to therefore win, because we are stronger). Since all of these patterns
were possible in Old English (see below, Table 9.5), Table 9.1 thus also attests
to changes and subsequent gaps in the system of causal connection in the his-
tory of English.
Table 91: Causal connectors in Present Day English and Present Day German
A. Parataxis
(98) Sie ist strker Deswegen wird sie gewinnen [V2; post-posed]
CR
(99) She is stronger. Therefore she is going to win [post-posed]
CR
(100) Sie wird gewinnen Sie ist nmlich strker [V2; post-posed]
RC
B. Hypotaxis
(101a) Weil sie strker ist, wird sie gewinnen [V-fnal; pre-posed]
CR
(102a) Because she is stronger she is going to win [pre-posed]
CR
or
(101b) Sie wird gewinnen, weil sie strker ist [V-fnal; post-posed]
RC
(102b) She is going to win because she is stronger. [post-posed]
RC
C. Correlatives
(103a) Weil sie strker ist, deswegen wird sie gewinnen [V-fnal + V2]
CR
(103b) Sie wird deswegen gewinnen, weil sie strker ist [V2 + V-fnal]
RC
The acknowledged traditional criteria for distinguishing these different types
of connectors are topological: the position of the connector, the sequence and
position of the respective connected elements, the possibility of collocations of
connectors and, in German, word order. In contrast to Modern English, Pres-
ent Day German also differentiates main clauses from subordinate ones by
employing verb-second for the former and verb-fnal word order for the latter
(for details, see above, Chapter 2.3).
The focus on topological criteria such as word order, sequence and colloca-
tions is still persistent in most grammars and publications on the subject, as, for
136 Shifting deictics in the history of English causal connection
instance, in the Handbuch der deutschen Konnektoren (Pasch et al. 2003), which
considers topological criteria only (see, e. g., the terms Postponierer postpon-
ers or Verbzweitsatzeinbetter V2-embedders). For Present Day English, Quirk
et al. (1985) establish six, predominantly topological, parameters similar to the
ones illustrated in Chapter 2.3 above; these yield a coordination subordin-
ation gradient (Quirk et al. 1985: 13.18). In addition to three discrete poles
coordinators (such as and and or), conjuncts (adverbial connectors such as
however and therefore) and subordinators (such as if or because) there are also
forms, among them the causal connectors for and so that, which are situated in
the middle of this gradient, sharing three parameters with paratactic and three
parameters with hypotactic connectors. This has led to much uncertainty or even
confusion in particular about the status of PDE for, which is classifed as a subor-
dinator rather than a coordinator but is said to be more coordinator-like than the
more typical subordinators if and because (Quirk et al. 1985: 13.19), mainly
on the grounds that clauses with initial for unlike those headed by because and
since are restricted to post-position in Present Day standard English.
3
Such a
discussion of the niceties of the coordination-subordination gradient turns out to
be rather elusive (see also section 9.3.1) and does not really get to the heart of the
matter, since a number of recent publications on causal connection have shown
that an analysis of connectors should choose a wider perspective and should, in
addition to topological and syntactic criteria, also consider aspects of semantics,
pragmatics and text linguistics (see, for example, Thompson 1985; Diessel 2005
and in particular the studies collected in Lefvre 2000).
9.2.4. Semantic and pragmatic parameters
For the distinction of subgroups of causal clauses, a number of semantic and prag-
matic parameters are employed. One set of distinctions differentiates three major
semantic-pragmatic sub-groups, separating external reason clauses (cause based
in the external reality) from internal ones (cause based in the speakers world
of reasoning) and rhetorical ones (cause based in the speech act) (for these dis-
tinctions see, e. g., Halliday and Hasan 1976: 238244; Quirk et al. 1985: 15.20,
15.4515.47; Sweetser 1990: 7684; Traugott 1992: 252; Higashiizumi 2006).
3 Thus Altenberg (1984: 41, footnote 3) says he follows Quirk et al (1985) in clas-
sifying for as a subordinator. Kortmann (1997: 331), on the other hand, says that for
qualifes as a coordinator in current English. For such a highly frequent element
as for, these differences in classifcation obviously cause major problems for the
comparative value of quantitative data for coordination and subordination (see also
Rissanen 1997; Claridge and Walker 2002).
Causal connectors 137
Because, the central causal connector of Present Day English, may be em-
ployed for all three subtypes: First, because may give an inherent objective
connection in the real world, such as, for example, physical causes and their
effects (external reason clauses marking real-world causality), as in
(104) The fowers are growing so well because I sprayed them.
Secondly, because may also like PDE since, for, as and now that give the
speakers inference of a connection and signal their way of presenting argu-
ments (internal reason clauses; epistemic because), as in
(105) He must be here because his bicycle is outside (meaning The reason I
think he is here is that his bicycle is outside).
Thirdly, the reason given need not be related to the situation in the matrix
clause but is a motivation for the implicit speech act of the utterance (rhetori-
cal or speech act because), as in
(106) Percy is in Washington, because he phoned me from there (meaning
The reason why I say this is that he phoned me from there).
While because may be employed in all three functions in todays English,
other conjunctions are more restricted in their use: PDE since and as are in-
ternal explanation causals, as can be seen from the fact that they do not al-
low why-questions or cleft-sentence constructions (for this terminology and the
criteria, see Quirk et al. 1985: 15.2015.22, 15.4515.47). Similarly, for also
functions internally as explanation and ground rather than assertion of a true
causal relation in the external reality (Quirk et al. 1985: 15.45).
Many languages are more rigid in specifying these different functions, at
least in the written standard, obligatorily by distinctive connectors (cf. Kroon
1995: 1017). For the present investigation of causal connectors in translations
of Boethius De Consolatione Philosophiae, the data for Latin and French (one
of Chaucers exemplars by Jean de Meun) are of particular interest:
Table 92: Cross-linguistic taxonomy of causal relations
external (sociophysical) internal (epistemic, conversational)
explanation justifcation
Latin quia; quod quoniam nam; enim
French parce que puisque car
German weil da denn; nmlich
PDE because because; because
since, as; for
138 Shifting deictics in the history of English causal connection
9.2.5. Information processing
With respect to the states of affairs, there is no difference whether the relation
of causality is expressed by a causal relation (95) or a relation of result as in
examples (96) and (97); (the examples are repeated here for the sake of clarity).
(95) John wont come tonight because he is ill. cause
(96) John is ill so that he wont come tonight. result
(97) John is ill. Therefore he wont come tonight. result
While the causal constant AB remains the same in all cases, the differences
between the various means of marking causal connection are to be found in the
speakers choice of information structure and therefore in the intended and/or
highlighted aim of the message. In the sentence highlighting the causal rela-
tion (95), the speakers interest is focussed on the frst connect John wont come
tonight, refecting the main information and single focus of this sentence. This
information is thus presented in the main clause. As for the relation of result
(96), the speakers interest is focussed on John is ill. Therefore this information
is presented in the main clause, from which the speaker looks to the result B
(He wont come tonight).
These differences are highly important for the interpretation of complex
sentences, but even more so for the general organization of the text (for details,
see also above, Chapter 2.4). Since both these complex sentences only have
one focus, subordinate clauses are not only syntactically but also semantically
subordinate (see Quirk et al. 1985: 13.3). They thus work on a local level of
textual connection, but with different functions for pre- and post-posed subor-
dinate clauses. Initial clauses present given information an interpretation
which is often emphasized by marking cohesion also lexically; cf. hillsides and
hilly in (107) and state a problem within the context of expectations raised by
the preceding discourse:
(107) The houses were perched precariously up the hillsides <> Because it
was so hilly the area seemed constantly to be in a dark blue haze. (ex-
ample from Biber et al. 1999: 835, FICT).
Final adverbial clauses (cf. (96)), on the other hand, play an even more local
role and state the cause for the action named in the immediately preceding
clause. In a recent study, Diessel (2005) shows that with regard to parsing or ut-
Causal connectors 139
terance planning, complex sentences are easier to process and thus preferred if
the adverbial clause follows the main clause. Only rarely, discourse pragmatic
impulses override this processing preference for fnal position. In Diessels cor-
pus, only scientifc articles exhibit a substantial number of pre-posed causal
clauses, causal clauses in this genre being often used to provide a common
ground for a subsequent conclusion (Diessel 2005: 465). Across all genres,
however, most of the causal clauses follow the main clause: only 1 per cent of
all causal clauses in conversation are initial, 10 per cent in fction, and 27 per
cent in scientifc writing (Diessel 2005: 454).
As we might have expected from these considerations, the numbers are very
similar for Old English: if we include correlative constructions, the Old English
adaptation of Boethius Consolatio attests ca. 12 per cent pre-posed clauses
(thirteen out of 109 instances; see below, Table 9.6). The text chosen as an
example for scientifc writing in Old English, Byrhtferths handbook on as-
tronomy, has about 21 per cent pre-posed clauses (ffteen out of seventy-one
instances; see below, Table 9.7).
In contrast to the single focus of complex clauses comprising a subordinate
causal clause, the result construction by means of an adverbial connector in
(97) consists of two main clauses (John is ill Therefore he wont come tonight)
and hence has two foci, i. e. two separate information blocks. In this view,
adverbial connectors are in contrast to subordinators strong indicators of a
great illocutionary weight of the second connect, whose proposition may then
be pursued in the following discourse. This independent focus may be signalled
by syntactic means, such as verb-second (main clause) word order in Present
Day German and, more generally, by prosodic means (see above, Chapter 2.4).
Subordinate clauses, on the other hand, are syntactically and prosodically inte-
grated: the whole complex sentence only has one intonation contour for only one
thematic structure (see Wegener 2000: 36, and Givn 2001: 327 for English).
To avoid ambiguities, the connection by means of an adverb marks the rela-
tion most prominently on the surface level, and is thus highlighted by transpar-
ent lexical adverbial connectors such as PDE consequently or PDG nmlich or
by means of transparent deictic pronominal adverbial connectors such as PDG
deshalb for; therefore (PDG article, genitive singular. masc. des + lexical
element halb because of; cf. PDG halber for the sake of). Because of this
strong independent illocutionary weight, adverbial connectors may not only be
used to link clauses but also whole chunks of discourse.
In this context, two patterns which may cause ambiguity are particularly
relevant for the history of causal connectors in English. Firstly, pre-posed
subordinate clauses which state given information might in the spoken
medium be easily mistaken for post-posed ones or for adverbial connectors
140 Shifting deictics in the history of English causal connection
introducing a causally related new main clause (and are probably for this rea-
son rather infrequent). Similarly, adverbial connectors also have to be as un-
ambiguous as possible to signal overtly that the following main clause is on
some level connected to the preceding clause or discourse and forms a complex
proposition (external reason clause) or is an explanation or justifcation (in an
internal causal relation); adverbial connectors have to be as unambiguous meta-
linguistic indicators of the text structure as possible, since they explicitly state
the voice of the speakers, their opinion on the state of affairs. The following
account will illustrate that changes in the system of English causal connectors
have been induced by exactly these factors of information structure (see the
emergence of for as moche as, since and because) and may themselves, on the
other hand, also lead to changes in the options of information processing (see
the restricted use of for in Present Day English).
For a frst survey, the central causal conjunctions and adverbial subordina-
tors of Present Day English are summarized in Table 9.3. As has been men-
tioned above, Present Day English remarkably in contrast to most other lan-
guages lacks an adverbial connector for the sequence result cause (cf.,
for instance, Latin nam and enim or German denn or nmlich).
Table 93: Present Day English connectors: cause/result
Conjunctions cause: external true reason clauses: because
internal explanation clauses: since, as;
for; now that
Conjunctions result: so that
Adverbial Connectors result: accordingly; consequently; hence; so;
then; therefore; thus
9.3. OE form, foron, fory
9.3.1. Forms and functions of form, foron, fory
When we compare the Present Day English system of causal connectors with
the system in Old English, it becomes evident that all Present Day English
connectors are new coinages or developments in the history of English. Old
English has only one central, highly polysemous and polyfunctional connector
marking the semantic relation cause or reason, which appears in the forms
form (preposition for + m dative singular masculine/neuter demonstra-
tive pronoun), foron (preposition for + on instrumental singular masculine/
neuter demonstrative pronoun) and fory (preposition for + y instrumental
singular masculine/neuter demonstrative pronoun).
OE form, foron, fory 141
The Dictionary of Old English lists these three forms as variants in one
single entry and counts altogether about 15,500 occurrences in a wide variety
of spellings, which do, however, not carry distinguishing force (see DOE, s. v.
for-m, for-on, for-).
4
The item for-m (e) is in all these orthograph-
ic variants found in slots which in Present Day English are flled by adverbs
(therefore; A, examples (108) and (109)) or conjunctions (B; examples (110)
and (111)), but which could also be realized by so-called correlative construc-
tions (C; examples (112) and (113)). Table 9.4 (structured in A, B, C to al-
low comparison with Table 9.1) gives a frst survey with prototypical examples
taken from Byrhtferths Enchiridion (ByrM; ed. Baker and Lapidge 1995).
Table 94: Functions of OE form
A: CAUSE RESULT (cf. PDE Adverbial Connector therefore)
(CAUSE) Foron VSA (= RESULT)
(108) (Se solice by niwe geenda xxix on v kalendas Septembris). Foron
by niwe mona on IIII kalendas September .
Therefore is new moon on 29 August (ByrM 2.2.137).
(CAUSE) Foran SVA (= RESULT)
(109) as ing we gemetton on Ramesige urh Godes miltsigendan gife. Foran ic
ne swigie for ra bocre getingnyssum ne for ra gelredra manna ingum
e .
We found these things at Ramsey through Gods merciful grace. Therefore I
shall not be silent either on account of the eloquence of the literate or for the
sake of those learned men who (ByrM 1.1.158).
B. RESULT CAUSE (cf. PDE Conjunction because/for/since/as)
(RESULT) foran SOV (= CAUSE)
(110) as ing we swa hwonlice her hrepia on foreweardum worce foran we hig
encea oftor to hrepian and to gemunanne.
We discuss these things so briefy at the beginning of this work because we
intend to discuss and recall them more often (ByrM 1.2.250).
(RESULT) foron e SVA (= CAUSE)
(111) on am feoran geare he hf nigon and twentig, foron e an dg awyxst
binnan feower wintrum .
in the fourth year it has twenty-nine, because one day grows over four years
(ByrM 2.1.17).
4 Rissanen (1997: 393) speaks of at least eight different forms. Since he does not
really fnd consistent chronological or dialectal criteria for their distinction, I here
follow the DOE, Mitchell (1985: 30103051) and Traugott (1992: 252254) in
regarding these forms as variants of one single connector.
142 Shifting deictics in the history of English causal connection
C. Correlative Construction: RESULT CAUSE, CAUSE RESULT
foron SOVA (=RESULT) foram SVC (= CAUSE)
(112) Foron Romani hine gelogodon on issum mone (t ys on Februario)
foram he ys scyrtest ealra mona .
The Romans placed it [therefore] in this month (in February) because it is
the shortest of all months (ByrM 2.1.36).
foron SVC (= CAUSE) foron SVO (= RESULT)
(113) and Februarius, foron he ys se lsta and he hf twegra daga ls onne
a ore, foron he forlt <eahta> and feowertig tida.
and February, since it is the shortest and has two days fewer than the oth-
ers, looses therefore forty-eight hours (ByrM 2.1.372).
These instances illustrate one of the most striking features of causal connec-
tion in Old English: the syntactic distinctions between coordination and sub-
ordination as well as the semantic distinctions between internal and external
reason clauses are diffcult, if not impossible, to establish on morphological or
syntactic grounds (Mitchell 1985: 30073051; Traugott 1992: 252). There are
a number of studies which have tried to establish different criteria for coordi-
nation and subordination on a large textual basis (see, e. g., Liggins 1955, who
analyses all Old English prose texts), but all of these fnally had to conclude
that, apart from a few tendencies observable in certain authors, Old English
did not draw a distinction between coordinate and subordinate causal clauses
(Liggins 1955: 205; Mitchell 1985: 30133015; Stockwell and Minkova 1991;
Donoghue and Mitchell 1992), neither by a consistent use of different con-
nectors or different forms of only one connector, nor by differences in word
order (verb-second vs. verb-fnal), nor by the appended subordinating particle
e (see, in particular, Kivimaa 1966: 157).
5
Further, Anglo-Saxon linguists themselves do not seem to have felt a
strong need to differentiate between coordination and subordination. In his
grammar, lfric makes no attempt to distinguish coordination from subordi-
nation and thus draws no distinction between coordinators, subordinators and
adverbial connectors in his section De Coniunctione (Zupitza [1880] 2001:
257266; see also above, Chapter 4). Traugott rightly summarizes this dilem-
5 The number of instances of form e (i. e. with addition of the particle e) increas-
es towards the end of the Old English period, also showing a higher correlation of
form e with subordination (particularly in the writings of lfric). This could
be seen as a developing consciousness of subordination (Rissanen 1997: 394), but
there is still such a wide functional variety that it cannot be said to be decisive for
Old English in general.
OE form, foron, fory 143
ma of anachronistic classifcation of causal connectors by pointing out that it
is usually assumed that the because-clause is subordinate in OE, largely be-
cause the equivalent clause-type in PDE is subordinate (Traugott 1992: 253).
Form and its variants thus belong to those Old English items which are
traditionally called ambiguous adverbs/conjunctions (see above, Chapter
5.3). Except in a few cases (such as answers to questions), there is no way
of distinguishing the various functions on formal grounds on the basis of
our extant Anglo-Saxon material, because this in contrast to material from
later periods does also not provide orthographic clues such as capitaliza-
tion or punctuation or indications of the prosodic differences mentioned above
(though there were most certainly such distinctions in spoken Old English;
see Mitchell 1985: 3015). In sum, there are no unambiguous criteria such as
individual forms of the connector or word order: form and its variants may
be employed in all kinds of sequences for a number of different relations. In
contrast to the wide variety of forms of Present Day English, Old English thus
virtually only employs one form to mark the various causal relations.
6
Table 95: Distribution and functions of OE form
result cause cause result
R for m (e) C
for m R (e) C
for R m (e) C
for m R for m (e) C
C form R
form (e) C R
C for R m
9.3.2. Expressions for causal relations in Early and Late West Saxon
For a survey of quantitative fndings showing the central status of the pronomi-
nal form and its variant forms, I will now briefy summarize the fndings for
all items marking causal relations in two Old English texts which are compara-
tively independent of Latin originals, namely the Early West Saxon adaptation
of Boethius De Consolatione Philosophiae (COBOETH), thought to be writ-
ten in the circle around King Alfred (end of ninth century; for serious doubts
about King Alfred as the translator, see Godden 2007) and Byrhtferths Late
West Saxon Enchiridion (ByrM), a handbook mainly on astronomy (around
1000, after 996; see Baker and Lapidge 1995: xxvi).
6 The following table is compiled from Mitchell 1985: 30103051, Traugott 1992:
252255, Wiegand 1982 and my own material.
144 Shifting deictics in the history of English causal connection
In the passages of the Old English Consolatio selected by the Helsinki Cor
pus (ca. 10,000 words), we fnd altogether 109 forms of causal connectors. In
addition, there are fve instances of lexical expressions and eight of inferential
onne then in post-frst-position (see above, Chapter 5.4). These results are
comparable to the fndings for Present Day English (see section 9.2.2): the rela-
tion of cause is only very rarely expressed by full lexical phrases or preposi-
tional phrases such as be m inge for this reason:
Table 96: Causal connection in the Old English adaptation of Boethius
De Consolatione Philosophiae
cause result (21 forms/variants)
Sum
simple forms
fory 4
form 14
(pre-posed) form 1 19
e 1 1
correlatives
forme form 1 1
result cause (88 forms/variants)
simple forms
form 44
form e 23
foron e 1
fory 1 69
nu 4 4
correlatives
fory form 2
fory fory 1
fory e 2
fory y 1
fory for y e 1
form e 1
form form 2
forme form 1
forme forye 1 12
nu nu 2 2
y y 1
y e 1 2
Other means (5 forms/variants)
be y for this; hence 4
for m inge 1 5
onne 8 8
OE form, foron, fory 145
This survey shows that all of the forms of form (e) are employed for the var-
ious subtypes of the sequences cause result and result cause (form
t is not included here because it is only used for purpose clauses; four in-
stances). With altogether 100 of the 109 instances, variants of simple or correla-
tive form amount to ca. 92 per cent of all causal connectors (if we include the
ambiguous cases of temporal-causal onne and the lexical/prepositional means
of clausal connection, we still get ca. 82 per cent). The only other construction
which is used more than once (in ca. 5.5 per cent of all cases), is nu, also either
in simple (114) or correlative constructions (115; see also above Chapter 5.2.2):
(114) Se ilca God is, swa swa we r sdon, t hehste good & a selestan
gesla, nu hit is openlice cu t a selestan gesla on nanum orum
gesceaftum ne sint buton on Gode (COBOETH, 34.84.3).
The same God is, as we before said, the highest good, and the best hap-
piness since it is evidently known that the best felicities are in no other
things but in God.
In the correlative construction (115), the causal interpretation of nu is strength-
ened by the incongruity of a temporal reading of nu now and onne then. The
givenness of the pre-posed causal clause is highlighted by the deictic reference
to the preceding text by a verb of communication (witan know in u wast):
(115) a cw he: Nu u onne wast hwt a leasan gesla sint & hwt
a soan gesla sint, nu ic wolde t u leornodest (COBOETH,
33.78.27).
Now (that) you know what the false goods are, and what the true goods
are, (now) I would like that you should learn .
Other causal connectors which are commonly also listed in surveys compiled
by a dictionary-cum-grammar method (see Kortmann 1997: 331) turn out to
be fairly infrequent. For one out of three examples using the instrumental form
y of the demonstrative, see the correlative construction in (116):
(116) y hi seca anwald & eac eall oru good e we r ymb sprcon, y hi
wena t hit sie t hehste good (COBOEHT, 34.88.4).
Therefore they seek power, and also the other goods, which we before
mentioned, because they think that it is the highest good.
These fndings for Early West Saxon basically agree with the results for causal
connectors in an also rather independent Old English text from the Late West
146 Shifting deictics in the history of English causal connection
Saxon period, Byrhtferths Enchiridion (complete Old English text examined;
altogether seventy-one causal connectors in ca. 20,000 Old English words).
Table 97: Causal connectors in Byrhtferths Enchiridion
cause result (30 forms/variants)
Sum
simple forms
foron/foran 15
(pre-posed) foron 1 16
(pre-posed) nu 3 3
correlatives
foron foran 2 2
nu nu 9 9
result cause (41 forms/variants)
simple forms
foron/foran 35 35
correlatives
foron foran 6 6
The forms foron (and its orthographical variant foran) are as predominant
in this text as they are in the early West Saxon adaptation of the Consolatio
(59 occurrences), amounting to ca. 83 per cent of all expressions of the causal
relation. The rest (12 instances) are expressed by nu, either in the simple form
or in correlative expressions. In contrast to the translation of the Consolatio,
however, Byrhtferth employs nu solely for the relation cause result, most
often in pre-posed topic-forming causal clauses:
(117) Nu ic ealles ymbe as ing sprce hbbe, me ing behefe ing t ic
swa mycel ymbe issum getle preostum gecye, swa me ne gesceami-
ge (ByrM, 3.3.275).
Now that I have spoken about these things in detail, it seems to me ne-
cessary to make known to priests enough about this numbering system
that I will not be ashamed .
(118) Nu we habba sceortlice amearcod ra hiwa gefeg e boceras gyma,
nu ing hyt us gedafenlic t we heom gecyon t we r geheton
(ByrM, 3.3.234).
Now that we have briefy written the series of fgures that writers use,
[now] it seems ftting to us to tell them what we promised before .
Discourse deixis 147
In most of these cases, the temporal textual deixis and givenness of the
pre-posed causal clause is, like in (115) above, further emphasized by a verb of
communication, namely sprecan in (117) and mearcian in (118).
Form and its variant forms can certainly be called the central causal con-
nector in Old English, since they mark eighty to ninety per cent of all instances
of causal connection. In spite of the wide variety of its possible functions, how-
ever, there are only few ambiguous instances. In most of the cases, the seman-
tics and pragmatics are clear. In fact, the instances which cause problems for
the understanding of a text can be listed individually (see the discussion in
Mitchell 1985: 30113014). In the following chapters, it will be argued that
this low number of ambiguous attestations and also the essential patterns in
the later development of causal connectors in English is attributable to the
morphological make-up of the Old English connectors and in the deictic value
they inherently contain (for similar approaches, see Wiegand 1982; Traugott
1992; Markus 2000). They belong to category II of Raibles continuum of con-
nectors (introduced above in Chapters 5.4 and 7.4) because they are explicitly
anaphoric or cataphoric.
9.4. Discourse deixis
9.4.1. Form: morphological make-up and discourse deixis
The morphological make-up of the forms form, foron and fory is unprob-
lematic. They are prepositional phrases consisting of the preposition for gov-
erning the distal demonstrative pronoun in the dative (m) or instrumental
(on and y).
7
In all functions, they may, but need not, be followed by the par-
ticle e (see Campbell 1959: 708709; Mitchell 1985: 30113051):
preposition + distal demonstrative pronoun [+ e]
for + dative m instrumental on/y [+ e]
for + m/y [+ e]
Traditional accounts of the history of the for-causals state that the original
prepositional phrase was re-analysed as a conjunction in (pre-)Old English. A
prototypical use of one of the common forms of for m in an external reason
7 Campbell (1959: 290, 708, n.4), however, remarks in his discussion of the forms
on and y: The classifcation of these forms as instrumental is traditional, but
refects neither their origin, nor their prevailing use.
148 Shifting deictics in the history of English causal connection
clause is example (119), where the form foron traditionally classifed as a
subordinating conjunction points anaphorically to the preceding clause.
(119a) Do rto ffe foron unresdg hf ff regulares.
Add thereto fve, because [Conj.] Thursday has fve regulares (ByrM,
1.2.236).
This construction is commonly seen as a re-analysis of
(119b) *Do rto ffe for on: unresdg hf ff regulares.
Add thereto fve for that [reason] [pp as a]: Thursday has fve regulares.
In construction (119b), assumed to be underlying (119a), the prepositional
phrase for on functions as an adverbial in the frst clause, and cataphorically
refers to the following clause. The demonstrative on has a dual function: with
respect to the frst clause, it is the noun phrase in the prepositional phrase,
which itself functions as an adverbial. At the same time, on indexes, i. e.
points cataphorically, to the second, causal clause, identifying the adverbial
relation cause which must necessarily follow here.
Since demonstratives are by defnition deictic (cf. Brugmann 1904), the noun
phrase m of the construction necessarily needs a point of reference which
it points to. Theoretically, the point of reference required by for m could be
found in the extra-linguistic reality and the speaker could identify the cause in-
dexed by m (this one) by the pointing of his fnger. Commonly, however,
the point of reference is present in the co-text, i. e. the preceding or following
discourse. This is most clearly seen in the so-called correlative constructions,
which are rather frequent in the Old English texts examined (ca. 12 per cent in
COBOETH, ca. 25 per cent in ByrM). They are the most explicit surface mark-
ers of causal connection because they index one another and thus reinforce the
conjoining force of the respective connectors (see the defning properties of
Raibles category II).
(120) Foron Romani hine gelogodon on issum mone (t ys on Februario)
foram he ys scyrtest ealra mona (ByrM, 2.1.36).
The Romans placed it [therefore] in this month (in February) because
it is the shortest of all months.
These constructions are therefore employed for stressing result cause se-
quences as in (120), but can also be used for stressing the topic-forming quality
of pre-posed causal clauses, as in
Discourse deixis 149
(121) Aprelis, Iunius, September, and Nouember habba feower and twentig
ls, and Februarius, foron he ys se lsta and he hf twegra daga
ls onne a ore, foron he forlt eahta and feowertig tida (ByrM,
2.1.372).
April, June, September, and November have twenty-four fewer, and
February, since it is the shortest and has two days fewer than the others,
[therefore] loses forty-eight hours.
The forms form, for y etc. are intrinsically deictic, because their demon-
stratives m or y require an element in the near co-text to which they relate:
this can either be the clause or also a much larger piece of discourse. They are
thus inherently phoric, either cataphoric or anaphoric, and point to a pragmati-
cally governed use of deixis. Wiegand (1982) even goes as far as to maintain
that the phrases for m with variants are not yet conjunctions in Old English
because the prepositional phrase is still so transparent in its deictic reference
in its respective context that we do not have to assume a univerbated use of
form with variants as a conjunction (Wiegand 1982: 388).
For a better understanding of the following argument and, in particular, the
diachronic and cross-linguistic comparisons, it has to be stressed that the pres-
ent investigation uses in accordance with most studies on similar topics the
term deixis in its widest sense. With respect to discourse deixis, a clear dis-
tinction between deixis (e. g. reference) and anaphora (discourse deixis), as
this is supported by, for instance, Knig (2009: 151), cannot easily be drawn
because the written, but also the spoken, preceding and following co-text itself
can be seen as an extra-linguistic point of reference (for a discussion of this
issue and defnitions of deixis and anaphora, see Lenz 1997: 7108 and Con-
sten 2004: 458).
9.4.2. Pronominal connectors in Present Day German
This pattern of so-called pronominal adverbs (Pasch et al. 2003: 7) is fre-
quent in Latin and, though often opaque, in many of the Romance languages
(see above, Chapter 5.4.2) and in all Germanic languages. In Old English we
fnd it for all kinds of semantic relations: see, for instance, r m (e) before,
fter m (e) after, mid m (e) during or wi m e provided that.
In Present Day German, the pattern is not only extant but actually the pre-
dominant one for causal adverbial connectors (cf. dem- in demnach or des- in
deswegen):
150 Shifting deictics in the history of English causal connection
Table 98: Present Day German connectors cause/result
German Conjunctions cause DA, WEIL, weshalb, weswegen
German Conjunctions result so dass
German Adverbial Connectors result also, DAHER, DARUM, demNACH,
demzufolge, deshalb, deswegen,
folglich, infolgedessen, so, soMIT
German Adverbial Connectors cause DENN, nmlich
semibold: pronominal elements; underlined: lexical elements; capitals: other types of
deixis (for instance, space or time deixis)
Virtually all of the causal connectors in Present Day German, but predominantly
adverbial connectors marking result apart from folglich, which is purely lexi-
cal contain a deictic element (see items in bold), most often a demonstrative pro-
noun in the dative (dative singular masculine/neuter dem) or, in younger forms,
the genitive (genitive singular masculine/neuter des), which allows its analysis as
a prepositional phrase with anaphoric or cataphoric reference.
Another group of important deictic connectors marked by capital letters are
those relating to time and space deixis, such as PDG weil (cf. PDG Weile while,
(short) period of time; time) or PDG denn (variant of PDG dann then, see Kluge/
Seebold 2002, s. v. denn; time). Originally spatial da (there space; by metonymic
transfer also then time) is attested in PDG da, PDG daher and PDG darum.
These are text deictics (see above, Chapter 5.2.2) because the point of reference is
the text in its temporal and spatial extension. Da there in PDG daher, for instance,
relates the following to the preceding element of discourse which is the cause for
the result mentioned in the daher-clause (cf. PDG -her from (there)). In contrast
to the pronominal connectors, which explicitly require a point of reference in the
co-text, this signalling of deictic reference is more subtle and thus asks for a more
sophisticated cognitive process by the listener/reader (Consten 2004: 2637).
Table 99: Deictic expression in Present Day German causal connectors
Pronominal deixis demNACH, demzufolge, deshalb, deswegen,
infolgedessen, weshalb, weswegen
Time/Space deixis DA, DAHER, DARUM, DENN,
?
WEIL
Deictic so
8
so (dass), somit,
?
also
Lexical means folglich
No longer transparent
?
also,
?
WEIL
?
of uncertain status as to their transparency
8 So is a very strong deictic indicator but is in German as well as in English notori-
ously diffcult to analyse because of its polyfunctionality; see OED, s. v. so and for
a full account of the functions of so in Old English Schleburg 2002 (see also above,
Chapter 7.5.1).
Discourse deixis 151
Present Day German hence shows three different patterns for causal connec-
tors: in addition to the pronominal connectors, such as deshalb or deswegen,
it uses linguistic items which employ time and space deixis (see daher, somit);
only rarely lexical elements, such as PDG folglich lit. followingly, are found.
9.4.3. Deictic elements in English causal connectors
An examination of the system of causal connectors in Present Day English
shows that there is not a single fully transparent remnant of the principal pat-
tern of Old English, i. e. pronominal connectors such as form.
Table 910: Deictic expression in Present Day English causal connectors
9
Pronominal deixis
Time/Space deixis SINCE, NOW (that), HENCE, THEN,
?
THEREfore
Deictic so, thus so, so (that), thus
Lexical means because, accordingly, consequently, for
Not transparent as,
?
THEREfore
semibold: pronominal elements; underlined: lexical elements; capitals: other types of
deixis (for instance, space or time deixis)
With respect to the conjunctions, we fnd the lexical because (cf. cause) and the
polyfunctional since, which is transparent in its time deixis because it still has
its temporal meaning. In as (< eall swa all so), on the other hand, the deictic
so is no longer transparent; the same is true in the case of for, which is no longer
related to the prepositional phrase for m, but is only semantically transpar-
ent through the preposition for because of that.
This is in stark contrast to Old English, where we only fnd like in Pres-
ent Day German expressions marked explicitly for deixis, either pronomi-
nal connectors such as form with variants or originally instrumental on
or y in the second elements of correlatives. Time deixis (which may origi-
nally also denote space)
10
is attested in nu now, a then and onne then
(and variant forms), either as simple forms or in correlatives. Like form,
9 For further polyfunctional connectors, which may also be used for the causal rela-
tion, such as after all, indeed, in fact or of course, see below, Chapter 12.
10 Clark Hall (1984), s. v. heonu lists the meaning therefore among a wide variety
of concessive meanings for this word. In all of its seventy-eight attestations in the
Rushworth and Lindisfarne Gospels, however, it exclusively glosses Latin ecce and
is thus not causal (see DOEC).
152 Shifting deictics in the history of English causal connection
these are ambiguous adverbs/conjunctions (marked by in this and the fol-
lowing tables).
Table 911: Deictic expression in Old English causal connectors
Pronominal deixis form, foron, fory
in correlatives: y, e
Time/Space deixis nu, a, onne
swa?
This summary shows that Old English only employs explicitly deictic elements
to mark causal relations, mainly when the high token number of the variants
of the form is taken into account pronominal connectors.
9.5. Causal connectors in the history of English
9.5.1. Causal connectors in English translations of Boethius
De Consolatione Philosophiae
Most of the causal connectors used in Present Day English testify to dramatic
changes which have occurred in the English system after the Old English pe-
riod. For a frst survey of the development see Table 9.12 summarizing various
causal connectors as attested in translations of Boethius De Consolatione Phi
losophiae after the Old English period (for Old English, see Table 9.6). It lists
the different connectors for the relations cause result and result cause,
ordered by diminishing frequency in the respective texts.
Table 912: Causal connectors in English translations of Boethius De Consola
tione Philosophiae (absolute numbers in brackets; for details on the
corpus texts, see Appendix C.2.2)
cause result
Chaucer, ME Boece (CMBOETH)
for whiche (8), forwhy (3), therefore (3), forthy (2)
pre-posed: for as moche as (7), for (2), syn (1), syn that (1)
Colville 1556 (CEBOETH1)
therefore (21), whereby (7), so that (6), wherefore (5), so (3)
pre-posed: for by cause (3), synce that (1)
Causal connectors in the history of English 153
Queen Elizabeth 1593 (CMBOETH2)
therefore (17), wherefore (7), so (7)
pre-posed: because (2), since (2)
correlative: because now (1), now that now (1)
Preston 1695 (CMBOETH3)
therefore (17), hence (7), wherefore (4), so (1)
pre-posed: since (3), because (2), now that (2)
James 1897 (CLBOETH3)
wherefore (6), therefore (5), accordingly (3), so (2), so then (2)
pre-posed: since (12)
Watts 1969
therefore (14), so (8), indeed (3), consequently (1)
pre-posed: since (11), because (2)
result cause
Chaucer, ME Boece (CMBOETH)
for (63)
post-posed: for (15), syn (2), syn that (1)
correlative: wherefore for as moche as (1)
Colville 1556 (CEBOETH1)
for (53)
post-posed: for 13, for by cause (2)
Queen Elizabeth 1593 (CEBOETH2)
for (60)
post-posed: for 8
Preston 1695 (CEBOETH3)
for (53)
post-posed: since (7), because (6)
James 1897 (CLBOETH3)
for (28)
post-posed: since (6), because (2)
Watts 1969
for (22)
post-posed: because (7), since (4), for (3)
This table attests to the wide variety of forms that have been coined as replace-
ments of OE form (and its variants), which, of course, cannot be discussed in
detail here. Essentially, it shows that the dramatic structural changes can best
be illustrated by the changes affecting adverbial connectors in the (Early) Mid-
154 Shifting deictics in the history of English causal connection
dle English period (see 9.5.2.) and by the changes of connectors in pre-posed
causal clauses (see 9.5.6. below).
11
9.5.2. Causal adverbial connectors in the history of English
Adverbial connectors are the kind of connectors which have to signal clausal
linkage most clearly because they present speakers opinions of the relation of
the states of affairs explicitly in two information blocks with independent illo-
cutionary weight (see above, Chapter 2.4). Table 9.13 provides a list of linguistic
elements which have served the function of a causal adverbial connector in any
period of the history of English, based on the data in Appendices A.1 and B.3.
To allow a comparison, the layout of the table is modelled on the tables in Kort-
mann (Kortmann 1997: 342 for causal subordinators). It only lists the central
causal/resultive connectors, i. e. no borderline cases such as PDE after all.
12
This table illustrates that the Early Middle English period in particular was a
period of experiment and transition: in the periods ME1 to ME3 (11501350),
nine adverbial connectors are coined, but most of them are ephemeral. Only two
of them have survived into Present Day English, namely therefore
13
and thus (a
late-comer frst attested in the adverbial connector function in 1380; see OED,
s. v. thus).
14
11 There are some differences in the relative frequencies, but not in the forms of causal
connectors between translations and comparatively independent texts. In Chaucers
Astrolabe (CMASTRO complete text), for example, we fnd for the sequence
cause result: therefore (10), wherefore (3) thanne (2), forwhy (1), pre-posed:
for as much as (2), by cause that (2), for (1), sin that (1). For the sequence result
cause, we fnd: for (20), for (8) and prepositional by-cause that (7).
12 Since connectors for the semantic relation cause form a closed class, there are only
very few other causal connectors which are not listed here because they are attested
very infrequently (such as Latin ergo; see OED, s. v. ergo). Similarly, connectors
which may also be used for the relation addition/transition, such as indeed, in
fact, of course are not listed here (for these, see also Table 9.15). The meanings and
functions of after all are still only given in the entry for the preposition / adverb
after in the OED (s. v. after).
13 For the proliferation of there-compounds from the beginning of the Middle Eng-
lish period, see sterman 1997, and above, Chapter 8.3. The history of therefore is
rather dark (cf. OED, s. v. therefore): I suggest a similar path to that which has led
to Latin propter-ea (cf. OE for-m) vs. ea-propter (cf. ME therefore).
14 The adverb us is, of course, attested in Old English, but it was only employed as a
circumstance adverb modifying a verb, in phrases such as us sprecende speaking
thus. The different uses are, however, often hard to distinguish.
Causal connectors in the history of English 155
T
a
b
l
e

9

1
3

:

C
a
u
s
a
l
/
r
e
s
u
l
t
i
v
e

a
d
v
e
r
b
i
a
l

c
o
n
n
e
c
t
o
r
s

i
n

t
h
e

h
i
s
t
o
r
y

o
f

E
n
g
l
i
s
h

(
b
a
s
e
d

o
n

A
p
p
e
n
d
i
c
e
s

A
.
1

a
n
d

B
.
3
)
Old
English
11001350
(ME 1/2)
13501420
(ME 3)
14201500
(ME 4)
15001570
(EModE1)
15701640
(EModE2)
16401710
(EModE3)
17101780
(LModE1)
17801850
(LModE2)
18501920
(LModE3)
1920
(PDE)
[
c
u

l
i
c
e
]
s
w
a
s
w
a
s
o
s
o
s
o
s
o
s
o
s
o
s
o
s
o
s
o

o
n
n
e
t
h
a
n
e
/
t
h
e
n
n
e
t
h
e
n
t
h
e
n
t
h
e
n
t
h
e
n
t
h
e
n
t
h
e
n
t
h
e
n
t
h
e
n
t
h
e
n
[

s
]

i
f
o
r

m
/
f
o
r

o
n
f
o
r

o
n
f
o
r

y
f
o
r

i
f
o
r

i
f
o
r

i
f
o
r
h
w
i
f
o
r
h
w
i
f
o
r
h
w
i
f
o
r
f
o
r
f
o
r
f
o
r
f
o
r
f
o
r
f
o
r
f
o
r
f
o
r
f
o
r
f
o
r

t
h
a
t
f
o
r

t
h
a
t
f
o
r

t
h
a
t
t
h
e
r
e
f
o
r
e
t
h
e
r
e
f
o
r
e
t
h
e
r
e
f
o
r
e
t
h
e
r
e
f
o
r
e
t
h
e
r
e
f
o
r
e
t
h
e
r
e
f
o
r
e
t
h
e
r
e
f
o
r
e
t
h
e
r
e
f
o
r
e
t
h
e
r
e
f
o
r
e
t
h
e
r
e
f
o
r
e
w
h
e
r
e
f
o
r
e
w
h
e
r
e
f
o
r
e
w
h
e
r
e
f
o
r
e
w
h
e
r
e
f
o
r
e
w
h
e
r
e
f
o
r
e
w
h
e
r
e
f
o
r
e
w
h
e
r
e
f
o
r
e
w
h
e
r
e
f
o
r
e
h
e
r
e
f
o
r
e
h
e
r
e
f
o
r
e
[
h
e
r
e
f
o
r
e
]
[
h
e
r
e
f
o
r
e
]
h
e
r
e
f
o
r
e
f
o
r

w
h
i
c
h
f
o
r

w
h
i
c
h
f
o
r

w
h
i
c
h
f
o
r

t
h
e

w
h
i
c
h
[
e
r
g
o
]
t
h
u
s
t
h
u
s
t
h
u
s
t
h
u
s
t
h
u
s
t
h
u
s
t
h
u
s
t
h
u
s
t
h
u
s
[
c
o
n
s
e
-
q
u
e
n
t
l
y
]
c
o
n
s
e
-
q
u
e
n
t
l
y
c
o
n
s
e
-
q
u
e
n
t
l
y
c
o
n
s
e
-
q
u
e
n
t
l
y
c
o
n
s
e
-
q
u
e
n
t
l
y
c
o
n
s
e
-
q
u
e
n
t
l
y
h
e
n
c
e
h
e
n
c
e
h
e
n
c
e
h
e
n
c
e
h
e
n
c
e
t
h
e
n
c
e
w
h
e
n
c
e
w
h
e
n
c
e
a
c
c
o
r
d
i
n
g
l
y
a
c
c
o
r
d
i
n
g
l
y
a
c
c
o
r
d
i
n
g
l
y
a
c
c
o
r
d
i
n
g
l
y
156 Shifting deictics in the history of English causal connection
A closer look at the various forms coined in these centuries shows that the
coinages refect the consequences of the loss of transparency in the central
causal connectors at the end of the Old English period. Foran (from instru-
mental for on or from levelled forms of the dative for m) and the instru-
mental fori (< for y) lost their transparency when the paradigms of both of the
Old English demonstratives collapsed (see above, Chapter 7.5). We fnd rem-
nants of infected forms of the Old English demonstratives in some southern
and western texts as late as the thirteenth century, but most of these infections
are number concord and thus irrelevant for form and fory. Consequently,
these forms lose their deictic value and are soon lost altogether. In this process,
phonologically weakened forms such as foran [-(n)] are given up much ear-
lier than the stronger fori (cf. the long vowel in OE for), which survives as
a univerbated form until the end of the Middle English period (see MED, s.vv.
for-than and for-). For fory, orthographical forms showing a reduction of the
fnal long vowel (such as fore) are only attested from the late thirteenth and
fourteenth century. The earlier forms seem to refect the full (long) vowel and
not a reduced schwa (cf. vor-ui or feri; see OED, MED, s. v. forthi, for-).
The forms of OE form, on the other hand, are in spite of the Old English long
vowel regularly attested as -am, -on, or en (see MED, s. v. for-than).
Studies examining the history of OE form in detail fnd a very rapid sim-
plifcation of the for-formulae, which is frst attested as simple for around 1100
in manuscript F of the Anglo-Saxon Chronicle (see the tables in Kivimaa 1966:
214 and 250). This rapid rate is particularly evident in the continuations in the
Peterborough Chronicle, where the use of simple for increased from 27 per
cent in Continuation I to 86 per cent in Continuation II (for the years 1132 to
1154). In the Early Middle English texts investigated by Kivimaa (1966), for is
the form used almost universally, and in several texts it is the only one attested.
This line of development is very similar to that which has above been summa-
rized as the change attested in French-based creoles, where parce que appears
as non-transparent pas (Raible 1992: 201; see above, Chapters 5.5.2 and 5.5.3).
In pas, parce que has lost both of its deictic elements ce and que, a process
identical to the development of form e into for in Early Middle English.
This parallel development, of course, brings the Middle English creole hy-
pothesis back to mind, i. e. the conjecture that Early Middle English underwent
creolization at the time of either the Norse or Norman Conquests, or during
both periods. This hypothesis, which was very popular in the 1980s, was main-
ly based on the assumption of an extreme reduction in infected forms from Old
English to Middle English. While a distinct process of creolization (with all its
properties) is certainly to be rejected for Middle English (for a summary of the
arguments, see Grlach 1986), it is nonetheless interesting to see that the reduc-
Causal connectors in the history of English 157
tion of forms has also led in both languages discussed here to a collapse of
the system of causal adverbial connectors.
At the same time, we see in English and in the creoles attempts at innova-
tions in the feld. Like its Old English precursor form, the reduced form for
was ambiguous as to its status as an adverb, a coordinating or subordinating
conjunction (see also Jucker 1991). While this does not seem to have present-
ed any problems in Old English, we increasingly fnd tendencies for a better
mapping of form and function by specifcally marking the most problematic
pre-posed subordinate clauses (by for as moche as; see below, section 9.5.6)
and adverbial connectors.
9.5.3. Deixis in new adverbial connectors
With respect to the novel adverbial connectors in Early Middle English, it is
obvious at frst glance that there are not very many different patterns among
the new coinages.
Table 914: Adverbial connectors: new coinages in Middle English (ME1 to ME3)
Pronominal Connectors
with demonstrative for-that
with interrogative/relative forhwi, for which, wherefore
Text-Deictics
spatial distal: therefore
proximal: herefore
Deictic
thus
Non-Deictics
no longer transparent for
lexical ergo
These innovations show that Middle English frst of all tried to use the new ma-
terial for coining causal connectors similar to the make-up of OE form. The
earliest coinages are pronominal connectors which employ the new form of the
demonstrative that (MED, s. v. for-that adv. & conj.; the lexicalized status of for
which is not as clear). The new coinages are also similar to their Old English
models in that many of them are so-called ambiguous adverbs/conjunctions,
which can be employed as adverbial connectors or conjunctions (cf. for-that,
for which, forhwi forwhy and wherefore). This pattern can be exemplifed by
the rather frequent for whiche (found eight times in Chaucers Boece; see Table
9.12) and the functional extension of forms such as the relative and originally
158 Shifting deictics in the history of English causal connection
interrogative forhwi. From the thirteenth century onwards, it is not only used as
an interrogative or relative but also as an adverbial connector signalling a new
information unit, thus carrying the meaning therefore (see OED, s. v. forwhy)
This use is only attested as a conversational implicature in Old English (see DOE,
s. v. forhwm, forhwon, forhw B.).
15
Forhwi and forwhich are, however, the last
coinages which follow the Old English pattern of pronominal connectors.
The other feld of experiment are new connectors employing time or space
deixis, similar to OE nu From early Middle English onwards, deixis of space in
particular is becoming more important (see above, Chapters 8.2 and 8.3). The
co-text is taken as a point of reference in patterns using the relative where and
the distal and proximal forms there and here (cf. therefore, wherefore), which
relate the following to the preceding discourse. Many studies have shown that
spatial cognition is at the heart of our thinking and that most of our fundamen-
tal concepts are organized in terms of one or more spatialization metaphors
(Lakoff and Johnson 1980: 17; Levinson 2003). With respect to the use of time
and space deictics as connectors, however, there is a difference between orality
and literacy. While time deixis may be used for both the spoken and the written
medium, the metaphoric extension of spatial terms into metalinguistic ones is a
function of literacy, since it is only with literacy that language is objectifed in
visual space (cf. Ong 1982; for details, see above, Chapter 8.3).
Though this means of establishing cohesion is not as explicit as the anaphoric
or cataphoric linkage by a demonstrative, it is still comparatively transparent in
signalling deixis by reference to the written text in its chronological and spatial
extension. These forms become very frequent as adverbial connectors from the
beginning of the thirteenth century onwards (cf. sterman 1997 and Markus
2000), i. e. at exactly the time when the double paradigm of the demonstratives
was given up, yielding the indeclinable article the and the demonstrative this/
that, which are only infected for the referential category number.
9.5.4. ME for that
Considering these variants of pronominal forms and the fact that none of them
has survived into Present Day English, one of the central questions is why
speakers and writers of Early Middle English did not choose for that as the
central connector or conjunction. That is, Why did speakers not simply start
15 For the importance of the difference between conversational implicature and coded
meaning for a theory of regularities in semantic change, see Traugott and Dasher
2002.
Causal connectors in the history of English 159
substituting at for m, on and y, as they generally did for the demonstra-
tive objects of prepositions?
16
The fact that on and y were not replaced by
the new demonstrative form that, on the other hand, is not startling, since the
instrumental y was only marginally a case category even in Old English (see
Campbell 1959: 708, n. 4). Since y in particular was mainly used as an ele-
ment of complex conjunctions, it does not come as a surprise that fori survived
as a univerbated, lexicalized resultive connector therefore until the end of the
Middle English period.
Both the weak impact and the loss of for that, however, require a more de-
tailed investigation because they testify to the collapse of the pattern which
had been central in Old English. For this examination, Early Middle English
texts are chosen which employ for at as both an adverbial connector and a
conjunction (CMTRINIT ca. 1225, CMKENTSE ca. 1275, and CMAYENBI
ca. 1340). In these texts, for at is used in clear replacement and imitation of
OE form in correlative constructions with fori:
(122) and eft he us wile feie; anne we shulen arisen of deae. and fori [122a;
adverb] we clepe him fader for at [122b; conjunction] he us feide here.
and he will also join [our limbs together] when we will arise from
death. And therefore we call him father because he put us together
here (CMTRINIT, p. 26).
While the instances of for-at in examples (123b) and (123c) should be regard-
ed as conjunctions heading genuine internal reason clauses (they are the focus
of the negation; for this criterion, see above, Chapter 3.2.3), for-at in (123a)
unambiguously functions as an adverbial connector:
(123) And herodes iherde et o king was ibore et solde bi king of geus. for
et [123a] he was ofdred for to liese his king riche of ierusalem. And
al swo herodes iherde is. swo spac te o rie kinges. and hem seide.
hic wille go and anuri hit. et ne seide he nocht herodes for et [123b]
he hit wolde onuri; ac for et [123c] he hit wolde slon (CMKENTSE,
p.214215).
And Herod heard that a king had been born which should be the king of
the Jews. Therefore [123a] he was very afraid that he might lose his
kingdom of Jerusalem And as soon as Herod heard this, he spoke to
the three kings and said to them I will go and worship it [= the child].
16 I would like to thank Cynthia Allen for raising this very important issue at an ear-
lier presentation of this chapter.
160 Shifting deictics in the history of English causal connection
This he, Herod, did not say because [123b] he wanted to worship him,
but because [123c] he wanted to kill him.
In most of its attestations in Early Middle English, for at does not function as
an adverbial connector but as a subordinator because (often rendering Latin
quia). This employment is not at all unexpected, since it is in line with the
development of that as the general subordinator in Middle English. Kivimaa
(1966: 248) shows that the earlier conjunctive phrase employing the subordi-
nating particle OE e lost e during the twelfth century and began to be found
with that only in the thirteenth century. That also takes the place of OE e as
a relativizer and becomes the general subordinator in Middle English (see the
quotation from Fischer 1992 cited above, Chapter 7.5).
The use of that as the general subordinator helps to explain why ME for that
did not replace OE form as the central adverbial connector for the relations
cause/result. While the subordinating particle e was different from all of
the forms of the demonstratives in Old English, the Middle English subordina-
tor that is homonymous with the demonstrative. This means that sentence-ini-
tial for that would be conceptualized as introducing a pre-posed causal clause
and not as an adverbial connector. For this reason, it is more and more avoided
in this position and replaced by other, unambiguous connectors.
For examples of this unstable situation of experiment and transition at the
beginning of the thirteenth century, see the following examples from Vices and
Virtues (CMVICES1; ca. 1225):
(124) Foran [124a] hie bie godes wierwinen, alle o e willen hem sel-
uen heiin. Godd sei him self at hie sculen bien ineerede. Hierfore
[124b] ic am neer and unmihti, foran [124c] ic habbe eben prud and
modi (CMVICES1, p.5).
Therefore [124a] they are Gods enemies, even though they want to raise
themselves. God says himself that they shall be lowered. Therefore [124b]
am I low and powerless, because [124c] I was proud and conceited .
(125) Hie is anginn of alle cristendome, hie mai michele eadinesse of-earnin
at ure lauerde gode, for an e [125a] hie ilief at hie nure niseih. For
i [125b] sade Crist: Eadi bie a menn e on me belieuen and nure
me ne seien! (CMVICES1, p.25).
He is the beginning of all Christianity, he may earn much blessing from
our Lord, because [125a] he believes what he has never seen. Therefore
[125b] Christ said: Blessed are those, who believe in me and have never
seen me.
Causal connectors in the history of English 161
The form foran with phonological levelling is still the most polyfunctional
item: it is used as an adverbial connector marking result in (124a), as the second
part of a correlative construction marking internal cause in (124c) and still
with the particle e in an internal reason clause in (125a). In adverbial con-
nector function, we also fnd the spatial hierfore (124b) and the by then no
longer transparent form fori (125b).
Another text from the Early Middle English period, Hali Meidenhad (Kathe-
rine Group; CMHALI, ca. 1225) only differentiates the form for denoting
cause (126a, 126c, 126d) as against forr-i (126b) denoting result.
(126) Meihad is et an eoue iettet te of heouene: do u hit eanes awei, ne
schalt tu neauer nan oer swuch acourin; for [126a] meihad is heouene
cwen, & worldes alesendnesse, urh hwam we beo iborhen forri
[126b] u Ahest, meiden, se deorliche witen hit; for [126c] hit is se heh
ing, & se swie leof godd, & se licwure. & et an lure et is wituten
couerunge. ef hit is godd [leof], et is, him seolf swa ilich, hit na wun-
der; for [126d] he is leofukest ing, & buten eauer-euch bruche
(CMHALI, p.135).
Virginity is the one gift granted to you from heaven; if you once dis-
pose of it, you will never regain another quite like it. For [126a] virginity
is the queen of heaven, and the worlds redemption through which we
are saved Therefore [126b], maiden, you have to guard it carefully.
For [126c] it is the high thing and so very dear to God and so acceptable,
and one loss of it is without recovering. If what is so like God is dear
to him, it is no wonder, because/for [126d] he is the loveliest thing and
without any sin .
9.5.5. Recursive for Latin nam/enim
Already this text from the beginning of the thirteenth century shows a prolif-
eration of for, which has become non-deictic after losing its demonstrative. It
is almost infnitely repeatable and may be used for post-posed causal clauses
working on the local level, providing an explanation for the frst connect. In
most cases, however, non-deictic for is ambiguous (126d) or employed on a
more global level, sketching or justifying the line of argument of the author.
For is thus a prototypical coordinator, because it marks the second connect
as having an illocutionary weight of its own and, more importantly, explicitly
marking the voice of the speaker who comments on his view of the relation of
textual portions.
162 Shifting deictics in the history of English causal connection
It is thus comparable to the Latin adverbial connectors nam and enim (see
Kroon 1995: 131203), both of which mainly work on the global level of text-
ual organization (in the present study termed transitional). Instead of being
adverbial connectors in a prototypical, strict sense, they are rather connective
particles concerned primarily with the presentation and organization of the
information conveyed in the discourse (Kroon 1995: 203). They are thus situ-
ated at the interface between the clausal relations of cause/result and add-
ition/transition. Accordingly, they are commonly not rendered by form
and its variants in Old English, but by OE solice or witodlice lit. truly, cer-
tainly (see Lenker 2000 and above, Chapter 8.4.3 and below, Chapter 11).
From Middle English onwards, however, these adverbs are less frequently used
and for is preferred to mark the transitional relation. This can be shown by a
comparison of translations of Latin enim and nam in the various translations of
the Consolatio in the history of English.
Table 915: Rendering of Latin enim and nam in English translations of Boethius
De Consolatione Philosophiae
17
Translation text for No translation Adverbs cognitive
domain truth
Diverse
Chaucer ME
(CMBOETH)
50 4 certes 4 but 1
forhwy 2
and thus 1
Colville 1556
(CEBOETH1)
53 certes 3
surely 1
truly 1
and so 1
for why 1
Queen Elizabeth 1593
(CEBOETH2)
49 11
Preston 1695
(CEBOETH3)
45 9 but 2
now 2
so then 1
thus 1
Watts 1969
(PDEBOETH2)
25 21 indeed 1 and 1
because 1
but 1
17 Old English cannot be used for such a detailed analysis, because the Old English
Consolatio is not a genuine translation, but a rather free adaptation with many di-
vergences from and additions to the Latin text. Also, the numbers for Chaucer do
not fully agree with those of the other translations, because Chaucer only addition-
ally uses the Latin text. Chaucers main source is an Old French version by Jean de
Meun (ed. Dedeck-Hry 1952), which renders Latin nam and enim by OF car.
Causal connectors in the history of English 163
This comparison shows that for has indeed become a transitional connector. In
this discourse function, it is also called recursive for (Mueller 1984: 135). It
has lost much of its causal meaning denoting internal reason clauses and main-
ly works on the textual level, indicating This is my line of argument. In Chau-
cers translation of the Consolatio, this is evident in the many collocations with
other connectors, such as for so as (127a), for yif (127b) and for certes (127c).
(127) Philosophie: that God, , is good. For [127a], so as nothyng mai
ben thought betere than God, it mai nat ben douted thanne that he that
no thinge nys betere, For [127b] yif God nys swyche, he ne mai nat
be prince of alle thinges; for [127c] certes somthing possessyng in itself
parfyt good schulde be more worthy than God For [127d] we han
schewyd apertely that alle thinges that ben parfyt ben frst er thyng-
es that ben inparft; and forthy [127e] we owe to graunte that the
sovereyn God is ryght ful of sovereyn parft good (CMBOETH, p.432).
[It is the universal understanding of the human mind] that God is
good. For [127a], since nothing can be conceived better than God, then
it may not be doubted that nothing is better For [127b], if God is not
such, he cannot be the prince of all things; for [127c], certainly, there
would have to be something else possessing perfect goodness over and
above God For [127d] we have shown overtly that all things that are
perfect are superior to things that are imperfect; and therefore [127e],
, we must grant that the supreme God is very full of supreme and per-
fect goodness.
This transitional use of for highlighting the line of argument is the predominant
one from the middle of the fourteenth century until the middle of the eighteenth
century (see the high number of occurrences in Table 9.15). For Early Modern
English, this interpretation is also supported by the increasingly more consistent
punctuation: more than half (58 per cent) of the for-clauses analyzed by Clar-
idge and Walker are preceded by heavy punctuation (Claridge and Walker 2002:
4244). It has to be stressed again that in these cases for should be classifed as a
coordinator, even if there is no way to confrm this by the syntactic or topological
criteria summarized above (cf. Chapter 2, Table 2.2 and also Table 9.1). Con sider-
ing issues of pragmatics and text-linguistics, however, for has to be analyzed as
a coordinator because it gives an independent illocutionary weight to the second
connect, in many instances by explicitly indicating that the speaker thinks that
the propositions of the two sentences are connected in some way or other.
Much of the uncertainty about the status of for in Present Day English is
probably due to the fact that for, and in particular the rhetorical uses of for,
164 Shifting deictics in the history of English causal connection
have decreased rapidly since the end of the eighteenth century. In Present Day
English translations of the Consolatio (Watts 1969), the number of instances
of for translating Latin nam and enim is halved. In the other half of the cases,
enim and nam are not translated at all. These asyndetic constructions, how-
ever, are not complex and do not present the speakers opinion on the state of
affairs. This is entirely different in the Modern German translation (Neitzke
1959 in Grasmck 1997), which employs the regular connectors PDG nmlich
(twenty-eight instances) and PDG denn (twenty-one instances) in about equal
numbers and only occasionally uses other forms such as nun (1), jedenfalls (1),
in der Tat (1), freilich (2) and wirklich (1). In only three instances, nam or enim
are not translated into German. These differences are not due to individual
translators or errors of the Modern English translation (the fgures are almost
identical for the translations by Green 1962 and Watts 1969), but to the fact that
English has lost a regular adverbial connector for the sequence result cause.
9.5.6. Subordinators: for as much as, since, because
This transition of the all-purpose causal connector OE form to a transitional
connector indicating the line of argument has also yielded changes in the sys-
tem of the subordinators, i. e. the emergence of for as much as, since (that) and
because, frst mainly in pre-posed subordinate clauses (see Table 9.12). Be-
cause of the proliferation of instances of for as a rhetorical marker of the line of
argument, non-deictic for was less frequently used in these usually thematic
pre-posed clauses, since it did no longer mark these contexts unambiguously
enough. The comparison of the translations of the Consolatio in the various
periods of English shows that since, as and because are never used as render-
ings of Latin adverbial connectors such as igitur or quare (i. e. the sequence
cause result) or nam and enim (i. e. the sequence result cause) before
the Modern English period, but only for subordinators such as Latin quoniam
and causal (not temporal) cum.
While therefore and then (rendering Lat. igitur or quare) and for (render-
ing Lat. nam or enim) are rather stable from the middle or end of the Middle
English period until Modern English (in the case of therefore even until today),
we fnd more variation for quoniam and causal cum.
18
The earliest expression
18 Unfortunatly, Latin quia is only attested once in the selected corpus. This, however,
mirrors the relations in the full text rather precisely. The numbers for the full text:
igitur (157), nam and namque (162), enim (141), quoniam (49), quare (27), quia (9).
This also shows the centrality of the adverbial connectors in Latin.
Causal connectors in the history of English 165
coined to avoid ambiguities is ME for as moche as (see MED, s. v. for as moche
as), which is attested throughout Middle and Early Modern English. It is most
remarkable that for as much as is the only one of the subordinators coined in
Middle English which has not survived into Present Day English (see Kort-
mann 1997: 331332). It is most likely that this form (probably modelled on Old
French causal pourtant; see also Spanish por tanto therefore) was given up
because it was neither deictically nor lexically transparent and because it was
too close to the then proliferating sentence-initial rhetorical for
While the temporally deictic since is occasionally found from Middle Eng-
lish onwards, often rendering Latin temporal cum (with indicative) and causal
cum (with subjunctive), because is frst mainly attested in the collocation for
because (in fve instances in Colville 1556), i. e. in a combination of the causal
connector for and an original prepositional phrase (see also OED, s. v. because
and MED, s. v. because), but is subsequently also, though for some time rather
sparingly, used on its own. Only after 1750, because fnally replaces for and
becomes the all-purpose connector of Present Day English (for quantitative
data, see Claridge and Walker 2002). A rather typical instance of the various
renderings of a pre-posed causal clause introduced by Latin quoniam is
(128) Quoniam igitur agnovisti, quae vera illa sit, quae autem beatitudinem
mentiantur, nunc superest, ut unde veram hanc petere possis agnoscas
(Book III, Prose 9; Grasmck 1997: 154).
(128a) Thanne, for as moche as thou hast knowen whiche now by-
hoveth (Chaucer, COBEOTH).
(128b) Now for bycause thou hast knowen , then now (Colville
1556; CEBOETH1).
(128c) Because thou knowest now, what , now it followith (Queen
Elizabeth 1593; CEBOETH2).
(128d) Seeing then thou knowest already which it remains (Pres-
ton 1695; CEBOETH3).
(128e) Now then, since you know what true happiness is, and the things
that falsely seem to offer it, what remains now is that you have to look
for true happiness.
All in all, (for) because is still very rare in the Early Modern English material
investigated (eighteen instances including the collocation for because). While it
occasionally occurs to render pre-posed Latin quoniam in Colville (1556), it is
later increasingly used for marking post-posed external causal connection (see
the use in Preston 1695). In view of the argument presented above, it has become
the central connector in Present Day English because it is lexically transparent.
166 Shifting deictics in the history of English causal connection
9.6. Conclusions
This analysis of systematic changes in the forms of clausal connectors in the
history of English has shown that the system of subordinators and adverbi-
al connectors expressing the relation cause/result was almost completely
re-structured after the Old English period. While Old English virtually only
had one all-purpose connector the pronominal connector form , Modern
English uses different linguistic items for coding the various hierarchies and
relations of cause and result.
Similar to processes also attested for French creoles (French parce que
creole pas; see above, Chapter 5.5.25.5.3), the central Old English connector
form (and the variants foron and fory) lost its deictic value and was
throughout the Middle and Early Modern English period used in its highest
token number not as a causal connector, but as the text-structuring connec-
tor for. This excessive use of this rather unspecifc recursive for from late
Middle to Late Modern English (see Appendix B.3.3), however, fnally led to
its becoming more and more infrequent; for is now, if used at all, considered
to be archaic (Peters 2006, s. v. for). As a result, Present Day English lacks an
adverbial connector expressing the relation result cause.
Since for was in particular in specialized contexts such as pre-posed the-
matic causal clauses not unambiguous enough, new subordinators such as
for as much as, since and because were coined, which, when used in these
positions, i. e. introducing pre-posed causal clauses, helped to disambiguate
alternative structures and thus facilitated information processing.
The collapse of both of the paradigms of the demonstratives at the begin-
ning of Early Middle English, yielding the homonymic form that for the de-
monstrative and the subordinating particle, triggered the coinage of a number
of new adverbial connectors in the frst two sub-periods of Middle English
(ME1 and ME2), most of which were basically modelled on the Old English
pattern of pronominal connectors (cf., e. g., for that, therefore or wherefore).
The most likely candidate for a new adverbial connector, for that, was only
sporadically used because of the homonymy of that with the subordinating
particle that, which was increasingly used as pleonastic that in various kinds
of subordinate contexts (as a complementizer, in relative and all kinds of ad-
verbial clauses). Early Middle English thus emerges as a period of experiment,
when speakers sought to express the relations by other means, mainly by time
or space deictics. Yet, of the connectors formed by this experimental pattern,
only therefore has survived (though in Present Day English even therefore does
not seem to be transparent in its space deixis for most speakers; see OED, s. v.
therefore). This is different with younger connectors used since the end of the
Conclusions 167
Middle English period. Adverbial connectors such as hence and after all, for
instance, are still transparent in their space and time deixis, because they may
still be used as adjuncts with their original spatial (Go from hence; OED, s. v.
hence, I.) or temporal meanings (After all this had happened ).
Another group of connectors emerging from the end of the Middle English
period onwards indicates, however, that English has now almost completely
abandoned its original structural pattern which explicitly (by, for instance, de-
monstratives) or inherently (by time or space deixis) marked deixis in connec-
tors. English now also signals causal connection by purely lexical means: this
is true for adverbial connectors such as consequently or accordingly but in par-
ticular for the high-frequency item because (cf. the noun and verb PDE cause).
With this concentration on lexical means, English has again moved far away
from the Germanic system which is still well alive in Present Day German.
This exemplary scenario of causal/resultive connectors is most similar to the
developments in the feld of concessive/contrastive connectors, which will be
discussed in the following chapter. This does not come as a surprise because
various recent typological studies could show that the so-called cccc-relations
or four-c relations cause, concession, condition
19
and contrast often show
parallel developments (see the papers collected in Couper-Kuhlen and Kort-
mann 2000).
19 Condition is not relevant for the present study because it is not expressed by adver-
bial connectors, but only subordinators or, rather, correlative constructions (e. g. if
then, when then). For details, see above, Chapter 2.4.
10. contrast/concession
10.1. cause condition contrast: affnities
As has been pointed out in the research report on connectors above (Chapter 1),
the semantic relation of concession has received by far the most attention in
research to date. It has been discussed in a considerable number of lexico-se-
mantic studies analysing its general logical properties (on Present Day Eng-
lish, see, for instance, Halliday and Hasan 1976: 250256; Quirk et al. 1985:
15.4015.43). Cross-linguistic studies focus on its differences and similarities
with the other ccc-relations, i. e. the semantic relations condition, cause and
contrast. Because of these mostly still transparent affnities and interde-
pendencies, concession has also played a major role in studies on language
universals and long-term developments in language change, such as gram-
maticalization (see, e. g., Knig and Traugott 1982; Knig 1985a, 1985b, 1988;
Rudolph 1989, 1996; Gonzlez-Cruz 2007; Sorva 2007). Recently, a number
of studies have approached the topic from a discourse-functional perspective
(Couper-Kuhlen and Thompson 2000; Barth-Weingarten 2003; Barth-Wein-
garten and Couper-Kuhlen 2002).
In spite of all this research, however, it has not been possible to defne the
notion concession in a more generally accepted way: the many different ap-
proaches refect the complexity of this relation, which is in spite of its com-
plexity a very basic one. Its complexity is mirrored in the fact that it is acquired
very late in language acquisition (even after the already fairly late conditionals;
see Bowerman 1986). Its basicness is refected by the fact that concessives are
attested in most of the languages studied in broad cross-linguistic analyses of
subordinators (see Kortmann 1997: 343).
For the present study of the diachrony of adverbial connectors, the relation
of concession to contrast proves to be much more essential than its relation
to cause (concessivity can be said to express a causal relation that remains
unfulflled or ineffective) or, in particular, condition. For the development of
concessive subordinators, on the other hand, the relation to conditionals via
concessive conditionals has played a much more important role (see subordi-
nators such as PDE even if, even when consisting of a focus particle and a con-
ditional). These developments are of only marginal interest to the present study
because they are as their origin in conditionals suggests mainly relevant for
the coinage of subordinators and not adverbial connectors. These differences
in origin are related to the differences in information structure of hypotactic
CONCESSION CONTRAST 169
and paratactic structures: Conditional relations, which are only expressed in
subordinate constructions (see above, Chapter 2.4), are very likely to give rise
to subordinators and not at least not immediately to adverbial connectors.
10.2. concession contrast
10.2.1. concession
For a better understanding of the developments and, in particular, the instabili-
ties in the feld of adverbial connectors in the history of English, it is neces-
sary to provide working defnitions for the semantic relations concession and
contrast. As in all structures discussed in the present study, a prototypical
concessive relation consists of two propositions realized in two different (sen-
tential) structures, here called connects (see properties M3 to M5, discussed
above, Chapter 2.2). The concessive connection between the two propositions
is prototypically made explicit by a subordinator such as although:
(129) (Al)though he is poor, he is satisfed with his situation.
In the concessive relation although p, q we have one proposition in each of
the connects
(129a) p: He is poor.
(129b) q: He is satisfed with his situation.
Both p and q are thus entailed. Normally, however, the propositions p and q do
not go together (in spite of p, q). Thus the standard implication is normally
(if p, then not-q), i. e. the implication is:
(129c) If someone is poor, she or he is not satisfed with her or his situ-
ation.
10.2.2. contrast
In relations of contrast or adversativeness q, whereas p, on the other
hand, there is no standard assumption that under normal circumstances p and
q should not go together (as against).
1
1 All examples in these sections are taken from either Halliday and Hasan (1976:
250255) or Quirk et al. (1985: 15.3915.44).
170 CONTRAST/CONCESSION
(130) (130a) Peter is poor, whereas Harry is rich.
(130b) Peter is poor, but Harry is rich.
(130c) Mr. Larson teaches physics, while Mr. Corby teaches chemistry.
(130d) I ignore them, whereas my husband is always worried what they
think of us.
Just as in all semantic relations (for cause, see above Chapters 9.2.49.2.5),
contrast can not only express external, socio-physical aspects as in examples
(130ad). In (131), taken from Alice in Wonderland, but does not relate to the
socio-physical reality, but to the act of communication (i. e. expresses an inter-
nal aspect). Alice recognizes that her suggestion which was made with the
intention of being helpful may in fact not be of any use.
(131) you might catch a bat, and thats very much like a mouse, you know.
But do cats eat bats, I wonder?
contrast here is equivalent with internal concessivity, i. e. counter-expectan-
cy on the level of the communication process cf. external in spite of the
facts vs. internal in spite of the communicative roles we are playing, in spite
of the state of the argument. This internal concessive meaning as against
what the current state of the communication process would lead us to expect
may also be expressed by connectors which only signal these internal relations
emphatically in a kind of avowal (Halliday and Hasan 1976: 253255). This
term already shows that connectors such as in fact, as a matter of fact, actually,
to tell you the truth are at the interface between purely contrastive/concessive
connectors and transitional ones (see below, Chapter 12).
Other sub-categories of contrastive connectors which are also purely inter-
nal (i. e. relate to epistemic or conversational aspects) are those which imply a
sense of correction, of either a correction of meaning (cf. antithetic connec-
tors such as instead, rather or on the contrary) or a correction of the wording
(cf. reformulatory connectors such as at least, rather or I mean). Another
special case of connectors expressing predominantly internal contrastive/con-
cessive relations are generalized forms of the adversative relation no matter
(whether or not), still such as in any/either case, any/either way, any
how, or at any rate. Taken by themselves, these dismissives (for the term,
see Halliday and Hasan 1976: 254255) have nothing cohesive about them, but
they always presuppose that something has gone before, i. e. however that may
have been. Accordingly, this generalized sense is indicated by universal quan-
tifers such as all, any, or -ever. These originally non-cohesive forms recur-
rently yield by the conversational implicatures sketched above contrastive
connectors proper, the most prominent case in English being however.
CONCESSION CONTRAST 171
The various sub-categories of concession/contrast and the Present Day
English items which express these relations are summarized in Table 10.1
(adapted from Halliday and Hasan 1976: 255256):
Table 101: Connectors for external and internal relations of contrast/concession
(A) Concessive Relations proper in spite of (external and internal)
but;
yet; though
however, nevertheless, all the same, <despite this>
(B) Contrastive Relations (as against) (external)
but, and
however, on the other hand, at the same time, <as against that>
(C) Contrastive Relations (as against) (internal)
<in fact>, as a matter of fact, to tell the truth, actually, in point of fact
(D) Corrective Relations (not but)
Correction of meaning antithetic: instead, rather, on the contrary
Correction of wording reformulatory: at least, rather, I mean
(E) Dismissive (Generalized Adversative) Relations (no matter still)
(external and internal)
Dismissal closed: in any/either case/event, any/either way, whichever
Dismissal, open-ended: anyhow, at any rate, in any case
<> ambiguous examples
10.2.3. contrast/concession: information structure
In introductory accounts of the relation contrast/concession, we usually fnd
statements or charts claiming that the relations of contrast and cause may be
expressed by either subordinators or by adverbial connectors. See, for instance,
the account in the Comprehensive Grammar (Quirk et al. 1985: 1088):
Subordinators Conjuncts
concessive whereas, while yet
contrastive whereas, while on the other hand, instead
In their more detailed accounts of the hypotactic constructions, however, we
fnd that Quirk et al. (1985: 15.40) introduce examples such as
(132) No goals were scored, although it was an exciting game.
(133) It was an exciting game, although no goals were scored.
172 CONTRAST/CONCESSION
as illustrative of the concessive relation, but they go on saying that they also
imply contrast between the situations between the two clauses (Quirk et al.
1985: 15.40), in particular in examples where the propositions p and q may be
placed in either of the connects. This signals an equal weight of both proposi-
tions. This mixture of contrast and concession (Quirk et al. 1985: 15.43) is
primarily found in post-posed concessive clauses.
Because of the defning property of external counter-expectancy for conces-
sive relations proper, these are typically expressed by subordination, where there
is only one central information entity and the apparently contrary or dismissed
information is given in the subordinate clause. In accordance with the principles
of text information structure sketched above in Chapter 2.4, the information
given in the subordinate clause (in the case of concession: the dismissed infor-
mation) is not pursued in the further discourse. Even more specifcally, conces-
sion is commonly expressed by pre-posed subordinate clauses introduced by
although (informally though)
2
or while, and may be made even more explicit by
a correlative construction (i. e. by an adverbial connector in the second connect):
(134) (Al)though Sam had told the children a bedtime story, June told them
one too.
(135) While he has many friends, Peter is (nevertheless) often lonely.
The preferred pre-position of subordinate adverbial clauses can be explained in
terms of processing factors (see, for instance, Diessel 1996: 77). The introduc-
tion of the concessive connector at the beginning of the sentence guides the
hearers or readers in their interpretation by signalling that the expectations
raised by the assertion to follow will not be fulflled: an underlying expected
chain with the then following proposition will be broken. If, however, the intro-
duction of the concessive connector is deferred until the middle of a sentence,
a reinterpretation of the preceding assertion will be forced.
Speakers and, in particular, writers, who are not able to signal the non-va-
lidity of the proposition in the subordinate clause by intonation or gestures,
therefore often choose to mark the concessive relation in both the connects,
i. e. by correlative constructions (i. e. although nevertheless, although yet,
although nonetheless, although still), in all cases in the sequence subor-
dinate clause main clause. Until the eighteenth century, this double marking
was by far the most common construction for (unambiguous) concessives.
2 For examples from the LLC-Corpus, see the detailed analysis, below, Chapter 10.7.
Crosslinguistic patterns in the origin of concessive connectors 173
Because of the equivalent illocutionary weight of both of the respective con-
nects in paratactic constructions, adverbial connectors, on the other hand, inev-
itably carry a meaning of contrast (and not a pure meaning of concession).
For the present analysis of adverbial connectors, this means that it has often
been impossible to decide whether a construction is indeed unambiguously
concessive. For this reason, the label chosen is contrast/concession, since
contrast is predominant in most of the cases.
10.2.4. Subordinators adverbial connectors: mood distinctions
While the sub-categorization into contrast and concession is hard to make,
it is in contrast to the diffculties in differentiating causal/resultive adverbial
connectors from subordinators not at all problematic to disambiguate homo-
nymous forms functioning as subordinators or adverbial connectors in the rela-
tion of contrast/concession, at least until the time when verbal forms marked
mood distinctions. As has been shown above (Chapter 5.3.2), these differences
are clearly signalled by the mood of the respective verbs (i. e. subjunctive for the
subordinate clause vs. indicative in the clause with an adverbial connector). In
her analysis of mood distinctions in subordinate clauses, Moessner (2006) fnds
that the subjunctive was fairly persistent in concessive clauses in Early Modern
English, when still about 65 per cent of though-clauses exhibit subjunctive forms
(Moessner 2006: 258). This close connection of subordinate clauses and the sub-
junctive in all languages which distinguish subjunctive forms is not only evi-
dent in originally verbal phrases such as albeit or howbeit, comprising the sub-
junctive be (most probably modelled on French soit; for details see Sorva 2007),
but also in subordinate constructions without a subordinator (although they then
generally require a correlative adverbial connector in the other connect):
(136) Be that as it may, I (still) think .
10.3. Cross-linguistic patterns in the origin of concessive connectors
Cross-linguistic studies on the diachrony of concessive connectors have shown
that most of these, regardless of whether they are realized by free morphemes,
prepositions (PDE despite, notwithstanding), conjunctions (although) or adver-
bial connectors (nevertheless, nonetheless), are more complex in nature than
the connectors for the other relations. This is already attested in the inven-
tory of Old English adverbial connectors marking the relation concession:
174 CONTRAST/CONCESSION
apart from the additive rtoeacan also, only the concessives nayls and
swaeahhwere nevertheless consist of more than two morphemes. Both of
them are, however, more complex than the fairly straightforward compound
adverb rtoeacan (consisting of the space deictic rto thereto and the lexi-
cal eac also) because they require a processing of their morphological struc-
ture: OE nayls and swaeahhwere are thus clear examples of the iconic
principle that formal complexity corresponds to conceptual complexity (for the
compound rtoeacan see below, Chapter 11.2.4).
Most of the European languages have a wide variety of concessive markers
and, apparently, newcomers are constantly added to this class in all of the lan-
guages (Knig 1985a, 1985b). In comparison to connectors for other semantic
relations, most of these concessive connectors seem to have developed relatively
late in the history of the respective languages (for subordinators, see Kortmann
1997: 329, 333, 347; for the new set of French concessives, see Klare 1958).
In contrast to other markers, most concessives because of their compara-
tive youth and also because of their morphological complexity have a fairly
transparent history. Knig (1985b) has shown that, in a wide variety of unre-
lated languages, other meanings and functions of concessives can be easily
identifed and concessivity can be related to other semantic domains. On the
basis of such affnities, he distinguishes fve major groups which share certain
aspects in their origins:
(I) Concessives often contain a component which is also used as a univer-
sal quantifer, such as French tout, German and English all-: cf. French
toutefois, tout que, German bei all .; allerdings
(II) Originally conditional (if ) or temporal (when) are used together with an
additive focus particle or emphatic particle like English even, also, too,
such as German ob-gleich; ob-wohl; ob-schon; wenn-gleich; wenn
auch, Latin et-si, French quand mme, lorsque mme.
This pattern has been, as is shown above, important for English concessive
subordinators such as even though.
(III) The connectors may signal a remarkable co-existence or co-occurrence of
two facts as part of their literal meaning,
(a) asserting that one fact p does not prevent another fact q
(b) asserting that fact p does not pay attention to another fact q
(c) by the simultaneity of p and q
(d) by the unhindered continuation of q given p; cf. German ungeachtet,
unbeschadet, gleichwohl, dennoch, or French tout de mme or cependant.
The diachrony of English contrastive/concessive adverbial connectors 175
In addition to these three major sources, concessives may also be formed from
items which
(IV) directly lexicalize notions of conficts, obstinacy, dissonance or our re-
actions to such situations, such as French en dpit de, au mpris de, Ger-
man trotz, trotzdem or Spanish a pesar de (for sorrow/regret of).
Further, they may go back to original epistemic items which
(V) mark the factual character of something (for these, see also below,
Chapter 12), cf. German concessive gewiss or zwar (< mhd. zware from
ze ware (in) truth; Kluge and Seebold 2002, s. v. zwar).
10.4. The diachrony of English contrastive/concessive adverbial
connectors
10.4.1. General tendencies
Since most of the relevant structural and lexico-semantic features of the devel-
opment of contrastive/concessive connectors in the history of English are either
illustrated above (in Chapter 9 on the causal connectors) or are discussed in de-
tail in the relevant diachronic literature on concessives (see Knig and Traugott
1982; Knig 1985a, 1985b; Gonzlez-Cruz 2007; Sorva 2007), the following
sections will only summarize the main fndings for the history of adverbial
connectors in English (sections 10.410.5). A more detailed investigation will
then provide an in-depth study of the meandering paths in the history of Pres-
ent Day English adverbial though and, in particular, its functions when placed
sentence-fnally in spoken discourse.
10.4.2. Antithetic/reformulatory adverbial connectors
Table 10.2 (based on Appendix A.1 and Appendices B.4.2 and B.4.3) shows the
general tendencies in the history of antithetic and reformulatory connectors,
i. e. those connectors which signal a correction in the meaning and the word-
ings (Corrective Relations, see above, Table 10.1 (D)).
As is evident at frst glance, contrastive/concessive connectors are fairly
atypical adverbial connectors, because they have never been realized as pro-
nominal connectors; neither are other deictic means of linkage employed. Most
176 CONTRAST/CONCESSION
T
a
b
l
e

1
0

:

C
o
r
r
e
c
t
i
v
e

(
a
n
t
i
t
h
e
t
i
c

a
n
d

r
e
f
o
r
m
u
l
a
t
o
r
y
)

a
d
v
e
r
b
i
a
l

c
o
n
n
e
c
t
o
r
s

i
n

t
h
e

h
i
s
t
o
r
y

o
f

E
n
g
l
i
s
h
Old English
11001350
(ME 1/2)
13501420
(ME 3)
14201500
(ME 4)
15001570
(EModE 1)
15701640
(EModE 2)
16401710
(EModE 3)
17101780
(LModE 1)
17801850
(LModE 2)
18501920
(LModE 3)
1920
(PDE)
e
l
l
e
s
o
r

e
l
l
e
s
o
r

e
l
l
e
s
o
r

e
l
l
e
s
o
r

e
l
s
e
o
r

e
l
s
e
o
r

e
l
s
e
o
r

e
l
s
e
o
r

e
l
s
e
o
r

e
l
s
e
o
r

e
l
s
e
o

e
r

s
i
d
e

o
n

t
h
e

o
t
h
e
r

s
i
d
e
o
n

t
h
e

o
t
h
e
r

s
i
d
e
o
n

t
h
e

o
t
h
e
r

s
i
d
e
o
n

t
h
e

o
t
h
e
r

s
i
d
e
o
n

o

e
r

h
a
l
f
o
t
h
e
r
w
a
y
s

[
o
t
h
e
r
w
a
y
s
]
[
o
t
h
e
r
w
a
y
s
]
[
o
t
h
e
r
w
a
y
s
]
[
o
t
h
e
r
w
a
y
s
]
[
o
t
h
e
r
w
a
y
s
]
[
o
t
h
e
r
w
a
y
s
]
[
o
t
h
e
r
w
i
s
e
]
[
o
t
h
e
r
w
i
s
e
]
[
o
t
h
e
r
w
i
s
e
]
o
t
h
e
r
w
i
s
e
o
t
h
e
r
w
i
s
e
[
o
t
h
e
r
w
i
s
e
]
[
o
t
h
e
r
w
i
s
e
]
[
o
t
h
e
r
w
i
s
e
]
[
o
t
h
e
r
w
i
s
e
]
c
o
n
t
r
a
r
i
w
i
s
e

c
o
n
t
a
r
i
w
i
s
e
c
o
n
t
r
a
r
i
w
i
s
e
[
c
o
n
t
r
a
r
i
w
i
s
e
]
[
c
o
n
t
r
a
r
i
w
i
s
e
]
[
c
o
n
t
r
a
r
i
w
i
s
e
]
c
o
n
t
r
a
r
i
w
i
s
e
[
c
o
n
t
r
a
r
i
w
i
s
e
]
c
o
n
t
r
a
r
i
l
y

[
c
o
n
t
r
a
r
i
l
y
]
[
c
o
n
t
r
a
r
i
l
y
]
[
c
o
n
t
r
a
r
i
l
y
]
[
c
o
n
t
r
a
r
i
l
y
]
c
o
n
t
r
a
r
i
l
y
[
c
o
n
t
r
a
r
i
l
y
]
c
o
n
t
r
a
r
y

[
c
o
n
t
r
a
r
i
w
a
y
s
]
o
n

t
h
e

o
t
h
e
r

h
a
n
d

o
n

t
h
e

o
t
h
e
r

h
a
n
d
o
n

t
h
e

o
t
h
e
r

h
a
n
d
o
n

t
h
e

o
t
h
e
r

h
a
n
d
o
n

t
h
e

o
t
h
e
r

h
a
n
d
r
a
t
h
e
r
i
n
s
t
e
a
d
The diachrony of English contrastive/concessive adverbial connectors 177
of todays contrastive/concessive connectors are lexicalized nominal or prepo-
sitional phrases, which are still comparatively transparent, precisely because
they comprise lexical elements.
One sub-group expresses the correction by an adjectival or nominal lexeme
meaning contrary, such as conversely or contrary in contrary, contrarily,
contrariwise and contrariways (cf. also by/in contrast, on the contrary which
have not been included in the present study because they are fully transparent;
see above, Chapters 1.6 and 7.4).
Another sub-group conveys the sense of correction by phrases comprising
the adjective other, for example in lexicalized nominal phrases, such as ME
other side, on other half or on the other hand (with a metaphorical extension of
the spatial or body noun on the written space); the adjective other is also found
in original prepositional phrases, such as otherwise and otherways (< in/on
other wise; for the suffx -wise/-ways see above, Chapter 7.3.3).
Reformulatory else is almost exclusively attested in collocations with the
adversative/reformulatory coordinating conjunctions but and or and thus con-
trasted to additive ellis/else (which is found in collocations with and; see Ap-
pendix B.4.3). The OED (s. v. else) even bases its introductory remarks on else
on its synonymy with other: In this use else, like its synonym other, admits
contextually of two different interpretations: e. g. something else may mean
something in addition to what is mentioned, or something as an alternative
or a substitute.
Both of the forms which come to be used as adverbial connectors in Present
Day English instead and rather are formed from Germanic word material
and originally go back to Old English. In their original meanings, they intrigu-
ingly denote two opposite spatial-temporal concepts: OE stede standing still,
as opposed to movement (see OED, s. v. stead n., 1.) as opposed to OE hraor
faster, the comparative of hrae/hre quickly, rapidly (see OED, s. v.
rathe, adj. Denoting rapidity in the performance or completion of an action).
Although the noun stead is no longer used in English on its own, instead (short-
ened from the anaphoric, pronominal form instead of this) is still transparent
in its spatial origin through compounds such as homestead or place names in
-stead such as Grinstead, Halstead or Hampstead (on the diachrony of instead,
see Fischer 2007: 277280). Rather, on the other hand, does no longer seem to
be transparent in its temporal-spatial origin for speakers of Modern English
(see Rissanen 1999a). It is thus the only of these adverbial connectors marking
correction which is not transparent.
178 CONTRAST/CONCESSION
10.4.3. Contrastive/concessive connectors in English: general tendencies
Table 10.3 below lists the adverbial connectors expressing the other external and
internal sub-groups of the semantic relations contrast/concession (see Table
10.1, sub-groups (A), (B), (C) and (E)). This chart shows that the developments in
English adverbial connectors agree with what has been found for other languages
and also for the history of English subordinators. English has like most of the
European languages a wide variety of concessive markers, and newcomers have
been added to this category over all periods of English. Parallel to the develop-
ment in other languages, also many of the contrastive/concessive connectors have
developed rather late in the history of English. The adverbial connectors emerg-
ing from the middle of the Early Modern English period onwards have shown
themselves to be fairly persistent (only the variant howsoever has not survived
into Present Day English), in contrast to the concessive subordinators coined in
the younger periods of English, most of which were fairly ephemeral.
10.4.4. Shifting deictics in English contrastive/concessive connectors
The history of contrastive/concessive connectors reinforces most of the general
tendencies found for the development of English causal connectors. Old Eng-
lish employed a number of explicitly deictic items,
3
namely forms comprising
swa so, such as swaeah and swaeahhwere, and the though rare pro-
nominal connector nayls no + the + less comprising a form of the demon-
strative (locative or instrumental y/e).
Of these, only nevertheless has survived into Present Day English. Its constitu-
ent the, however, is no longer transparent as a form of the demonstrative and is
classifed as an adverb in the OED (s.vv. the, thy). The increasing loss of transpar-
ency of e/y after the Old English period (see above, Chapter 7.5) is illustrated
by the many variant forms of the word during the late Middle English and Early
Modern English period, in particular in spellings which fuse the original demon-
strative element and thus render the form fully opaque (see OED, s. v. natheless
and OED, MED, s.vv.): consider, for example ME nadleez, naughtles, nales or
ME/EModE nathles, naithlesse, nathlesse, nathless, nothless, or forms which
abandon this element altogether, such as ME nether-les, never-les(se) or never-
later. Interestingly, the MED (s. v.) only labels nether-les an aberrant form (cit-
ing about ffteen instances, all of them from the ffteenth century).
3 For the use of the term deixis in the present study, which includes discourse dei-
xis, see above, Chapter 9.4.1.
The diachrony of English contrastive/concessive adverbial connectors 179
T
a
b
l
e

1
0

:

A
d
v
e
r
b
i
a
l

c
o
n
n
e
c
t
o
r
s

m
a
r
k
i
n
g

c
o
n
t
r
a
s
t
/
c
o
n
c
e
s
s
i
o
n

i
n

t
h
e

h
i
s
t
o
r
y

o
f

E
n
g
l
i
s
h
Old
English
11001350
(ME 1/2)
13501420
(ME 3)
14201500
(ME 4)
15001570
(EModE 1)
15701640
(EModE 2)
16401710
(EModE 3)
17101780
(LModE 1)
17801850
(LModE 2)
18501920
(LModE 3)
1920
(PDE)
h
u
r
u
h
u
r
u

i
n
g
a

e
a
h

e
a
h

e
a
h

t
h
o

t
h
o
u
g
h

t
h
o
u
g
h

e
a
h
-
h
w

e
r
e

e
a
h
-
h
w

e
r
e
s
w
a

e
a
h
s
w
a

e
a
h
s
w
a

e
a
h
-

h
w

r
e
n
a

y
l

s
n
a
t
h
e
l
e
s

n
a
t
h
e
l
e
s
n
a
t
h
e
l
e
s
y
e
t

y
e
t
y
e
t
y
e
t
y
e
t
y
e
t
y
e
t
y
e
t
y
e
t
y
e
t
[
n
e
v
e
r
t
h
e
-

l
a
t
t
e
r
]
[
n
e
v
e
r
t
h
e
l
a
t
t
e
r
]
n
e
v
e
r
t
h
e
l
a
t
t
e
r
[
n
o
u
g
h
t
-
f
o
r
-
t
h
a
n
]
[
n
o
u
g
h
t
-
f
o
r
-
t
h
a
n
]
[
n
o
u
g
h
t
-
f
o
r
-
t
h
a
n
]
n
e
v
e
r
t
h
e
l
e
s
s

n
e
v
e
r
t
h
e
l
e
s
s
[
n
e
v
e
r
t
h
e
-
l
e
s
s
]
[
n
e
v
e
r
t
h
e
-
l
e
s
s
]
[
n
e
v
e
r
t
h
e
-
l
e
s
s
]
[
n
e
v
e
r
t
h
e
-
l
e
s
s
]
[
n
e
v
e
r
t
h
e
-
l
e
s
s
]
[
n
e
v
e
r
t
h
e
-
l
e
s
s
]
[
n
e
v
e
r
t
h
e
-
l
e
s
s
]
[
n
o
u
g
h
t
-
f
o
r
-
t
h
a
t
]

[
n
o
u
g
h
t
-
f
o
r
-
t
h
a
t
]
[
n
o
u
g
h
t
-
f
o
r
-
t
h
i
]

[
n
o
u
g
h
t
-
f
o
r
-
t
h
i
]

a
l
g
a
t
e
s

[
a
l
g
a
t
e
s
]
[
a
l
g
a
t
e
s
]
n
o
t
w
i
t
h
s
t
a
n
d
i
n
g

n
o
t
w
i
t
h
-
s
t
a
n
d
i
n
g
n
o
t
w
i
t
h
-
s
t
a
n
d
i
n
g
n
o
t
w
i
t
h
-
s
t
a
n
d
i
n
g
n
o
t
w
i
t
h
-
s
t
a
n
d
i
n
g
h
o
w
b
e
i
t

h
o
w
b
e
i
t
h
o
w
b
e
i
t
[
n
o
n
e
t
h
e
l
e
s
s
]

[
n
o
n
e
t
h
e
l
e
s
s
]
[
n
o
n
e
t
h
e
l
e
s
s
]
[
n
o
n
e
t
h
e
l
e
s
s
]
[
n
o
n
e
t
h
e
l
e
s
s
]
[
n
o
n
e
t
h
e
l
e
s
s
]
n
o
n
e
t
h
e
l
e
s
s
a
l
b
e
i
t

a
l
b
e
i
t
[
h
o
w
s
o
m
-
e
v
e
r
]
[
h
o
w
s
o
m
-
e
v
e
r
]
[
h
o
w
s
o
m
-
e
v
e
r
]
[
h
o
w
s
o
m
-
e
v
e
r
]
[
h
o
w
s
o
m
-
e
v
e
r
]
[
h
o
w
s
o
m
-

e
v
e
r
]
180 CONTRAST/CONCESSION
Old
English
11001350
(ME 1/2)
13501420
(ME 3)
14201500
(ME 4)
15001570
(EModE 1)
15701640
(EModE 2)
16401710
(EModE 3)
17101780
(LModE 1)
17801850
(LModE 2)
18501920
(LModE 3)
1920
(PDE)
h
o
w
e
v
e
r

h
o
w
e
v
e
r
h
o
w
e
v
e
r
h
o
w
e
v
e
r
h
o
w
e
v
e
r
h
o
w
e
v
e
r
h
o
w
s
o
e
v
e
r


h
o
w
s
o
e
v
e
r
a
t

l
e
a
s
t

a
t

l
e
a
s
t
a
t

l
e
a
s
t
a
t

l
e
a
s
t
a
t

l
e
a
s
t
a
t

l
e
a
s
t
(
i
n

t
h
e
)

m
e
a
n
t
i
m
e

[
i
n

t
h
e

m
e
a
n
-
t
i
m
e
]
[
i
n

t
h
e

m
e
a
n
-
t
i
m
e
]
i
n

t
h
e

m
e
a
n
-
t
i
m
e
[
i
n

t
h
e

m
e
a
n
-
t
i
m
e
]
[
i
n

t
h
e

m
e
a
n
t
i
m
e
]
(
t
h
e
)

m
e
a
n
-
w
h
i
l
e

[
m
e
a
n
w
h
i
l
e
]
[
m
e
a
n
w
h
i
l
e
]
[
m
e
a
n
w
h
i
l
e
]
[
m
e
a
n
w
h
i
l
e
]
[
m
e
a
n
-
w
h
i
l
e
]

[
a
f
t
e
r

a
l
l
]
a
f
t
e
r

a
l
l
a
f
t
e
r

a
l
l
a
f
t
e
r

a
l
l
a
f
t
e
r

a
l
l
s
t
i
l
l

s
t
i
l
l

s
t
i
l
l
s
t
i
l
l
s
t
i
l
l
a
t

t
h
e

s
a
m
e

t
i
m
e

a
t

t
h
e

s
a
m
e

t
i
m
e
a
t

t
h
e

s
a
m
e

t
i
m
e
a
t

t
h
e

s
a
m
e

t
i
m
e
o
f

c
o
u
r
s
e

o
f

c
o
u
r
s
e
o
f

c
o
u
r
s
e
i
n

f
a
c
t

i
n

f
a
c
t
i
n

f
a
c
t
[
a
n
y
h
o
w
]
[
a
n
y
h
o
w
]
[
a
n
y
h
o
w
]
i
n

a
l
l

e
v
e
n
t
s

i
n

a
l
l

e
v
e
n
t
s
i
n

a
l
l

e
v
e
n
t
s
[
a
n
y
w
a
y
]
[
a
n
y
w
a
y
]
[
a
l
l

t
h
e

s
a
m
e
]
[
a
l
l

t
h
e

s
a
m
e
]
i
n

a
n
y

c
a
s
e

i
n

a
n
y

c
a
s
e
The diachrony of English contrastive/concessive adverbial connectors 181
As in the case of the causal connectors (see above, Chapter 9.5), speak-
ers and writers of Middle English tried to preserve these pronominal forms
or, alternatively, coined new ones with the new demonstrative that (cf. ME
nought-for-than, nought-for-thi nevertheless and nought-for-that neverthe-
less, all of them ambiguous adverbs/conjunctions). Yet, with the exception of
nevertheless, none of these has survived the Middle or Early Modern English
period. Middle English thus again emerges as a period of experiment and tran-
sition (see above, Chapter 6.6).
After the Middle English period, we fnd, just as in the history of causal
connectors, a number of newcomers which instead of demonstrative deixis
employ time deixis as a frame of reference, such as (in the) meantime, (in the)
meanwhile, after all, still, and at the same time (see above, Chapter 8.2). While
German also employs originally time deictic elements (see the list in Rudolph
1996: 45), it has again as in the case of causal and additive connectors pre-
served many connectors comprising a specifcally pronominal element, such as
PDG dessen ungeachtet, indes, indessen, nichtsdestoweniger, nichtdestotrotz
(colloquial; mixture of nichtsdestoweniger and trotzdem), trotzdem or an ele-
ment of space deixis such as da there or hin hence, such as PDG dagegen,
hingegen or archaic hinwiederum.
As in the semantic relation of cause, English adverbial connectors express-
ing contrast/concession have virtually given up the old system of pronomi-
nal connectors, which is still very much alive in German, and have sought
other means for the coinage of new adverbial connectors.
10.4.5. Patterns in the origin of English concessive connectors
Following the abandonment of the coinage of pronominal connectors after
the Middle English period, speakers and writers of English have employed
those means which have above been introduced as cross-linguistic patterns for
new coinages in the feld of new contrastive/concessive connectors (see Knig
1985b, and above, Chapter 10.4, patterns (I)(V)). For the history of adverbial
connectors, Patterns (I) and (III) are most important:
Pattern (I): Universal quantifcation (OE hwer whether; PDE all, any, ever)
OE hwer which of two, each of two, either, any
OE hwer(e)
eahhwere
swaeahhwere
182 CONTRAST/CONCESSION
all
ME algates (lit. all gates every way; ON alla gtu)
EModE1 albeit
EModE2 after all
LModE3 in all events
LModE3 all the same
any
LModE2 anyhow
LModE3 anyway
LModE3 in any case
ever
EModE1 howsomever
EModE2 however
EModE2 howsoever
As has been pointed out above, these forms comprising universal quantifers
are intrinsically cohesive because they always presuppose that something has
preceded. Their path to becoming concessives is most clearly seen in forms
such as however, howsoever and howsomever, since they are relics of original
subordinate clauses how(so)(m)ever this may be (see OED, s. v. however),
comprising in the full clause an anaphoric element such as this. By the conven-
tionalization of these conversational implicatures, these dismissive connectors
may become contrastive connectors proper (cf. however).
Their generalized sense, however, recurrently leads to further semantic
bleaching, so that many of these items follow the adverbial cline from ad-
verb to discourse marker suggested in the semantic-pragmatic approach to
grammaticalization (see Traugott and Dasher 2002; for a detailed discussion,
see above, Chapter 8.4). It is therefore probably no coincidence that neither
anyhow (which is stated to be the fourth most frequent linking adverbial in
spoken British English) nor anyway (fourth most frequent linking adverbial
in spoken American English; Biber et al. 1999: 887) are attested in my corpus,
which mainly consists of written texts, all of which are non-interactive (on the
discourse structuring function of anyhow and anyway, see Lenk 1998; on after
all, see Traugott 1997 and 2004: 554560).
Pattern (III): Remarkable co-existence or co-occurrence of two facts
Pattern (IIIa): Negation (asserting that one fact p does not prevent another fact q)
As the many adverbial connectors (and subordinators) comprising an element
of negation (see na-/no, none-, never-) show, there is also a close correlation
between counter-expectancy as one of the properties of concession/contrast
The diachrony of English contrastive/concessive adverbial connectors 183
and negatives. Commonly, negatives are only articulated in contexts where the
corresponding affrmatives have already been discussed, or where the speakers
or writers assume the hearers or readers belief in or at least their familiarity
with the corresponding affrmative.
Assumptions like these have led to the concessive meanings of English items
comprising a negative. These are attested over all the periods of English, in
various pronominal and non-pronominal patterns.
4
no + for + demonstrative
ME nought-for-than, nought-for-thi, nought-for-that
no + y/e + less
OE nayls
ME natheless, nautheless naught + the + less, netheless,
not-the-lesse, notheless, noughttheless
EModE1 nonetheless
In a mixture of Pattern (I) (universal quantifers) and (III) (negation), we fnd
forms comprising the negative of the universal ever, never:
5
never + the + less
ME nevertheless (variants: ner-the-les, never-ther-les,
never-tho-les)
never + the + latter
ME ner-the-latere, ME/EModE never the later
As a further negative, we fnd notwithstanding (a calque on French non ob
stant), which is only used as a preposition in Present Day English, but is also
habitually employed as a contrastive/concessive adverbial in the prepositional
phrase notwithstanding this.
Pattern (IIIb): Simultaneity or unhindered continuation
Communicative principles have also led to the concessive meaning of certain
temporal adverbs signalling simultaneity or unhindered continuation, in par-
ticular yet and still (see Knig and Traugott 1982; Knig 1985a, 1985b), but also
4 In addition to the connectors listed below, the MED also lists the coinages not + the
+ more (ME no-the-mo, EModE nathemore, nathemo) and never-the-whether, for
which I could fnd an attestation as an adverbial connector neither in the dictionar-
ies nor in the corpus texts, i. e. they are not attested as adverbial connectors on the
level of the sentence or discourse.
5 Again, the MED lists forms such as neverthemore, for which I could not fnd an
attestation as an adverbial connector on the level of the sentence or discourse.
184 CONTRAST/CONCESSION
the phrases (in the) meantime and meanwhile, which are, however, only rarely
used as concessives in Present Day English (see OED, s.vv. meantime, 1b, 2.,
meanwhile B.2). In its draft revision entry on meanwhile, the OED (apparently
unaware of the research on concessives) inherently comments on these con-
versational implicatures by stating that although this is now rare as a distinct
meaning, the adversative force is still often present in the use of sense 1 to
mean at the same time. The two senses are not always readily distinguished
. For such ambiguous, contextually concessive uses in Late Modern Eng-
lish, see, for example,
(137) then the consequent practice which they direct shall be explained, and
rendered easy of adoption. In the meantime, let no one anticipate evil,
even in the slightest degree, from these principles; they are not innoxi-
ous only, but pregnant with consequences to be wished and desired be-
yond all others by every individual in society (CLOW2).
These conversational implicatures are, by contrast, fully conventionalized in
yet and still (for examples, see below, Chapter 10.5.3).
The origins for the subsequent conventionalization of conversational impli-
catures contained in the notion continuation are clearly seen in sentences
such as (138), where the originally temporal meaning can always be augmented
by concessive implicatures, if this is consistent with the background knowl-
edge of speaker and hearer:
(138) It is midnight and he is still working.
In view of the pragmatics of negation (Pattern IIIa) discussed above, such an
augmentation is even more likely if the second clause contains a negation, as in
(139) It is midnight and he still has not stopped working.
10.5. OE eah PDE though
10.5.1. Long-term developments: grammaticalization
While the contrastive/concessive connectors discussed so far support earlier
cross-linguistic and English-specifc research on concessives, a more detailed
investigation of the history of PDE though seems to challenge one of the prop-
OE eah PDE though 185
erties of grammaticalization, the principle of uni-directionality (for a summary
on the state of research, see Brinton and Traugott 2005: 6976).
The term grammaticalization owes its origin to work by Meillet, who
characterized grammaticalization as the attribution of grammatical character
to a previously autonomous word. It is now commonly understood as that part
of the study of language that is concerned with such questions as how lexical
items and constructions come in certain linguistic contexts to serve grammati-
cal functions or how grammatical items develop new grammatical functions
(Hopper and Traugott 2003: 1). In the morpho-syntactic approach to grammati-
calization, a number of parameters have been established as prototypical, such
as, for instance, increased bonding or syntactic scope reduction (for a full list
of criteria, see Lehmann [1982] 1995).
Among the well-known properties of grammaticalization processes is also
uni-directionality (i. e., e. g., from lexicon to grammar, from less grammatical
to more grammatical and not vice versa). Most of the newly coined gram-
matical elements can be shown to follow regular, unidirectional clines. A com-
mon cline is the coinage of new conjunctions from frst of all lexicalized
nominal and prepositional phrases; see, for instance, the formation of because
from prep. be + N. cause + subordinator (that) (for the distinctive, subsequent
processes of lexicalization and grammaticalization in the coinage of new con-
junctions in Spanish, see, e. g., Lehmann 2002: 812).
In view of these facts the path of English OE eah/PDE though merits a
more detailed investigation. This is not evident at frst glance, since histori-
cal dictionaries and grammars suggest that the various forms of this linguis-
tic item have always been syntactically polyfunctional, i. e. could be used
as either adverbs or conjunctions (with mood distinction in the verb of the
relevant connect). A closer inspection of the actual language data, however,
suggests that this scenario is too simplistic, since eah was only occasion-
ally used as an adverb from Middle English to the beginning or middle of
the twentieth century. In the complete corpus of the present study, adverbial
though is only used three times in the Early and Late Modern English pe-
riod (in three different texts; see Appendix B.4.1 and examples (149)(151)
below), predominantly in post-frst position in questions. This suggests that
though has not followed a strictly unidirectional path, but a meandering way
from adverb/conjunction to conjunction to, again, adverb/conjunction. Yet,
though has in contrast to such connectors as nu or form never been
a prototypical ambiguous adverb/conjunction because of mood distinc-
tions in the verb in earlier English and because of topological distinctions
(sentence-initial vs. sentence-medial or -fnal) in Present Day English (see
below, Chapter 10.6).
186 CONTRAST/CONCESSION
The following account of this path will frst sketch the situation in Old Eng-
lish and will then, passing a long period of time in one large step, analyse the
re-emergence of adverbial (sentence-fnal) though in spoken Present Day English
in the middle of the twentieth century by an analysis of the forms of although
and though in the London Lund Corpus of Spoken English (LLC; spoken part).
In contrast to the preceding chapter on causals, the section will thus not focus
on general developments in the expression of concessivity, since, as has been
sketched above, the concessive relation has over all periods more prototypically
been expressed by subordinate or, in particular, correlative constructions (and is
also discussed widely in the relevant publications; see Quirk 1954; Knig 1985a,
1985b; Kortmann 1997; Gonzlez-Cruz 2007; Sorva 2007). For the present study,
two other aspects are essential: the re-analysis of original concessives as mark-
ers of contrast, when used as adverbial connectors, and, at the same time, their
topological fxation to the middle (written mode) or fnal position (oral mode).
10.5.2. OE eah
Similar to the relation of cause, there is in essence only one central contras-
tive/concessive connector in Old English, eah (and its variants eahhwere;
swaeah, and swaeahhwere).
In the corpus used for the present study, other forms are only rarely at-
tested, and are restricted to certain texts and contexts. Thus, occurrences
of hwere whether > nevertheless are confned to the Blickling Homilies
(OE 3), in very generalized, dismissive contexts (however that may be; see
above, Category (E)).
(140) Ac mid re bysene, he gecyde t sofste men habba mid him
eofas & synfulle men; & hwere hie sceolan heora yfel geylde a-
refnan on him selfum (COBLICK, 175).
But with this example he told us that righteous men keep the company
of thieves and sinners; and yet [however that may be], they have to suf-
fer for their sins themselves with patience.
Forms comprising a negative, such as nayls nevertheless, are not attested at
all in my corpus (similar to nonetheless, which was according to the OED
coined in the sixteenth century, but is frst attested in my corpus material in the
Present Day English material of the FROWN Corpus).
These fndings from the corpus texts are corroborated by a survey count in
the Complete Corpus of Old English in Machine Readable Form (DOEC),
OE eah PDE though 187
which yields the following numbers (for the spelling variants eah/eah, eh/
eh with smoothing of the breaking, hw-/hw- and y/y/e/e):
Table 104: Concessives in Old English: DOEC
eah ca. 3380 occurrences
swaeah ca. 960 occurrences
eahhwere ca. 290 occurrences
swaeahhwere ca. 30 occurrences
hwere eah ca. 15 occurrences
nayls ca. 10 occurrences
eah (and its variants) do thus not have any serious competitors as central
markers of the concessive relation in Old English.
6
The following survey shows that eah just like OE form could be used
as an adverbial connector, but also as a subordinating conjunction introducing
pre-posed, medial and post-posed subordinate clauses.
(141) Oft seo brodige henn, eah heo sarlice cloccige, heo tosprt hyre fyera
& a briddas gewyrm (COBYRHTF, 2.1.225).
Often the brooding hen, though she clucks sorrowfully, spreads her
wings and warms her chicks.
Just as in Present Day English, pre-posed concessives are preferred since they
facilitate the understanding of the complex relation of counter-expectancy in
concessives (see above, Chapter 10.2.4).
(142) We wylla t hig understandon isne cwide, (\Vasa fctilia tanto solent
esse utiliora quanto et uiliora\), eah we wace syn & as ing leohtlice
unwreon, hig magon fremian bet onne a e beo on leowisan fgre
geglenged (COBYRHTF, 1.2.198).
We wish them to understand this saying: Earthen vessels are more
useful the cheaper they are. Although we are weak and we explain
these things cursorily, they may do more good than those that are fairly
adorned in verse.
6 The concessive instances of hwere have not been counted because of the poly-
functionality of this word. The comparatively few instances of the collocation
hwere eah (15 in comparison with the 290 for eahhwere), however, cor-
roborate my corpus fndings that hwere was only very rarely used.
188 CONTRAST/CONCESSION
The subordinate status of the pre-posed clause may be further signalled by the
general subordinating particle OE e, as in
(143) eah e ealle dagas lce geare habbon heora (\concurrentes\), syn-
derlice se dg e by on (\ix kl Aprl\) getacna hu fela beo on geare
(COBYRHTF, 1.2.92).
Though all the days in each year have their concurrents, the day that
falls on 24 March shows how many there are in the year.
The double function of eah is most obvious in correlative constructions. As
has been shown above (see Chapter 5.3.2), this does, in contrast to causal con-
nectors, not present major problems because the functions are differentiated by
the mood of the verbs in the respective clause (see the subjunctive scine in the
subordinate clause (144a), and the indicative astih in the main clause (144b)).
(144) & eah e [144a; conjunction] seo sunne under eoran on nihtlicere
tide scine, eah [144b; adverb] astih hire leoht on sumere sidan re
eoran (COTEMPO, 1.32).
And although the sun shines under the earth at night, her light still
comes on the summer side of the earth.
In contrast to concessive subordinators, contrastive/concessive adverbial con-
nectors tend to express a contrastive rather than the purely concessive meaning
of counter-expectancy.
(145) & eah is [3 Prs. Sg. Ind.] an God, se is stemn & staol eallra goda;
of m cuma eall good, & eft hi fundia to him, & he welt ealra
( COBOETH, 34.86.1).
And yet there is one God, who is the stem and the foundation of all
goods and from him comes all good; and again, they tend to him and he
governs all .
The adverbial function of eah is here frst illustrated by an example from
the Old English Boethius adaptation, not only because this is a comparatively
independent Old English text (a criterion most important for the use of the
subjunctive), but also because late Old English authors increasingly distinguish
the subordinator from adverbial functions by the use of different connectors
and different topologies.
Already in the Old English Boethius, post-frst or medial position is pre-
ferred for eah in its use as an adverbial connector (in contrast to the obligatory
clause-initial position for the subordinator):
OE eah PDE though 189
(146) t is eah micel syn to geencanne be Gode, tte nig god sie buton
on him, (COBOETH 34.84.31).
It is, however, great sin to imagine concerning God, that any good be
external to him .
In Late Old English, authors increasingly try to distinguish the different func-
tions by the use of different lexemes. For the adverbial connector, longer, more
complex and more explicitly deictic items are chosen, which signal the equal
weight of the propositions of the two connects iconically. Consider the fol-
lowing example with adverbial swaeah (so + eah), which is found almost
identically in Byrhtferths Enchiridion and lfrics De Temporibus Anni:
(147) Synd swa eah ma heofena, swa swa se witega cwy: (\Celi celorum\)
(COBYRHT, 2.1.257).
Sind swa eah ma heofenan, swa swa se witega cw; Celi celorum,
t is heofena heofenan (COTEMPO, 1.8).
Yet, there are more heavens, as the prophet says: Heaven of heavens.
For the adverbial connector, lfric in particular prefers the unambiguous
forms swaeah, eahhwere (in Catholic Homilies I) or the very complex
and therefore highly iconic swaeahhwere (in Catholic Homilies II), mainly
placed in post-frst or medial position. There are very few instances of simple
eah used as an adverbial connector (see Godden 2000: Glossary, s.vv. eah,
eahhwere, swaeahhwere). See, for example, the adverbial connector
swaeahhwere in post-frst position in
(148) Is swa eah hwere an ra eahta winda, (\Aquilo\) gehaten, se
blw noran & eastan . (COTEMPO, 10.23).
Yet, there is one of these eight winds, which is called Aquilo, which
blows from the north and the east.
As a subordinator, lfric almost exclusively employs the complex form eah
e, comprising the general Old English subordinating particle e. This shows
that even in Old English writers tried to differentiate the connectors for the
semantically and cognitively rather complex relation of concession and em-
ployed more explicit forms for the adverbial connectors signalling contrast.
These were also distinguished topologically and were predominantly placed in
post-frst or medial position.
190 CONTRAST/CONCESSION
10.5.3. Contrastive adverbial connectors from Middle English to Present
Day English
10531 though
In the discussion of the general tendencies in the developments of adverbial
connectors in the history of English, we have seen that the inventory of contras-
tive/concessive connectors is considered the least stable one over all the peri-
ods of English (see Kortmann 1997: 347). Yet, most of the new coinages from
the Early Modern English period at least, in the meantime, in the meanwhile,
after all, at the same time, of course, in fact, anyhow, in all events, anyway, all
the same and in any case express the dismissive sub-type of concessives
(Type (E)) only.
The inventory of connectors is much more stable with respect to the contras-
tive/concessive relation proper (Types (A), (B) and (C)). The forms continuing
OE eah the Scandinavian loan though
7
and the originally emphatic form
although (with the universal quantifer all stressing the concession) which
is used as a non-emphatic form from around 1400, are predominantly used as
subordinators from the Middle English period onwards. In my corpus texts, I
have only found very few instances of adverbial though from the late Middle
English and Early Modern English period (see Appendix B.4.1). Most of these
are, however, not attested in contexts prototypical for adverbial connectors. In
the example taken from Reynes Commonplace Book (ME3), both instances of
adverbial though work on the level of the phrase (subject complement: dredfull
and owg gracious):
(149) Whoso be born in Aries, he schall be dredfull and owg gracious.
Cancer: he schall ben pore and weyke, and owg gracious
(CMREYNES, p.241).
Whoever is born in Aries, he will be anxious, but nonetheless genial.
Cancer: He will be poor and weak, but nonetheless genial.
7 The form though is not a continuation of OE eah, but a Scandinavian loan attested
from the twelfth century onwards (probably from unattested ON *h). Samuels
argues that it spread out like its competitor, the plural pronoun they from
its original Danelaw area in answer to a functional need for clear communica-
tion during the ffteenth century. He fnds that they was the form frst adopted into
the southern area, which, however, already possessed that form in the meaning
though (cf. the attested Middle English spellings eigh, ei, ey, eigh, thigh or
thei continuing OE eah; see OED, s. v. though). The resulting homonymic clash
therefore is seen to have triggered the borrowing of though shortly afterwards (cf.
Samuels 1972: 7171 and 70, map 2 and OED, s. v. though).
OE eah PDE though 191
In the two instances from the Early Modern English period, though is used in
post-frst position in questions:
(150) (Dan.) It is not in these matters to be taken as wee imagine, but as the
word of God teacheth. What though a man think he worshippeth not
deuils, nor seeketh not help at their handes, as he is persuaded, nor hath
any such intent, is he euer the neere, when as yet it shall be found by
Gods word, that he doth worship them, and seek vnto them for help?
(CEHAND2A, p. B2R).
(151) But the vulgar cares not for this. What tho? (CEBOETH2 (Queen Eliza-
beth), p.87).
In example (152), it is not quite clear whether the second though (152c) in
spite of being strengthened by the disjunct methinks is indeed used as an
adverbial connector or rather as a post-posed subordinate clause, further modi-
fying the correlative construction although (152a) yet (152b):
(152) Now, though [152a; conjunction] I have with great diligence endeav-
oured to fnd whether there be any such thing in those (Microscopical)
pores of Wood or Piths, as the (Valves) in the heart, veins, And other
passages of Animals, that open, And give passage to the containd fuid
juices one way, And shut themselves, And impede the passage of such
liquors back again, yet [152b; adverb] have I not hitherto been able to say
any thing positive in it; though [152c], me thinks, it seems very proba-
ble, that Nature has in these passages, as well as in those of Animal bod-
ies, very many appropriated Instruments And contrivances, whereby to
bring her designs And end to pass, which tis not improbable, but that
some diligent Observer, if helpd with better (Microscopes) , may in time
detect (CESCIE3A. p.13.5,116).
In sum, we see that though was only rarely used as an adverb in the Middle
English, Early Modern and Late Modern English period (see also the paucity
of quotes for these periods in the OED, s. v. though I.a).
10532 Yet and still
Instead of adverbial though, speakers and writers choose different forms
coined on the patterns introduced above, in particular yet and still, which both
employ the temporal Pattern (IIIb) Simultaneity or unhindered continuation.
192 CONTRAST/CONCESSION
While the conversational implicatures are conventionalized for yet as early as
the beginning of the Middle English period, still only shows unambiguous con-
cessive meanings from the beginning of the eighteenth century onwards. These
developments will be summarized in the following sections.
In its earlier temporal meanings, yet implied a continuance from a previous
time up to and at the present (or some stated) time: Now as until now (or then as
until then): = still 4a. (OED, s. v. yet II.2.a). In lfrics Grammar, for instance,
gyt is given as an equivalent of Latin adhuc up to now, as yet, still (Zupitza
[1880] 2001: 237), but also used to illustrate forms of the Latin future or future
perfect, such as amabo ic lufge gyt to dg oe to merien or amato tu lufa
u gyt in contrast to amo ic lufge (Zupitza [1880] 2001: 130, 131, 133).
These temporal meanings are often emphasized by another temporal adverb,
as in the collocation with nu in (153):
(153) Ac u ana hwilum bescylst mid ore eagan on a heofonlican incg, mid
ore u locast nu giet on as eorlican (COBOETH, 121.30).
But thou alone sometimes lookest with one eye on the heavenly things
and with the other thou lookest as yet on these earthly things.
The OED (s. v. yet III.9) and, following the OED, Knig and Traugott (1982)
see the frst instances of concessive meanings of yet in the Early Middle Eng-
lish Period. Examples such as (154), however, are certainly ambiguous even in
Old English and allow a concessive reading as a conversational implicature, a
reading supported by the collocation with the adversative conjunction ac but:
(154) a cw ic: Nis nan ing sore onne t; ne magon we nanwuht fndan
betere onne God. a cw he: Ac ic wolde giet mid sumre bisne e
behwerfan utan t u ne mihtst nnne weg fndan ofer (COBOETH,
85.2023).
Then said I, Nothing is truer than that. We are not able to discover any-
thing better than God. Then said he: But I would now/still prepare thee
by some examples, so that thou may not fnd any way of escaping.
A concessive reading is even more likely in instances such as (155), where giet
may be seen to function as the frst connector in a correlative concessive con-
struction (giet eah):
(155) Ac ic e mg giet tcan oer ing e dysegum monnum wile incan
giet
8
ungelefedlicre, & is eah genog gelic am spelle e wit fter-
spyria (COBOETH, 38.118.19).
8 Giet here functions as a focus particle on the phrase level, i. e. even.
OE eah PDE though 193
But I can now/still teach you another thing, which to foolish men will
seem even more incredible, and is nevertheless suitable enough to the
argument we are holding.
The frst unambiguous, concessive examples, showing a conventionalization
of these implicatures, are found in the Early Middle English period. Yet is,
however, not a prototypical adverbial connector in its concessive uses, be-
cause it is not free in its position in the sentence, but is restricted to clause-
initial position:
(156) A Meiden eode ut to bihalden uncue wummen. get ne sei hit nawt
et ha biheold wepmen (CMANCR, p.32).
A girl went out to watch unknown women. Yet, it does not say that she
watched soldiers.
Yets function as an adverbial connector is highlighted by its many occurrences
in correlative constructions, and in particular, in collocations with conjunc-
tions (ac giet, and yet and but yet) and other adverbial connectors (and yet
again, yet nevertheless or yet still) (see Appendix B.4.1).
While yet developed its additional contrastive/concessive sense as early as
the beginning of the Early Middle English period, still was exclusively used as
a circumstance adverb, modifying verbs such as stand in the sense of quietly,
until the end of the Early Modern English period (OED, s. v. still I.). From the
sixteenth century onwards, we fnd uses indicating the continuance of a previ-
ous action or condition, so that still acquires the temporal reading now and is
contextually used for as yet (OED, s. v. still 4.).
(157) Temporal still:
(Arch.) Dont mistake me, (Aimwell), for tis still my Maxim, that there
is no Scandal like Rags, nor any Crime so shameful as Poverty (Far-
quhar, Beaux Stratagem, CEPLAY3B, p.3).
From these temporal meanings, the contrastive/concessive sense arose at the
beginning of the eighteenth century (OED, s. v. still 6.). For very early ex amples
from Farquhars The Beaux Stratagem (1707), see (158) and (159).
(158) Contrastive still in correlative though still:
(Aim.) Tho the whining part be out of doors in Town, tis still in force
with the Country Ladies; And let me tell you (^Frank^), the Fool in that
Passion shall outdoe the Knave at any time (Farquhar, Beaux Strata gem,
CEPLAY3B, p.6).
194 CONTRAST/CONCESSION
The beginning of the eighteenth century then also sees still frst used as an
adverbial connector:
(159) Adverbial connector still nevertheless:
Rise, and know, I am a Woman without my Sex, I can love to all the
Tenderness of Wishes, Sighs and Tears But go no farther Still to
convince you that Im more than Woman, I can speak my Frailty, con-
fess my Weakness even for you (Farquhar, Beaux Stratagem, CE-
PLAY3B, p.58).
10533 However
Another recurrent pattern employed by speakers and writers for the coinage of
new adverbial connectors is the use of the full sentences as dismissives (Type
(E)), i. e. generalizing sentences which presuppose that something has gone
before, such as however that may be.
Accordingly, forms such as howbeit and albeit (see Sorva 2007), which addi-
tionally invoke a concessive meaning by the use of the subjunctive be, are used
as adverbial connectors in late Middle English and Early Modern English (see
Appendix B.4.1 and MED, s. v. hou-be-it (a)).
(160) The Front or the Forhead conteyneth nothing but the Skinne and Mus-
culus feshe, for the panicle vnderneth it is of Pericranium, and the bone
is of the Coronal bone. Howebeit there it is made broade, as yf ther were
a double bone, whiche maketh the forme of the Browes (CESCIE1A,
p.34).
(161) When I was before the Lordes at Lambeth, I was the frst that was called
in, all beit, Maister Doctour the Vicar of Croydon was come before me,
and diuers other (CEPRIV1, Corp. Morelet to Margaret Roper, l. 1).
The most prominent of these is, of course, however, which has come to be the
central adverbial connector marking contrast in Present Day English (see
Halliday and Hasan 1976: 255; Altenberg 1999: 250; Biber et al. 1999: 851), in
particular in its non-initial uses in a separate tone group (for details, see below,
Chapter 13). For a better survey, the data of my corpus fndings are repeated
here (see Appendix B.4.1):
OE eah PDE though 195
however Contrastive/Concessive
however, initial position
EModE3 CEHAND3B, CEEDUC2B, CEPRIV3, CEPLAY3A
LModE1 CLSERM1A, CLSERM1B, CLREY1, CLBUR1,
CLPRIES1, CLPAR1, CLPRIV1A, CLPRIV1B, CLDIA1A
LModE2 CLHAZ2, CLPRIV2B, CLOFFIC2A, CLOFFIC2C
LModE3 CLEWEL3, CLCARP3
PDE LOB-D, FROWN-D, FLOB-F
however, medial position
EModE3 CEPRIV3
LModE1 CLSERM1A, CLSERM1B, CLHUM1, CLJOHPF1, CLSMI1,
CLREY1, CLPRIES1, CLPAR1, CLDIA1A
LModE2 CLGOD2, CLBLAI2, CLMAL2, CLBABPF2, CLBAB2,
CLOWPF2, CLOW2, CLFOS2, CLHAZ2, CLELLI2, CLSERM2A,
CLSERM2B, CLPRIV2A, CLOFFIC2A, CLOFFIC2C
LModE3 CLBAGPH3, CLHUX3, CLARN3, CLPAT3, CLBOW3,
CLCHES3, CLHUL3, CLCARP3, CLCHEY3, CLJAM3,
CLBOETH3, CLBURY3, CLHOM3A
PDE LOB-D, LOB-F, FROWN-D, FLOB-F
however, fnal position
EModE3 CEPLAY3A, CEPLAY3B
LModE2 CLPRIV2A
Collocations: and however
EModE3 CESCIE3B
but however
EModE3 CEDIAR3A
LModE1 CLSERM1B
This entry shows that in the corpus texts however frst occurred as an ad-
verbial connector marking contrast/concession in the late Early Modern
English period (EModE3 = 16401710),
9
since when it has been attested in a
wide variety of texts. It is often found in initial position (also in the colloca-
tions and and but however), but also and more frequently in medial position
(see the underlined texts; underlining refers to more than 5 occurrences per
5,000 words). For this topological variation, see two examples from one text
in LModE3, a sermon by Joseph Butler (CLSERM1A). Both of them also very
9 Here the OED has a slightly earlier date in EModE2 (1613) for use of however as an
adverbial connector (OED, s. v. however 3.), so that this date has been used for the
overall analyses.
196 CONTRAST/CONCESSION
nicely illustrate the contrastive function of adverbial however (162a, 163a) in
contrast to the purely concessive, correlative constructions though yet (162b
and c, 163b and c):
(162) However [162a], though [162b] all this be allowed, as it expressly is by
the inspired writers, yet [162c] it is manifest that Christians at the time
of the Revelation, and immediately after, could not but insist mostly
upon considerations of this latter kind (CLSERM1A, Butler, Sermon 1).
(163) I must, however [163a], remind you that though [163b] benevolence and
self-love are different, though [163c] the former tends most directly to
public good, and the latter to private, yet [46d] they are so perfectly co-
incident that the greatest satisfactions to ourselves depend upon our hav-
ing benevolence in a due degree; and that self-love is one chief security
of our right behaviour towards society (CLSERM1A, Butler, Sermon 1).
Already in its earliest sub-period, however is also attested sentence-fnally,
but only in texts which record oral or colloquial language, notably two plays
(Vanburghs The Relapse and Farquhars The Beaux Stratagem; CEPLAY3A
and 3B) and a letter by Jane Austen from the second sub-period of Late Modern
English (CLPRIV2A).
(164) (Lov.) (running to him) I hope I hant killd the Fool however Bear him
up! Wheres your Wound? (CEPLAY3A, Vanbrugh, The Relapse, p. I, 39).
(165) (Dor.) O, Madam, had I but a Sword to help the brave Man?
(Boun.) Theres three or four hanging up in the Hall; but they wont
draw. Ill go fetch one however (CEPLAY3B, Farquhar, The Beaux
Stratagem, p.63).
(166) I danced twice with Warren last night, and once with Mr. Charles Wat-
kins, and, to my inexpressible astonishment, I entirely escaped John
Lyford. I was forced to fght hard for it, however. We had a very good
supper, (CLPRIV2A, Jane Austen, Letters to her Sister Cassandra
and Others, January 9, 1796).
This suggests that sentence-fnal uses of adverbial connectors such as how
ever, which are still much more common in spoken than in written Present
Day English, started in the oral mode. This diachronic observation and the
corpus fndings for Present Day English instigated the following analysis for
Sentencefnal connectors in Present Day English 197
the re-emergence of though as an adverbial connector in Present Day English
by an analysis of the data collected in the London Lund Corpus of Spoken
English.
10.6. Sentence-fnal connectors in Present Day English
10.6.1. Corpus fndings
Among the commonly accepted topological criteria used for distinguishing
conjunctions from adverbial connectors is the position of the connector: While
conjunctions are always placed clause-initially, adverbial connectors are fexi-
ble in their positions in the sentence (see above, Chapter 2). In most Indo-Euro-
pean languages, possible positions for adverbial connectors are the initial and
the medial position (for the different medial positions, among them the specifc
post-frst position, see above, Chapter 5.4 and below, Chapter 13.213.4) with,
of course, differences in the respective languages.
10
In Present Day English, however, we also fnd fnal adverbial connectors.
These prove to be very common in the spoken mode (see Table 10.5):
Table 105: Positions of adverbial connectors in conversation and academic
prose (adapted from Biber et al. 1999: 891, Biber et al.s Table 10.18)
% in initial position % in medial position % in fnal position
CONV

ACAD

= 5 %; = less than 2.5 %
In the corpus of the Longman Grammar, fnal adverbial connectors are attested
in both the spoken and the written medium. academic prose clearly favours
initial and medial position: only ca. 10 per cent of all adverbial connectors are
found sentence-fnally, while more than 40 per cent are attested in medial posi-
tion (see below, Chapter 13 for a discussion of this high proportion of medial
adverbial connectors and its history).
10 In German, for example, there are systematic differences between the positions of
conjunctions and adverbial connectors on the one hand and modal particles on the
other hand. The latter have to be placed in the middle feld. See Sie wollte tele-
fonieren, aber sie hatte kein Kleingeld She wanted to make a phone call but she
did not have any change (conjunction) vs. Das ist aber eine berraschung It is
indeed a surprise (modal particle) (see Diewald 1999).
198 CONTRAST/CONCESSION
In the sub-corpus conversation, on the other hand, fnal adverbial connec-
tors are used 40 per cent of all instances and are thus almost as frequent as initial
ones. A closer corpus analysis (Biber et al. 1999: 889) shows that this high pro-
portion of adverbials in fnal position is mainly due to three frequently occurring
forms then, anyway and though , which are commonly placed in fnal position.
10.6.2. Information structure
The high frequency of adverbial connectors at the end of a sentence in the
spoken mode asks for an explanation, not only because this position is only
rarely attested in other Indo-European languages (for sentence-fnal connector
in Japanese and other non-Indo-European languages, see Itani 1992). While
circumstance and with some restrictions stance adverbials are regularly
placed at the end of sentences, English alone allows this fnal position as a
regular one for adverbial connectors.
In terms of information structure, the introduction of a connector at the be-
ginning of a sentence guides the hearers and readers in their interpretation, by
signalling that two sentences or chunks of discourse are linked and, secondly,
by explicitly highlighting the specifc semantic relation. This explicit marking
at an early place of the second connect is of less importance for the pure
additive and transitional connectors, which only emphasize a relation which
would already have been there in the asyndeton. It is much more essential for
the impure connectors marking ccc-relations: these signify that an addi-
tional (causal) or even contradictory aspect (concession/contrast) has to be
considered in the relation of the two propositions.
Adverbial connectors which are placed at the beginning of a sentence function
as explicit signposts, guiding the readers and hearers through the text, because
they facilitate the rapid processing of a passage. For speakers and writers, they
help to make sure that their intentions about the semantic relations of the two
propositions are understood. The initial position can therefore be considered the
unmarked position for adverbial connectors (see also Biber et al. 1999: 889). Final
connectors expressing relations of cause, contrast and concession, on the oth-
er hand, force a re-processing or even reinterpretation of the preceding assertions.
10.6.3. Sentence-fnal connectors in Present Day German
A comparison with other Indo-European languages shows that English alone
regularly employs sentence-fnal adverbial connectors. The Handbuch der
deutschen Konnektoren, which bases its classifcation on topological criteria
Sentencefnal connectors in Present Day English 199
only, does not recognize this position as a standard one for Present Day Ger-
man, but lists a number of connectors which may in certain contexts be
placed in fnal position (Pasch et al. 2003: 505509). In the internet version
of the handbook, which is constantly being revised, the grammis-team later
introduced the new category Nachsatzposition bei Pronominaladverbien (cf.
grammis; revision of 28/10/2004, 16:22). Yet, this fnal position (Nachsatz-
position) of adverbial connectors in Present Day German is restricted to pro-
nominal connectors, in particular to additive auerdem and davon abgesehen,
contrastive/concessive trotzdem and causal darum, deshalb, and deswegen.
The only non-interactive example grammis cites is
(167) Ich muss heute auf dem Markt einkaufen. Bei uns gibts nichts Ge-
scheites. Deswegen.
I have to go shopping at the market today. Here, there isnt anything
appropriate (to buy). Therefore.
In this example, however, deswegen is not a prototypical fnal connector, which
should by defnition signal the link of the second connect to the frst one. It
rather gives the impression of an afterthought. This is evident from the fact
that deswegen semantically belongs to the frst, and not to the second connect
(cf. the Bei uns gibts nichts Gescheites. Deswegen muss ich auf dem Markt
einkaufen or Ich muss heute auf dem Markt einkaufen, weil es bei uns nichts
Gscheites gibt and not Ich muss heute auf dem Markt einkaufen. *Deswegen
gibt es bei uns nichts Gescheites).
All the other examples given on grammis are even less convincing, since
in these the speakers do not employ sentence-fnal connectors, but only use a
connector in elliptical constructions (i. e. instead of the full sentence here given
in brackets):
(168) A: Es wre schade, wenn das Gebude abgerissen wrde.
B: Es ist aber baufllig.
A: Trotzdem!
(= Trotzdem wre es schade, wenn das Gebude abgerissen wrde).
A: It would be a pity if the building would be pulled down.
B: But it is dilapidated.
A: Nonetheless!
(169) A: Morgen muss das Auto zur Inspektion!
B: Und wir mssen uns auch um eine Garage kmmern.
A: Auerdem!
(= Ja, das mssen wir auerdem noch tun!).
200 CONTRAST/CONCESSION
A: Tomorrow, the car will have to be taken in for a service.
B: And we also will have to see about a garage.
A: Moreover!
All in all, Present Day German does not employ fnal adverbial connectors as
a regular pattern.
10.6.4. Sentence-fnal connectors in the history of English
In English, fnal adverbial connectors are also a comparatively recent phenom-
enon: the passages quoted above for however in (164) and (165) are the earliest
occurrences I have found. As has been pointed out above, all of these early
instances are attested in texts which record an oral mode, such as plays or pri-
vate letters. The same restriction to these genres is found with sentence-fnal
too, which like though may not be placed sentence-initially, but only medi-
ally and fnally. The frst instances of sentence-fnal too are also recorded in
speech-based texts, such as court trials and plays.
In view of these early attestations of sentence-fnal connectors in speech-based
texts and the high proportion of sentence-fnal connectors in conversation in
Present Day English, the following sections will examine the uses of one of
the three highly frequent sentence-fnal connectors of Present Day English,
though. Though is a good candidate for such an investigation, not only because
of its high frequency, but, more importantly, because it expresses a contrastive/
concessive relation, i. e. an impure relation which requires a reinterpretation
of the relation of the two connects. It is furthermore interesting because as
has been sketched above though had not been used as an adverbial connector
for many centuries, but had predominantly been employed as a subordinating
conjunction.
10.7. Present Day English sentence-fnal though
10.7.1. Earlier research
In spite of these rather obvious developments in the newer history of though,
these phenomena have, surprisingly, not been examined in detail in the litera-
ture. Sentence-fnal though is mentioned, but not discussed in Halliday and
Hasan (1976: 251), Quirk et al. (1985: 644) and Altenberg (1986: 2324). Inter-
estingly, Quirk et al. and Altenberg give the same example:
Present Day English sentencefnal though 201
(170) (a) (Al)though he is poor, he is satisfed with his situation.
(b) He is poor. He is satisfed with his situation, though.
The fact that the use of sentence-fnal though is illustrated by the construc-
tion with a pre-posed concessive clause suggests that these authors see the
two constructions as functionally identical, i. e. as concessives proper marking
counter-expectancy (i. e. the implication is if someone is poor, he or she is
not satisfed with his or her situation).
This interpretation, however, runs counter to the differences in illocution-
ary weight of hypotactic vs. paratactic construction. As has been pointed out
several times above, the equal illocutionary weight of both connects implies a
contrastive rather than a purely concessive reading. This contrastive function
of sentence-fnal though is acknowledged in the Longman Grammar. In the
discussion of the fndings, however, the sentence-fnal use of though is related
to the complexities of the semantic relation concession:
Though as a linking adverbial is found primarily in conversation, where its frequen-
cy far exceeds that of though as a subordinator A possible factor at work is that
considerable forward planning is required to construct a sentence with a concessive
adverbial clause. The speaker has to have two propositions in mind, together with
a realization that the one runs counter to the other, before starting to speak. This
is not easy to manage in the online production of speech (Biber et al. 1999: 851).
Since this explanation focuses on speech processing in speakers, the Longman
Grammar characterizes instances in the written mode as surprising:
[t]hough as a linking adverbial is also sometimes used in writing that is informal
or intended to resemble speech. Surprisingly [my emphasis, U. L.], though is also
sometimes used as a linking adverbial in more formal written prose (851).
This research report frst of all mirrors ongoing language change: Sentence-f-
nal though was a comparatively rare phenomenon in the 1960s and 1970s when
the Comprehensive Grammar was written (for the numbers, see below Figure
10.2), but became a highly frequent connector by the 1990s (see the counts of
the Longman Grammar), since when it has also been increasingly found in
written, even formal genres.
The only publication focussing solely on sentence-fnal though is an article
by Barth-Weingarten and Couper-Kuhlen (2002), who observe a development
of sentence-fnal though from concessive marker to discourse marker (topic
shifter) and examine the similarities and divergencies of this process to prop-
erties of grammaticalization. They do not discuss, however, if sentence-fnal
though functions as an adverbial connector proper. Characterizing sentence-f-
nal though (and also, e. g. then) as an adverbial connector would imply that we
202 CONTRAST/CONCESSION
are dealing with a comparatively recent syntactic change in English, i. e. a new
position for adverbial connectors, specifc to Present Day English. The prag-
matic/textual function of though as a topic shifter would then not be seen as
its key function, but as only one of the functions of the contrastive/concessive
sentence-fnal adverbial connector though (since such a topic shifter function is
inherently connected to the contrastive/concessive origins of an item and thus
to every concession marker and every concessive construction).
10.7.2. PDG obwohl
The close relationship and interdependencies of concession and contrast
with differences between parataxis and hypotaxis are also evident in the case
of the German concessive subordinator obwohl. In spoken Present Day Ger-
man, obwohl shows ongoing changes which are in some aspects similar to the
changes which have affected the Present Day English concessive subordinator
although.
Obwohl is one of the central subordinators for marking the relation conces-
sion in Present Day German. In accordance with the properties of parataxis
and hypotaxis in German, the main and subordinate clause are distinguished
by differences in word order (verb-second vs. verb-fnal). The subordinate
clause (verb-fnal) introduced by clause-initial obwohl may be placed before
or after the main clause.
(171) (a) Obwohl sie hart gearbeitet hat, ist sie durchgefallen.
(b) Sie ist durchgefallen, obwohl sie hart gearbeitet hat.
She failed, although she had worked hard.
Yet, in colloquial German, clause-initial obwohl is increasingly used with
verb-second, i. e. main clause word order, as in
(172) Sie hat hart gearbeitet. Obwohl sie ist durchgefallen.
The hyphen in this example signals that obwohl is marked prosodically by a
pause separating it from its connect. Examples (173) and (174) illustrate this use
of obwohl by corpus examples collected by Gnthner (2000: 444, 446):
(173) A: ich dt SO GERN mit. echt
A: obwohl ZEHN stunden fug des fnd ich doch net so gut.
A: I would love to come along (really)
A: although a ten hour fight I would not like that.
Present Day English sentencefnal though 203
(174) A: klingel einfach;
A: ich komm dann runter
B: okay
A: obwohl. ich komm doch besser bei dir vorbei
A: des isch glaub ich gschickter.
A: just ring the bell;
A: Ill come down
B: okay
A: although. Id better come to your place
A: thats more convenient I think.
These examples illustrate that obwohl, when used as an adverbial connector
with main clause word order, no longer marks a concessive relation. In these
interactive examples, the equal illocutionary force of the connects does not
indicate a concessive relation on the propositional level, but rather a contrast
on the interpersonal or speech act level. It specifcally signals correction, in
both of these cases self-correction.
10.7.3. PDE but
While PDG obwohl with main clause word order has recently appeared in this
function in spoken, colloquial German, (self-)correction on the interpersonal
level is much more frequently expressed by the coordinator aber.
11
This agrees
with the fndings for Present Day English but (cf. Barth-Weingarten 2003; Al-
tenberg 1986). In her analysis of public spoken discourse in the discourse-func-
tional perspective on concession, Barth-Weingarten (2003) fnds that in about
96 per cent of the cases with an explicit linker it is the coordinator but which
serves as such a connector.
These fndings widely correspond to the proportions found in the only quan-
titative study which has as yet analysed contrastive/concessive linking in both
spoken and written English in any detail (Altenberg 1986). Altenberg analyzes
a sub-corpus of 100,000 words each from the London Lund Corpus of Spo
ken English (LLC) and the Lancaster-Oslo/Bergen Corpus of British English
11 In accordance with other studies in the feld, I here follow the three-level approach
(for conjunctions, see, for example, Sweetser 1990), which distinguish the proposi-
tional level (i. e. the semantic content proper) from the interpersonal level (i. e. the
social and expressive functions of communicative acts and moves) and the textual
level (i. e. the methods of organization which create a coherent discourse).
204 CONTRAST/CONCESSION
(LOB), and offers the fgures for the coordinator but, the adverbial connec-
tors however, anyway, yet, still, nevertheless, though and the subordinators
although, though, while, whereas and even if.
Table 106: but adverbial connectors subordinators (after Altenberg 1986)
This table frst of all shows that the rank order (though not the proportion)
is the same in both corpora. Speakers and writers are guided by much the
same structural priorities when expressing contrastive relations: but is by far
the most frequent connector for the relation of contrast/concession in both
spoken and written English. With 843 in contrast to 85 occurrences, its number
is about ten times higher than that of subordinators. With more than 56 per cent
of all instances, but is, however, also the most frequent contrastive/concessive
connector in the written corpus. In percentages, these fndings yield the num-
bers in Table 10.7.
Table 107: but adverbial connectors subordinators: percentages
(after Altenberg 1986)
Coordinator but Adverbial Connectors Subordinators
Spoken English 78 % 14 % 8 %
Written English 56 % 22 % 22 %
But obviously functions as a potential all-purpose link that may replace other
(lexical and grammatical) alternatives in most contexts, whereas subordin -
ation is as we have seen above only employed under certain circumstances,
triggered by information processing considerations. The spoken material ex-
hibits a strong tendency to prefer coordination over subordination (which only
adds up to 8 per cent of all constructions). These proportions refect tendencies
Present Day English sentencefnal though 205
which facilitate information processing under conditions of real-time planning
in spoken discourse, among them the preference for right-tending construc-
tions (which has also been observed for causal/resultive clauses and complex
noun phrases) and the presentation of two assertions in loosely linked sequence
(Altenberg 1986: 20).
Yet, Altenberg found that these general differences do not suffce to explain
the much greater frequency of but in his spoken material. Much more import-
ant than these general tendencies is the fact that obviously speakers use but
to express a number of contrastive discourse functions that are rare or absent
in writing (Altenberg 1986: 2732). None of these do express the semantic
relation of concession or contrast on the propositional level, but may be
characterized as contrastive moves on the interpersonal and textual level of
discourse. The following table (Table 10.8) provides a summary of the specifc
sub-types of contrastive moves relevant for the analysis of sentence-fnal
though.
Table 108: Sub-types of contrastive moves expressed by but
Interactive yes-but countering (interpersonal level)
used turn-initially to introduce an objection or deviant opinion
But after disarmers (interpersonal level)
face-saving device indicating that the speaker is aware of possible objec-
tions to his opinion: I know but, Im terribly sorry but
Topic resumption after disruption (interpersonal/textual level)
not necessary in a written text
Topic shifting but (textual level)
weak countering force: does not signal objection to the previous speakers
utterance but a deviation from the current discourse topic
Passage (175) illustrates instances of but which signal contrastive moves. The
only instance of but which marks contrast on the propositional level is (175d).
(175)
B if they ^don`t get *t\aught#*
A *^y\es#* (175a) yes-but countering
A but the ^p\oint is#
A that [i @] that if we ^ch\ange the _system#
A to al^l\ow for this#
A that they`re *^st\arting up*
B *but we ^still* don`t propose to !d\o it# (175b) turn-initial countering
A ^I think we *d/\o#*
206 CONTRAST/CONCESSION
A ^this was
B ((*^y/\es#*))
B but ^how do we do our !f\inals ex_amining# (175c) yes-but countering
A ^w\ell#
A I ^think the _f\inalists will have to st/op#
A but the ^others will carry !\on# (175d) contrast (propositional)
B ^\I see#
A but the ^th\ing is# (175e) turn-initial countering
A for *^\instance# .
(LLC, Sample 1.4.714740).
In many instances, interactive yes-but counterings are found together with dis-
armers such as I know (176a, 176b), which serve as face-saving devices indicat-
ing that the speaker is aware of possible objections to his opinion.
(176)
D well she was the only one of the family there who/
D could do it .
B yes I kn\ow# (176a) yesbut countering
B but ^also he [?]/ought to have s\old the whole
B thing#
B for his ^hundred thousand *!p\ounds#*
C *^yes !I* kn\ow# (176b) yes-but countering
C but ^he {s/aw} . th/ere you see#
C a ca^reer for D=an#
C and a ca^reer for *((:M\argaret#))* /
(LLC, Sample 1.13.83484).
(177) then illustrates the use of but as a topic shifter. In an indirect speech act,
speaker C suggests opening another bottle of wine, but speaker B changes the
topic, enquiring from interlocutor A whether she was in the Archaeology or
English Department.
(177)
C *we`ve* switched to ((a)) bottom of another **
C ** bottle now ((haven`t we))
B **(laughs )**
A ^[/\m]#
B but ^you`re archae!\ology#
B I mean you`re ^s\eparate from /English#
(LLC, Sample 1.9, 673679).
In sum, the various functions of but in spoken interaction can be summarized
as follows:
Present Day English sentencefnal though 207
Figure 101: Contrastive functions of PDE but: three-level approach
10.7.4. Although and though in the LLC
10741 Quantitative fndings
For testing the various functions of sentence-fnal though in Present Day English,
I chose to examine Samples 1 to 9 from the London Lund Corpus of Spoken Eng
lish (LLC), i. e. the spontaneous recorded conversations comprising ca. 425,000
words. This corpus was selected because the corpus data for Present Day Eng-
lish and the accounts in research suggested that these conversations taped in the
1960s and 1970s would refect the origins of the now highly frequent use of sen-
tence-fnal though. The analysis was restricted to interactive texts, because only
these have the potential of attesting the diverse functions on the propositional,
interpersonal and textual level. To allow a comparison with the subordinator (al)
though, all instances of although and though have been examined.
Figure 102: Although and though in the LLC (Samples 1 to 9)
but
propositional level
counter-expectancy
expressing a contrast
etc.
interpersonal level
self-correction
other-correction
objection, different
opinion, etc.
textual level
topic management
topic change
discourse management
summarizing, etc.
208 CONTRAST/CONCESSION
This diagram shows that the subordinator (al)though was still much more fre-
quent than the adverbial connector though in these interactive texts of the 1960
and 1970s (in contrast to Present Day English, where adverbial though is found
to be more frequent; see above, Chapter 10.6.1). Adverbial though is, however,
already attested ffty times. Only seven of these occurrences are sentence-medi-
al. In accordance with the fndings for Present Day English, the majority of ad-
verbial though is found sentence-fnally, in altogether forty-three occurrences.
Subordinate clauses are in accordance with the rules of grammar or style
mainly introduced by although when they are pre-posed. More regularly, they
are introduced by either although or though and are post-posed (21 per cent
pre-posed; 59 per cent post-posed). Although this pattern is as has been ar-
gued above more diffcult to process, it is still more frequent in spoken in-
teraction, because of the preference for right-tending constructions, facilitat-
ing information processing under conditions of real-time planning in spoken
discourse. With respect to their functions, we fnd the following distribution:
Table 108: Functions of although/though in the LLC
Total Propositional level
concession
Interpersonal/textual
level
pre-posed although/though 39 39
post-posed although/though 129 104 25
fnal though 43 43
10742 Functions of the subordinator (al)though
This table shows that pre-posed subordinate clauses introduced by (al)though
are exclusively used to express the semantic relation of concession on the
propositional level. The counter-expectation in (178) is that the speaker cannot
put the CSC stuff away.
(178) B I`ll be at ^h/\ome#
B and al^though I`ll be doing CS/C _stuff#
B and ^that kind of th=ing#
B I can always put it on one *s/ide#* (LLC, Sample 1.1, 116118)
In (179), a correlative construction with adverbial nonetheless in the second
connect, B expresses her embarrassment that the term ends without her being
involved in any exams.
Present Day English sentencefnal though 209
(179) B ^I`ve been !{v\ery} emb\arrassed ((by)) the f/act#/
B that ^though I`m not . ex:\/amining#/
B ^none the l=ess# . /
B ^teaching seems to :come to an /end# . (LLC, Sample 1.4, 629633).
The semantic relation of concession in its propositional sense is even more fre-
quently (104 times) expressed by post-posed subordinate clauses. In (180), speak-
er B would expect her children to go and see her mother, but they do not do so.
(180) B children never go* to see my m/other#/
B al^though they`re !only on the :other side of/
B B\/irmingham#+
D ^y/eah# (LLC, Sample 1.12, 580590).
In (181), speaker A had been doubtful of whether her offer would be accepted.
(181) A ^=and [@m]#
A ^I !\asked her if {she would ^h\ave him#}#
A ^though we`re now !s\everal miles *aw/ay#*
C ^[/m]#* (LLC, Sample 1.8, 900930).
In twenty-fve of the 129 instances, however, post-posed subordinate clauses
introduced by although/though do not work on the propositional, but on the
interpersonal or textual level in exactly the sub-types that have been found for
the coordinator but. Example (182) is a prototypical example of self-correction
(stressed by in fact), when B acts as if she suddenly remembered that she had
got her PhD the week before.
(182) B hel^l\o Mr B/aines#
B ^this is _Mrs !\Edgton#
C ^\oh#
C (good ^after!n\oon#))
B [@:m] . ^((though)) in fact I`m :Doctor Edgton
B n/ow#
B so ^you ( laughs) you can . ob:serve my new
B !st\atus# .
B !got my :Ph:D last !w\eek# (LLC, Sample 8.3b, 270300).
Similarly, speaker B corrects her statement on the price of photo-copying by
excluding a certain shop which is obviously still very expensive in (183).
210 CONTRAST/CONCESSION
(183) A xerography is not terribly expensive after/
A all -
B ^n/\o#
B ^no it /\isn`t# -
B ^although Rank :X\erox
B are ^doing their !best to make :sure that it :\/is#/
(LLC, Sample 2.1, 938942).
In the next example (184), the post-posed subordinate clause has the potential
of functioning as a topic shifter on the textual level. This is evident from the
fact that speaker A again corrects herself, saying that she does not want to pur-
sue this topic (i. e. does not want to gossip about the dud ones).
(184) A ^and there is !H\erman#
A who is ^\also known#
A ^in that :frst b\atch [@:]#
A who ^came to us *:f\irst#*
B *^[=m]#*
A ^these are the !tw\o#
A al^though there were !three [@] !d\ud ones#
A [@:] that I ^don`t think at :this stage [@:m]
A I want to [@:] to :sp\/eak about#
(LLC, Sample 2.6, 530534).
In sum, these selected examples from the LLC illustrate that post-posed subor-
dinate clauses introduced by though or although are found in all the functions
which have been established for the coordinator but. These adverbial clauses
may not only work on the propositional level, but also have the potential to
signal contrastive moves, such as self-correction on the personal level, or may
function as topic shifters on the interpersonal level.
10743 Functions of sentencefnal though
With respect to the functions of sentence-fnal though, I have found only one
example which might allow an interpretation on the propositional level.
(185) A I ^l\ike chests# - -
B ^y\eah#
B they`re ^\useless bits of furniture#
B ^r\/eally though#
Present Day English sentencefnal though 211
B be^cause they :only go _sort of of !up to
B slightly below :t\able *h/eight#*
(LLC, Sample 4.2, 300340).
While a purely concessive meaning may be possible (implication: one should
not like useless bits of furniture), an interpretation as a contrastive move or
self-correction is much more likely, because the speaker stresses her corrective
refections by the epistemic really. All of the other instances of sentence-fnal
though only allow an interpersonal or textual reading.
(186) A *we`re going* to the Lake District for Christmas
C ^y\es#
C ^you !s\aid#
C ^that should be n/ice#
B [m] -
C ^I !love D/\orset though#
C it`s ^so b/\eautiful# (LLC, Sample 2.7, 110170)
These analyses are supported by the fact that in nineteen of the forty-three
instances though is followed by a question tag with falling intonation (sym-
bol ^ in the LLC). In negative question tags following a positive statement,
speakers signal by falling intonation that they assume that the person spoken
to will agree (Quirk et al. 1985: 10.57). The whole phrase thus functions as a
self-correction and a foor-holder.
(187) C *there* ^might have been m\ore varied#
A ^y\es# -
C it`s ^quite n\ice {to ^l\ook at as [@:]#}#
C ^just a :series of :p\/ictures though#
C ^\isn`t it#
A ^oh v\ery#
A ^y\es#
A and I ^do th\ink#
A that ^these three p\ortraits# (LLC; Sample 1.8, 582590).
(188) C so I ^thought ^I`d ^I`d rather p/ay my eighty
C p\ence#
C than ^go through :th\at# -
C ^it`s a !!lot of :m\oney though#
C ^\isn`t it# -
212 CONTRAST/CONCESSION
B 20 [m]
C so I ^think (LLC, Sample 2.7, 914920).
Instead of tag-clauses, we also fnd collocations of sentence-fnal though with
you know. These may be interpreted as co-texts complementary to disarmers
such as I know which are frequent in collocations with contrastive, interper-
sonal but (see above, example (176)).
(189) B and ^that really *((3 sylls))*
A *but if you`re* typing it up now [@:] **((why can`t/
A yes**
B **it`s ^going so sl\owly** though

you _know#
B it`s these awful these awful s/ymbols# .
(LLC, Sample 2.1, 160200).
In sum, we see that sentence-fnal though is predominantly employed as a typi-
cal adverbial connector: it marks the relation of contrast, in particular on the
interpersonal level in contrastive moves, such as self-correction. In its origins,
contrast here is clearly emerging from the idea of internal concessivity, i. e.
counter-expectancy on the level of the communication process in spite of the
state of the argument or as against what the current state of the communica-
tion process would lead us to expect. Since this relation of contrast can also
be marked by the coordinator but, or by post-posed subordinate clauses intro-
duced by (al)though, it is an inherent feature of contrastive/concessive connec-
tors in general and is thus not specifc to sentence-fnal though. This means
that it is certainly not correct to essentially characterize sentence-fnal though
as a discourse marker.
12
The current re-emergence of though is more likely
to be seen as a re-analysis of the subordinator though. However, since though
has been attested as an adverb in certain contexts (see above, examples (149)
to (152)), it is also problematic to regard its development as a counter-example
for unidirectionality (i. e. from the more grammatical conjunction to the less
grammatical adverbial connector).
The sentence-fnal uses of though (and also then) should rather be seen par-
allel to developments of Present Day German obwohl (and weil) with verb-sec-
ond, i. e. main clause, word order. The establishments of a regular, new position
for adverbial connectors in Present Day English may be seen as the uninten-
tional result of a certain rhetorical strategy on the part of the speakers (see, for
12 Barth-Weingarten and Couper-Kuhlen (2002) acknowledge the function of sen-
tence-fnal though as a concessive marker, but do not further comment on it.
Present Day English sentencefnal though 213
a similar account on the grammaticalization of adverb markers, Detges 1998:
7), who use a subordinator with illocutionarily heavier main clause word
order. By doing so, speakers highlight the relation contrast in the specifc
co- and contexts of contrastive moves, mainly self-correction. PDG obwohl is
used as an adversative adverbial connector (instead of as a subordinator) and
thus provides an alternative to the functionally rather overloaded PDG aber. A
similar path is evident in PDE though. Since English, however, does no longer
have any distinctions in the constituent order, which would differentiate pre-
dominantly concessive hypotaxis from contrastive parataxis (and thus the
relations concession and contrast), it chooses a structural alternative and
places the connector in a position which clearly differentiates hypotaxis from
parataxis: in contrast to the conjunctions which are placed sentence-initially,
the adverbial connector is placed sentence-fnally. This starts in spoken, inter-
active texts on the interpersonal level, but is increasingly found in written texts,
even in rather formal registers. The fnal placement of adverbial connectors
(PDE contrastive however, though, causal/inferential then, additive too) is thus
a change in constituent order, triggered by the lack of other distinctive means
(such as verb-second vs. verb-fnal word order): in the past centuries and, in
particular, the last decades, English has thus established a new slot for the
placement of adverbial connectors, the sentence-fnal position.
11. addition
11.1. General tendencies
In most of the grammars and metalinguistic accounts on sentence relation past
and present, adverbial connectors and conjunctions have been categorized in
groups defned by the semantic relations they mark (see Lockes criticism cited
at the beginning of this study). Traditionally, the different semantic relations
addition, cause, concession/contrast, and transition (see above, Chap-
ter 3.3) are not divided further, but are described as if they were on equal
levels. Only the authors of the Cambridge Grammar have recently proposed a
distinction between pure and impure connectives (Huddleston and Pullum
2002: 775779). While pure connectives have no other function than that of
connecting their clause to the respective surrounding text, impure connec-
tives, in addition to their linking function, also express a ccc-relation, which
adds another (cause/result) or even conficting (concession) aspect to the
relation of the two propositions connected.
The long-term developments in the feld of adverbial connectors clearly illus-
trate these differences in rank, since they mirror the hierarchy of the increas-
ing complexity of the semantic relations addition, cause and concession/
contrast. In the developments of adverbial connectors for the ccc-relations,
we have seen recurrent patterns triggered by changes in the typology of Eng-
lish, such as the loss of polyfunctionality in the differentiation of subordinators
from adverbial connectors, or the replacement of pronominal by temporal or
spatial deictics. Because of the differences in information structure separating
paratactic from hypotactic structures, the semantic relation of pure addition
is never expressed by subordination. This would contradict the very idea of the
pure addition of an equally important, further proposition in the second con-
nect. Issues such as problems of differentiating subordinators from adverbial
connectors and the respective paths the Old English polyfunctional items take
(cf. in particular OE causal/resultive form and OE concessive eah) do thus
not play any role in the diachrony of additive adverbial connectors.
11.2. The diachrony of reinforcing adverbial connectors
11.2.1. General tendencies
This relation of pure addition, inherently present as the unmarked case in
the asyndeton, may be highlighted by the coordinator and, which has at all
The diachrony of reinforcing adverbial connectors 215
times been by far the most common marker of the additive relation, not only
on the phrasal level, but also on the level of the sentence or discourse (see
Chafe 1982; Culpeper and Kyt 2000 and the summary of corpus fndings in
Appendix B.1.5; for mode and genre-differences, see below, 11.7). This addi-
tive force of and can be stressed and highlighted, i. e. reinforced, by its use
in collocations such as and moreover or and furthermore (for the relevance of
these collocations, see below, Chapter 13). These so-called reinforcing con-
nectors can, however, also be used on their own.
Table 11.1 lists the development in the sub-group of these reinforcing con-
nectors (excerpted from Appendix A.1 and Appendices B.1.2B.1.4).
11.2.2. Also
OE eall-swa, the only Old English additive connector comprising an explicitly
deictic element (OE swa so), is also the only additive adverbial connector which
has survived from Old English to Modern English (see DOE, s. v. eall-swa). In
all periods, however, it is often hard to decide whether eall-swa/also expresses a
reinforcing (furthermore) or an equative (similarly) additive relation.
3
(190) Eallswa eac nygontyne gear gefylla one circul e uwitan hata luna
rem (ByRM 2.3.116).
Furthermore/Similarly, nineteen years also (= eac) make up the cycle
that scholars call lunar.
(191) And also, certes, if I governed me by thy conseil, it sholde seme that
I hadde yeve to thee over me the maistrie, and God forbede that it so
weere! And also if I wolde werke by thy conseillyng, certes, my con-
seil (CMCTPROS, p.220, C.1).
And furthermore/similarly, in fact, if I behaved according to your ad-
vice, it would seem as if I had given you control over me. God forbid if
it were so. And furthermore/similarly, if I were to act according to
your advice, in fact, my advice .
The history of also in particular also refects deliberate choices of certain
authors or stylistic preferences of certain periods: also, which had been fre-
quently used in Old English (OE eall-swa) and Middle English, is only rarely
attested from the Late Modern English period onwards (see Appendix B.1.2).
3 For the development of OE eall-swa wholly so, quite so, just so which has yield-
ed the adverb also and the conjunction as, see the OED, s. v. as.
216 ADDITION
T
a
b
l
e

1
1

:

R
e
i
n
f
o
r
c
i
n
g

a
d
v
e
r
b
i
a
l

c
o
n
n
e
c
t
o
r
s

i
n

t
h
e

h
i
s
t
o
r
y

o
f

E
n
g
l
i
s
h

Old
English
11001350
(ME 1/2)
13501420
(ME 3)
14201500
(ME 4)
15001570
(EModE 1)
15701640
(EModE 2)
16401710
(EModE 3)
17101780
(LModE 1)
17801850
(LModE 2)
18501920
(LModE 3)
1920
(PDE)

a

g
i
e
t
e
f
t
e
f
t
[
e
f
t
s
o
n
a
/
-
e
s
]
e
f
t
s
o
n
a
/
-
e
s
e
a
c

s
w
y
l
c
e
e
a
c

s
w
l
y
c
e

r
t
o
e
a
c
a
n
[
t
o
e
k
e
n
]
e
a
c
e
e
k
e
e
k
e
l
l
e
s
e
l
l
e
s
e
l
l
e
s
e
l
s
e
e
l
s
e
e
l
s
e
e
l
s
e
e
l
s
e
e
l
s
e
e
l
s
e
e
l
s
e
e
a
l
l
-
s
w
a
a
l
s
w
a
a
l
s
o
a
l
s
o
a
l
s
o
a
l
s
o
a
l
s
o
[
a
l
s
o
]
a
l
s
o
a
l
s
o
a
l
s
o
[
m
o
r
e
]
[
m
o
r
e
]
[
f
u
r
t
h
e
r
]
f
u
r
t
h
e
r
[
f
a
r
t
h
e
r
]
f
u
r
t
h
e
r
[
f
u
r
t
h
e
r
]
[
f
u
r
t
h
e
r
]
[
f
u
r
t
h
e
r
]
f
u
r
t
h
e
r
f
u
r
t
h
e
r
f
e
r
t
h
e
r
m
o
r
e
f
u
r
t
h
e
r
m
o
r
e
f
u
r
t
h
e
r
m
o
r
e
f
u
r
t
h
e
r
m
o
r
e
f
u
r
t
h
e
r
m
o
r
e
f
u
r
t
h
e
r
m
o
r
e
[
f
u
r
t
h
e
r
-
m
o
r
e
]
[
f
u
r
t
h
e
r
-
m
o
r
e
]
[
f
u
r
t
h
e
r
-
m
o
r
e
]
f
u
r
t
h
e
r
m
o
r
e
[
f
a
r
t
h
e
r
o
v
e
r
]
,

f
u
r
t
h
e
r
o
v
e
r
[
f
u
r
t
h
e
r
o
v
e
r
]
o
v
e
r
m
o
r
e
o
v
e
r
m
o
r
e
o
v
e
r

t
h
a
t
o
v
e
r

t
h
a
t
o
v
e
r

t
h
a
t
[
o
v
e
r

t
h
i
s
]
o
v
e
r

t
h
i
s
o
v
e
r

t
h
i
s
m
o
r
e
o
v
e
r
m
o
r
e
o
v
e
r
m
o
r
e
o
v
e
r
m
o
r
e
o
v
e
r
m
o
r
e
o
v
e
r
m
o
r
e
o
v
e
r
m
o
r
e
o
v
e
r
m
o
r
e
o
v
e
r
m
o
r
e
o
v
e
r
i
t
e
m
i
t
e
m
i
t
e
m
a
f
t
e
r
2
a
g
a
i
n
a
g
a
i
n
a
g
a
i
n
a
g
a
i
n
a
g
a
i
n
a
g
a
i
n
a
g
a
i
n
[
o
v
e
r

a
n
d

b
e
s
i
d
e
s
]
o
v
e
r

a
n
d

b
e
s
i
d
e
s
a
b
o
v
e

a
l
l
a
b
o
v
e

a
l
l
a
b
o
v
e

a
l
l
a
b
o
v
e

a
l
l
a
b
o
v
e

a
l
l
a
b
o
v
e

a
l
l
b
e
s
i
d
e
b
e
s
i
d
e
s
b
e
s
i
d
e
s
b
e
s
i
d
e
s
b
e
s
i
d
e
s
b
e
s
i
d
e
s
b
e
s
i
d
e
s
o
v
e
r

a
n
d

a
b
o
v
e
o
v
e
r

a
n
d

a
b
o
v
e
t
o
o
t
o
o
t
o
o
t
o
o
[
p
l
u
s
]
2

T
h
i
s

a
d
d
i
t
i
v
e
,

r
e
i
n
f
o
r
c
i
n
g

u
s
e

o
f

a
f
t
e
r

m
o
r
e
o
v
e
r

i
s

e
x
c
l
u
s
i
v
e
l
y

f
o
u
n
d

i
n

C
M
V
I
C
E
S
4
.
The diachrony of reinforcing adverbial connectors 217
In the full corpus of texts analysed for the present study, it is, for example, not
attested at all as an additive adverbial connector in the Late Modern English
period (LModE 1 to LModE3).
3
The low esteem of also as an adverbial connector in Early and Late Modern
English is also corroborated by the evidence of metalinguistic texts. In the six
bilingual dictionaries conveniently accessible in the Early Modern English Dic
tionaries Database (EMEDD; ed. Ian Lancashire),
4
furthermore, moreover and
besides emerge as the central additive connectors of Early Modern English.
5
Table 112: Reinforcing adverbial connectors in Early Modern English dictionaries
Th. Thomas Latin English 1587
Etiam Also, more over, yea; yea further, yea; why
Insuper moreouer, ouer and beside, furthermore .
Itemque And moreouer, and also
Praeterea Furthermore, more ouer, beside, else
Proporro Further more, moreover
Quinetiam Moreouer, be side that, yea that more is
Florio Italian English 1598
Alsi also, moreover, eftsoone
Altresi also, eke, moreouer, ouer and besides, likewise,
Dauantaggio moreouer, besides, more
Et, a coniunction and, moreouer, besides
Oltraci, Oltradici moreouer, besides that, furthermore
Minsheu Spanish English 1599
deMs more, besides this, moreover, ouer and aboue, that which
remaineth
3 Interestingly, the OED also asserts that also was [n]ot common in 16th c., refer-
ring to a frequency count of Shakespeare, who has it only 22 times (OED, s. v.
also, Etymology).
4 John Palsgrave (1530; English-French), William Thomas (1550; Italian-English),
Thomas Thomas (1587; Latin-English), John Florio (1598; Italian-English), John
Minsheu (1599; Spanish-English), and Randle Cotgrave (1611; French-English).
5 It is still not very frequent as an adverbial connector in the Present Day English
sub-corpora analysed for the present study (two instances of also, seven of further
more, eight of again, and ffteen of sentence-medial too).
218 ADDITION
Cotgrave French English 1611
Dabondant ouer and besides, moreouer
Arriere also, moreouer, furthermore, besides, ouer and besides
Davantage Moreouer, further, furthermore, else, besides, ouer and be-
sides, aboue and besides
De creu Ouer and besides, moreouer
Outreplus Furthermore, moreouer, besides that
Dadvantage Moreouer; furthermore; ouer-and-besides
Puis Then, afer [sic], moreouer, furthermore
It is also interesting to note that in most of the dictionaries full pronominal be
sides this/that is preferred over the simple, lexicalized beside(s). The following
passage from a letter of Cuthbert Tunstall to King Henry VIII illustrates the
uses and also the effects of these longer, pronominal connectors besides that
and the lexicalized over this.
(192) And iff your Grace shuld accepte the said Election therby ye must con-
fesse your realme to be under subjection off thempire to the perpetual
prejudice off your successor, or ells the said Election wer voyde as made
off a person not eligible. Besids that the forme off the Election con-
tenyth that frst he must be Kinge of Romains and the coronation at
Rome makith hym have the name off the Emperor, wher befor he is
callyd but Kinge off Romains. Over thys yff themperor which nou is
remain stil Kyng off Romains as I understond he entendeth to doo, then
yff your Grace wer eligible and undir thempire, yet ye coud not be cho-
sen Emperor, by cause ye were never Kinge of Romains. And also he
remanyng ye could not be chosen Kynge off Romains, bycause the King-
dome is not voyde, and noon can be chosen therto but when it is voyd
edyr by dethe or ellys when the Kinge off Romains is crownyd Emperor,
wherby undir hym may be chosen a Kinge off Romains. (CEPRIV1,
Cuthbert Tunstall to King Henry VIII, P I, 136137).
The chain of arguments begins with the simple coordinator and. The fol-
lowing arguments are then highlighted by pronominal besides that and over
this. In the last instance, we do not fnd simple also, but the collocation and
also. This illustrates that additive (and also transitional) connectors mainly
serve rhetorical functions: they are employed to explicitly highlight a re-
lation which is already intrinsically present in the linear sequence of the
propositions.
The diachrony of reinforcing adverbial connectors 219
11.2.3. New coinages: item, plus, too
The large majority of all newcomers in the feld of adverbial connectors are
space deictics in origin. The only exceptions are the Latin loans item (ME3)
and plus (PDE), which are, however, confned to certain genres or modes
(item to documents and laws; plus is labelled colloquial by the OED; see
above, 7.1).
The most successful non-spatial reinforcing connector is too, which is re-
stricted to sentence-medial or sentence-fnal position. Sentence-fnal and
sentence-medial too are frst attested at the time when we also fnd the frst
instances of other sentence-fnal connectors (LModE1; on however, see above,
Chapter 10.5.3.3) and when there is generally a sharp increase of medial con-
nectors (see Chapter 13). The emergence of too is thus not specifc to the addi-
tive relations, but rather to be explained by these general topological changes
affecting all adverbial connectors (for examples, see below, Chapter 13).
11.2.4. Iconic principles
Most of the new coinages in the feld of additive connectors make use of the
source domains time and, in particular, space, even more naturally than the
adverbial connectors for the other semantic relations. In the case of additive
connectors, the transfer from the temporal or spatial sequence to the linear
textual and therefore argumentative sequence is straightforward and does not
need any additional construal in contrast to, for example, post hoc, propter
hoc interpretations for cause or the conventionalisation of the conversational
implicatures of simultaneity or concomitance for concession. Since most of
the relevant patterns of transfer from the domains time and space have been
described in Chapters 8.3 and 8.4 above, the following sections will only sum-
marize the most important fndings for the semantic relation addition.
The most regular pattern of newcomers in the feld of reinforcing adverbial
connectors are polymorphemic items coined from the source domain space.
By their polysyllabic structure and their resulting extraordinary length (in
comparison to other etymologically Germanic grammatical elements in Eng-
lish), their reinforcing function is signalled iconically. It has been pointed out
above that the additive rtoeacan thereto-also is among the longest con-
nectors in Old English. Similarly, the new coinages from Middle English on-
wards are typically compounds, so, for example, the double (often pleonastic)
comparatives, such as furthermore, furtherover, overmore or moreover (for
their morphology, see above 7.3.4). Even heavier phrases are attested from the
220 ADDITION
Early Modern English period: most frequent among them are over and besides
(which flls two dimensions of the spatial extension) and the again pleonas-
tic over and above.
The pleonastic expressions in particular illustrate that these polymorphemic
items are not coined for lack of other processes of word formation, but are
formed and used intentionally in order to give a greater weight to the second
connect, or to highlight that there is a relation of addition. It is important
to distinguish these compound connectors and phrases from polymorphemic
complex connectors, such as Old English contrastive/concessive nayls and
swaeahhwere. These complex concessive connectors are typical examples
of the iconic principle that formal complexity corresponds to conceptual com-
plexity (i. e. the complex relation concession signalling counter-expectancy;
see also Kortmann 1997: 113136). The pure semantic relation addition
is certainly not a complex concept and is accordingly not expressed by mor-
phologically complex connectors which require a certain amount of process-
ing. The compounds OE rtoeacan and ME overmore, and also the phrase
EModE over and besides are fairly straightforward as far as their morphologi-
cal make-up is concerned: their polymorphemic structure and length, however,
serve to iconically underline their reinforcing function.
This attempt to create iconic additive connectors may also be one of the
reasons why also comes to be predominantly employed as a circumstance ad-
verbial on the level of the phrase. It does not highlight the additive relation
expressively enough. In (193a), for instance, also is supplemented by the full
sentence it is further sayd. Without the punctuation, it could also be interpreted
as working on the level of the phrase, modifying the verb say; ouer and besides
(193b), on the other hand, is unambiguous and used on its own.
(193) Also, it is further sayd, [193a] that a small part of your potentiall Cau-
teryes, doth and will worke as forcibly on a soft and tender bodye, as
a great quantity thereof will doe vpon a stronge and grosse obdurate
person. Ouer and besides, [193b] the greater abcessions are to bee Cau-
terized one way, and the lesser an other way, and that with good consid-
eration (CESCIE2A, p.30).
11.3. Equative connectors
The different, rhetorical character of additive connectors is even more obvious
in the history of the so-called equative adverbial connectors, which do not
reinforce or highlight the proposition of the second connect, but place it on an
Equative connectors 221
equal level. There have always only been very few of these connectors in the
English language and, more importantly, all of them are formed from still fairly
transparent phrases (OE eall-swa
6
, eac swlyce also such, EModE like(wise),
LModE in the same way) or are derived from adjectives (semblably, corres
pondingly, similarly), a pattern which is only rarely used for the formation of
adverbial connectors of the other semantic relations (see above, Chapter 7.3.4).
New coinages are
EModE1 likewise, likeways, semblably
EModE3 videlicet (on the very few occurrences and disputable status of the
Latin loan videlicet, see above, Chapter 7.1)
LModE3 in the same way
PDE correspondingly, similarly
Their different status is also refected in the fact that no such adverbial con-
nectors were coined in the Middle English period, which had been the most
productive period for all kinds of other connectives (see above, Chapter 6.6).
11.4. Summative connectors
In their predominantly lexical character, these equative connectors are similar
to the sub-type of summative ones, which mark the end of an argument or
text. They commonly summarize a large chunk of preceding discourse and
thus have to be very explicit. Appendix B.2 shows that these are even rarer and
exhibit a still lower degree of lexicalization. The only candidates which show a
certain amount of lexicalization are the calque in fne (on French fn) and the
phrase in sum, which requires a transfer of a mathematical concept to the line
of argument in the text. Often, adverbials which are classifed as summative are
also ambiguous as to their status of adverbial connectors or style disjuncts, i. e.
they may be seen not to link chunks of discourse, but to comment on the style
of the text (in short). In order to emphasize that a following passage is going to
summarize a whole chunk of discourse, authors much more frequently employ
full nominal or verbal phrases:
(194) (194a) The sum is, men have various appetites, passions, and particular
affections, quite distinct both from self-love and from benevolence: all
of these have a tendency to promote both public and private good, and
may be considered as respecting others and ourselves equally and in
common (CLSERM1A, Butler).
6 On the polysemy of eall-swa, see above, Chapter 11.2.2, examples (190) and (191).
222 ADDITION
(194b) The sum is, the true reason why any man is an Atheist is because
he is a wicked man (CESERM3A, Tillotson, p.II,: ii, 421).
(194c) and therefore to conclude this point, I will say no more,
(CEEDUC2B; Bacon, Sample 1, p.23rv).
11.5. Appositive connectors
The high amount of lexical transparency and the consequently low degree of
lexicalization is also very evident in all the other sub-groups of additive ad-
verbial connectors. In the class of the so-called appositive connectors, such
as in other words, to wit, for instance or for example, all of them apart from
the archaic to wit namely show a very low degree of lexicalization and are
fully re-construable (the only property of lexicalization they fulfl is that they
commonly do not admit a pre-modifying adjective, but are uninterrupted in
their sequence; i. e.
?
for another instance). These items have not been included
in the quantitative analyses of the present study precisely because they are not
(yet) fully lexicalized and also because they mainly work on the phrasal level,
modifying a sentence constituent rather than connecting sentences or chunks
of discourse. They commonly do not signal a two-place relation (see above,
Chapter 2.2).
(195) And these the Horse standing in a true proportion, you shall fasten to
the foure straps of leather; to wit one of them to his neere forelegs, and
his nere hinder leg, and the other to his farre fore leg, and his far hinder
leg; which is cald amongst horsemen trauelling (CEHAND2B, p.72).
11.6. Enumerative (listing) connectors
The most peripheral category of adverbial connectors are those of the enumer-
ative/listing sub-type of the additive relation, i. e. numerals such as secondly,
thirdly, fourthly etc. They alone form an open class and new items can always
be added to the inventory of this sub-type. The only members of this class
showing a certain amount of lexicalization are those denoting the frst and the
last in a list, i. e. frst(ly), last(ly), fnally and the phrases at (the) last. At the last,
for example, is recorded as alast, i. e. in univerbated forms showing phonetic
attrition in Middle English (see MED, s. v. a-last). Furthermore, they may also
show some degree of semantic bleaching, because frst(ly) need not necessarily
be followed by second(ly).
Enumerative (listing) connectors 223
These enumerative adverbial connectors illustrate that their use is triggered
by rhetorical or stylistic considerations. In his sermons, Jonathan Swift uses
lists of enumerative connectors, even on various levels in one and the same
passage: level 1 (196a, 197a), level 2 (196b, 197b), level 3 (196c).
(196) First (196a), I shall produce several instances to show the great neglect
of preaching now among us. Secondly, I shall reckon up some of the
usual quarrels men have against preaching. Thirdly, I shall get forth the
great evil of this neglect and contempt of preaching, and discover the
real causes whence it proceedeth. Lastly, I shall offer some remedies
against this great and spreading evil. First (196b), I shall produce certain
instances to show the great neglect of preaching now among us. These
may be reduced under two heads. First (196c), mens absence from the
service of the church; and secondly, their misbehaviour when they are
here. The frst instance (196d) of mens neglect is in their frequent ab-
sence from the church (Swift, On Sleeping in Church; CLSERM1B).
(197) First (197a), I shall produce certain points wherein the wisdom and vir-
tue of all unrevealed philosophy in general fell short and was very im-
perfect. Secondly, I shall show, in several instances, where some of the
most renowned philosophers have been grossly defective in their lessons
of morality. Thirdly, I shall prove the perfection of Christian wisdom
from the proper characters and marks of it. Lastly, I shall show that the
great examples of wisdom and virtue among the heathen wise men were
produced by personal merit, and not infuenced by the doctrine of any
sect; whereas, in Christianity, it is quite the contrary. First (197b), I shall
produce certain points wherein the wisdom and virtue of all unrevealed
philosophy in general fell short and was very imperfect . (Swift, On
the Wisdom of this World; CLSERM1B).
Since this strategy is used in an almost identical way in both (196) and (197), it
is certainly not only appropriate for sermons entitled On Sleeping in Church
(196). It is, also for other authors of that time (see Appendix 1.1), rather a stylis-
tic preference of this period and often further highlighted by phrases compris-
ing a spatial element, such as place (see Appendix B.1.1):
(198) (198a) I proceed, therefore, in the third place, to show the perfection of
Christian wisdom from above (CLSERM1B; Swift, On the Wisdom of
this World).
(198b) Let me now, in the last place, offer some remedies against this
great evil (CLSERM1B; Swift, On Sleeping in Church).
224 ADDITION
It is such a clear marker of argumentative texts and their argumentative struc-
ture that Swift a great language critic also employs it ironically in his A
Modest Proposal:
(199) For frst, as I have already observed, it would greatly lessen the num-
ber of papists, with whom we are yearly overrun, being the principal
breeders of the nation as well as our most dangerous enemies Sec
ondly, The poorer tenants will have something valuable of their own,
Thirdly, Whereas the maintenance of an hundred thousand children
Fourthly, The constant breeders, Fifthly, This food would likewise
bring great custom to taverns Sixthly, This would be a great induce-
ment to marriage, (Swift, A Modest Proposal; 1729: 2127).
(200) First, as things now stand, how they will be able to fnd food and rai-
ment for an hundred thousand useless mouths and backs. And secondly,
there being a round million of creatures in human fgure throughout
this kingdom, whose whole subsistence put into a common stock would
leave them in debt two millions of pounds sterling, adding those who
are beggars by profession to the bulk of farmers, cottagers, and laborers,
with their wives and children who are beggars in effect (Swift, A Modest
Proposal; 1729: 32).
11.7. Genre-dependency
These examples show that the use of additive connectors in particular is highly
genre-specifc. The corpus fndings of the Longman Grammar corroborate this
view: additive linking adverbials are much more frequent in academic prose
than in all the other registers. This is related to the deliberate decision of the
authors of this register to mark the links as explicitly as possible, thus guiding
the reader through the texts by as many in particular dispensable signposts
as possible (Biber et al 1999: 880, Table 10.25).
Corpus studies of Present Day English have also shown that the employment
of additive connectors is very much medium-dependent, since speech-based
registers generally prefer the simple coordinator and. In a small corpus analy-
sis of Present Day English, Chafe (1982: 39) fnds that and is employed about
four times more often in the spoken than in the written medium (44.2 vs. 10.1
tokens per 1,000 words). These correlations of different modes and genres and
the employment of the coordinator and are not as clear-cut in the earlier peri-
ods of English, however. Culpeper and Kyts (2000) study on the frequency
Genredependency 225
of and in four different registers, for instance, yielded a rather vague pattern
(tokens per 1,000 words): there was no signifcant difference between the texts
chosen to exemplify the written mode science (17 tokens) and history (20.2
tokens) and those exemplifying the spoken mode drama (15.5 tokens) and
trials (20.5 tokens). We can only surmise that the low frequency of and in
the trial and drama texts is probably due to the fact that these are not oral, but
only speech-based texts, and that many of the ands were not jotted down when
transferring the spoken version to the written medium. This example again
shows that it is hard to carry out data-driven, quantitative studies on issues like
the present one.
In spite of these problems, we may still gain some information from a more
detailed analysis of the number of different adverbial connectors used in se-
lected texts of the various sub-periods, at least as far as the semantic relation
addition is concerned. The graphs in Figure 11.1 are based on a count of all
tokens of adverbial connectors in my corpus of Treatises and Homilies (see
Appendix C.2.1). Figure 11.1 gives the absolute numbers of tokens for the re-
spective semantic relations per 20,000 words.
Figure 111: Adverbial connectors per 20,000 words (corpus texts C.2.1)
The unsteady, even erratic graph for the causal connectors mainly refects
metho dological problems concerning the status of for, which had to be counted
as an adverbial connector marking the relation cause until sub-period ME3, but
was then no longer included because it as has been shown above came to be
a rather loose connective, functionally equivalent to and (or even but) from the
last Middle English period onwards (see above, Chapter 9.5.5). This, however,
226 ADDITION
means that neither the numbers for Middle English nor those for Early and Late
Modern English are statistically robust, since for, of course, still expresses the
relation of cause in some contexts. Trying to differentiate these uses would,
however, have also implied a very high degree of subjective interpretation. All
kinds of quantitative approaches which have been tested have unfortunately
failed to yield solid, explicable results (and were thus not used in the respective
chapters).
This is different for the relations concession/contrast, and, in particular,
for the semantic relation addition. The graph for concessives/contrastives in
Figure 11.1 refects the dominance of the contrastive coordinator but and the
contrastive/concessive constructions comprising although (simple although,
and correlative although yet) by showing that there was not very much
change in the absolute number of contrastive/concessive connectors until the
Late Modern English period, when sentence-medial however took over many
of the functions which had until then been expressed by the coordinator but
(see below, Chapter 13).
The graph representing the number of the additive adverbial connectors in
Figure 11.1 shows that there was a steady increase of additive adverbial con-
nectors until the Early Modern English period, followed by a slight decrease in
Late Modern English. Early Modern English thus emerges as the period when
additive adverbial connectors were employed most frequently. Since corpus
studies on Present Day English have shown that this kind of adverbial connec-
tors is particularly frequent in academic prose, the increase in additive con-
nectors in Early Modern English is most probably connected to the fact that the
vernacular was increasingly used in academic registers, which had before been
recorded in Latin and French. It also mirrors the tendency found for academic
prose in Present Day English, namely that authors writing texts of this register,
for which English was emancipated in the Early Modern English period, delib-
erately decide to mark the links as explicitly as possible. This again illustrates
that the changes in adverbial connectors marking the relation addition are
mainly triggered by rhetorical or stylistic considerations (and not by typologi-
cal demands), at least from Early Modern English onwards.
12. transition
12.1. Preliminary considerations
One of the most diffcult decisions which had to be taken in the earlier phases
of the present investigation was to decide whether to include the linguistic
elements which will be discussed in the following sections in the quantitative
parts of this study. In the end, my apprehensions have proved to be negligible,
at least as far as the quantitative analyses are concerned. The items discussed
here show different patterns of change than those found for the connectors
marking the other semantic relations: most of them follow the adverbial
cline as proposed by Traugott (Traugott 1995, Traugott and Dasher 2002; see
above, Chapter 3.6). Since they are thus recurrently replaced onomasiologi-
cally (see below, 12.2.4), their inclusion or non-inclusion does not affect the
relative numbers of adverbial connectors as analysed in Chapter 6 above (as
test analyses have shown). Their different patterns of long-term development,
however, demonstrate that connectors marking the relation transition are
indeed peripheral.
These items, such as OE solice, OE witodlice, OE a, ME certes, ME for,
or PDE of course and PDE in fact, are certainly not prototypical adverbial con-
nectors, since they mainly work on the interpersonal and textual levels (for their
specifc characteristics and the reasons for their inclusion, see Chapter 3.4). If
discussed at all, they have recurrently been labelled discourse markers in vari-
ous research contexts (OE a in Enkvist and Wrvik 1987 and Kim 1992; OE
solice and witodlice in Lenker 2000, 2007b, EModE indeed and in fact in Trau-
gott and Dasher 2002). In the present study, they are termed transitional, be-
cause they are best comparable to connectors such as now and here (for the cat-
egory transition, see Quirk et al. 1985: 634). This classifcation also highlights
their peripheral position in the feld of adverbial connectors: In essence, these
connectors are different from most of the other adverbial connectors because
they do not express a core meaning such as the reinforcing additive relation or, in
particular, any of the ccc-relations cause, concession and contrast.
With regard to the origins of transitional connectors, we can establish two
unambiguous source domains (see Appendix B.5.1 and 2):
Source domain time and place/space (see above, Chapter 8.3):
OE a, OE (hwt) a; here, now (all periods of English)
228 TRANSITION
Source domain truth/fact (see above, Chapter 8.4):
OE eornostlice lit. earnestly, solice lit. truly, witodlice lit. certainly,
truly
ME certainly, certain, certes, forsooth, iwis, sekirly, truly, verily
EModE indeed, sure, surely
LModE in fact, in truth
As has been shown above (Chapters 3.4 and 8.4), these items share the essen-
tial properties of other adverbial connectors (for these properties, see above,
Chapter 2.2), i. e. they signal at least at certain stages of their developments
a two-place relation in that they combine two connects which may be of sen-
tential or even higher status (M3) and they furthermore do not contribute to the
propositional meaning of either segment.
By contrast, one or both of these defning conditions are not (unambigu-
ously) fulflled by two other groups of connectors which are listed in Appendix
B.1.3 and B.1.4 but were not analyzed in the quantitative parts of this study, i. e.
adverbs of the source domain uncertainty/doubt (12.2) and interrogatives
(see 12.3).
12.2. Source domain uncertainty/doubt:
peradventure, perchance, perhaps
In the draft version of the third edition of the OED (s. v. peradventure, 5.) a con-
nector function is suggested for ME peradventure (peraunter), EModE per
chance and perhaps. All of these forms are lexicalized prepositional phrases
comprising the Anglo-Norman preposition par/per for and a noun meaning
chance, good luck
1
with an epistemic meaning expressing a hypothetical,
contingent or uncertain possibility. In their core meaning, they do commonly
not indicate a two-place relation, but contribute to the propositional meaning
of the connect: they are stance adverbials (content disjuncts) with a scope over
the sentence only.
(201) All-be-itt at ou arte man, gitt parauntur ou arte not cristened; and
giff ou be cristeynd, gitt parauntur ou leueste not as Criste biddeste e.
But sir, ou seiste parauntur, Criste calle e frend and er-fore he
will do well with e and shewe e of is mercye ( CM ROYAL, p.16).
1 Peradventure and perchance are loans, whereas perhaps is a hybrid formation on
the pattern of peradventure and perchance (cf. hap < ON happ chance, hap, good
luck; see OED, s. v. perhaps).
Interrogatives 229
Even though you are human, yet perhaps you may not be baptized; and
if you are baptized, you do perhaps not live as Christ wanted you to. .
But, Sir, you perhaps say, Christ calls you friend and therefore he will
do you well and show you of his mercy.
(202) And nowe, perchance, some enuious reder wyll hereof apprehende oc-
casion to scorne me, sayenge that I haue well hyed me (CEEDUC1A,
p.28).
Yet there are also instances of semantic bleaching of peradventure, when it is,
for example, used in a statement of fact. The OED here suggests a transitional
reading as it happened (OED, s. v. peradventure, 5):
(203) Ensample as thus: The xiij day of March fyl upon a Saturday, peradven
ture (CMASTRO, 12, p.673).
Example like that: The 13th day of March happened to be on a Satur-
day, perhaps .
Although the exact date in March seems to rule out a modal meaning, even
this instance of peradventure is not unambiguous. The passage is quoted from
Chaucers handbook on the use of the astrolabe, and the dates are thus not real
dates but example dates created for pedagogical purposes (the OED, interest-
ingly, does not quote the introductory Ensample as thus).
12.3. Interrogatives
The other category which has been excluded are interrogatives such as OE
hwt (but not the phrase hwt a), EModE why and how. Most of these are
predominantly used in speech-based, interactive genres. They usually do not
signal a two-place relation either, but are similar in function to interjections
and rather encode an entire, separate message (termed pragmatic idioms by
Fraser 1999: 942943).
(204) (Lord Fop.) Why, thats the Fatigue I speak of, Madam: For tis impos-
sible to be quiet, without thinking: Now thinking is to me, the greatest
Fatigue in the World.
(Aman.) Does not your Lordship love reading then?
(Lord Fop.) Oh, passionately, Madam But I never think of what I read.
(Ber.) Why, can your Lordship read without thinking? (CEPLAY3A,
Vanbrugh, The Relapse, p.I, 36).
230 TRANSITION
In Old English, these uses are further distinguished from proper adverbial uses
in that they are like all particles not followed by inversion, but by subject
verb word order (for the relevance of this criterion, see in particular Fischer
2007: 280297). For an example illustrating the interrogative (205a) and the
interjection function of hwt (205b) see
(205) a cw ic: Hwt [205a] is t, la?
a cw he: Hwt, [205b] u wast t ic e r sde t sio soe
gesl wre good (COBOETH, 34.86.1516).
Then I said: What [205a] is that? Then said he, What, [205b] you know
that I before said to you that the true happiness was good.
12.4. The diachrony of transitional connectors
12.4.1. Paradigm shifts: truth fact
time and space have been shown to be the central source domains for all kinds
of adverbial connectors. Accordingly, now and much less frequently here
have been used as transitional connectors over all periods of English (on OE
a, see above, Chapter 5.3.1). The second major source category truth/cer-
tainty/fact is specifcally used for transitional and as a conversational im-
plicature concessive connectors (see above, Chapter 8.2).
A more detailed look at the chronology shows that connectors from the
source domain fact are only attested from the Early Modern English period
onwards in the fully lexicalized and univerbated indeed and the still even more
transparent in fact. At the same time, linguistic items formed from the source
domain truth are no longer predominantly used as transitional connectors,
but are mainly employed in contexts which suggest conversational implicatures
of concession. The data collected in Appendix B.5.2 for instance show that
verily and truly are only rarely attested from the Late Modern English period
onwards. Sure(ly) and the full phrase it is true, which often have conversa-
tional implicatures of concession, gain ground, as do indeed and in fact for
the trans itional relation.
It is, of course, tempting to relate these fndings to the larger changes in
socio-cultural paradigms or, in the terminology of Foucault (1966), epistemes
(for wider implications, see Mahler 1997): the substitution of connectors from
the source domain truth by those of the source domain fact might be seen
to refect a socio-cultural shift of the leading paradigm truth to the leading
para digm fact. In such a scenario, the paradigm of truth would be connected
The diachrony of transitional connectors 231
to theo-centricity and dominating role of religion and theology in the Middle
Ages; the leading paradigm of fact, on the other hand, could be said to refect
the dominating role of the natural sciences and the ideas of the Enlightenment.
The diachrony of transitional connectors in English at least suggests such a
replacement of the leading paradigms also in this feld, an idea, however, which
will have to be substantiated by material from other languages.
Table 121 Innovations of transitional connectors source domain truth/fact
O
l
d

E
n
g
l
i
s
h
1
1
0
0

1
3
5
0
(
M
E

1
/
2
)
1
3
5
0

1
4
2
0
(
M
E

3
)
1
4
2
0

1
5
0
0
(
M
E

4
)
1
5
0
0

1
5
7
0
(
E
M
o
d
E
1
)
1
5
7
0

1
6
4
0
(
E
M
o
d
E
2
)
1
6
4
0

1
7
1
0
(
E
M
o
d
E
3
)
1
7
1
0

1
7
8
0

(
L
M
o
d
E

1
)
1
7
8
0

1
8
5
0

(
L
M
o
d
E

2
)
1
8
5
0

1
9
2
0

(
L
M
o
d
E

3
)
1
9
2
0

(
P
D
E
)
eornostlice
(ge)wislice
witodlice witodlice
solice soothly soothly soothly
iwis
[forsoothe] forsoothe forsoothe
[certes] certes certes certes certes certes
truly truly truly truly truly truly truly truly
surely surely surely surely surely surely
sure sure sure sure
of
course
of
course
of
course
in fact in fact in fact
12.4.2. Regularities in semantic change
If we compare Table 12.1 with the tables summarizing the diachrony of ad-
verbial connectors of all the other semantic relations, we at frst glance note a
fundamental difference. In all the other relations, we fnd at least one central
connector which has remained stable from the middle of the Middle English
period onwards (after the periods M1 and M2 at around 1300; for the relevance
of this date for the question When did English begin?, see Lutz 2002): there
fore/for for the relations of cause/result, yet for the relation of concession/
contrast and furthermore/frst and at last etc. for the relation addition. In
the feld of transitional connectors, however, the scenario is different. We fnd
a constant exchange and replacement of linguistic material by near-synonyms.
This fnding corresponds to the ideas of regularities in semantic change as
proposed in Traugott and Dasher (2002), in particular to the adverbial cline
(see Traugott 1995) clause-internal adverbial > sentence adverbial > discourse
particle. In their closer analysis, Traugott and Dasher (2002) point out that
232 TRANSITION
it is important to note that no lexeme is required to undergo the type of change sche-
matized [here] The hypothesis is that if a lexeme with the appropriate semantics
undergoes change, it is probable that the language change will be of the type speci-
fed. More importantly, a reverse order of development is hypothesized to be ruled
out except under special circumstances such as language engineering (Traugott and
Dasher 2002: 281).
The data of the present study confrm this hypothesis for the adverbs which
have been analysed in detail in Chapter 8.4 above (OE solice, ME soothly, ME
forsoothe, ME verily). All of them follow the suggested cline: on the phrase
level, they may be used as manner adverbs, mainly in direct speech with a frst
person subject (e. g. in phrases such as ic secge solice I tell you truly) or they
are employed as emphasizers. As sentential adverbs (intermediate stage), they
function as transitional connectors, often in collocations with the conjunctions
and, but and for on the local level of discourse. On the global level of discourse
then, they are employed to mark episode boundaries.
All of these items thus followed the suggested adverbial cline once they
had started it, and none of them shows a reverse movement. Most interesting
in this respect is the highly frequent solice, which died out after trewely had
acquired its epistemic and pragmatic functions. Another interesting case is for
soothe, which, after having been forced out by trewely, has only survived in a
highly intersubjective and negatively connotated function, i. e. parenthetically
with an ironical or derisive statement (OED, s. v. forsooth; see above, Chapter
8.4.5 and Lenker 2003: 283286). This also supports Traugott and Dashers
view that items after they have reached the right end of the cline are re-
placed onomasiologically (Traugott and Dasher 2002: 284):
A third issue is what happens to those polysemies that have developed non-truth-con-
ditional, procedural, scope-over-discourse, and intersubjective meanings. Our hy-
pothesis is that Ls [= lexemes] with such meanings are replaced by newly recruited
polysemies from other Ls, i. e. do not continue further semasiologically, but are
replaced onomasiologically.
These instabilities and recurrent replacements in the feld of transitional ad-
verbial connectors coined from the source domain truth
2
again show that the
transitional category is different to the other categories, and that it is peripheral.
2 It is too early to see these developments in the adverbial connectors coined from the
source domain fact.
13. Perspicuity and the New Rhetoric
Of all the parts of speech, the conjunctions are
the most unfriendly to vivacity
(Campbell [1776] 1963: 395)
13.1. Collocations vs. medial position of adverbial connectors
At the very beginning of this study, it was pointed out that the use of adverbial
connectors instead of asyndetic constructions or conjunctions was not only
a matter of grammar and information structure, but also of rhetoric and style.
In the concluding chapter, I will now resume this topic and show that refec-
tions on the use of particles, such as Lockes (see above, Chapter 1), indeed had
an impact, since the middle of the eighteenth century marks a shift in the use of
adverbial connectors. These changes are related to the ideas of perspicuity and
propriety as argued for in the spirit of Locke and the members of the Royal So-
ciety, and were further made possible by developments in the system of punc-
tuation. This interaction of logical and rhetorical analysis was also affected by
a change which took place in the role of rhetoric as expressed in the writings
of the New Rhetoricians, in particular in George Campbells The Philosophy
of Rhetoric ([1776] 1963).
Let me introduce these issues by sample passages from texts of a compa-
rable text type, i. e. treatises, by renowned, mature authors of their respec-
tive periods, whom we would normally not accuse of not knowing how to
write better prose. (206) is taken from Tale of Melibee by Geoffrey Chaucer
(ME3; 1343?1400), and (207) is from Adam Smiths The Wealth of Nations
(LModE1; 17231790).
(206) (a) but certes [206a] what ende that shal therof bifalle, it is nat light
to knowe. For soothly [206b], whan that werre is ones bigonne, ther is
ful many a child unborn of his mooder that shal sterve yong by cause
of thilke werre, or elles lyve in sorwe and dye in wrecchednesse. And
therefore [206c], er that any werre bigynne, men moste have greet con-
seil and greet deliberacion. And whan [206d] this olde man wende to
enforcen his tale by resons, wel ny alle atones bigonne they to rise for
to breken his tale, and beden hym ful ofte his wordes for to abregge.
For soothly [206e], he that precheth to hem that listen nat heeren his
wordes, his sermon hem anoieth (CMCTPROS, Chaucer, Tale of Me
libee, p.219.C2).
234 Perspicuity and the New Rhetoric
(b) And therfore [206f] ye shul venge yow after the ordre of right; that
is to seyn, by the lawe and noght by excesse ne by outrage. And also
[206g], if ye wol venge yow of the outrage of youre adversaries in oother
manere than right comandeth, ye synnen. And therfore [206h] seith
Senec that a man shal nevere vengen shrewednesse by shrewednesse
(CMCTPROS, Chaucer, Tale of Melibee, p.232, C.1).
(207) (a) This portion, however [207a], may still be considered as the natural
rent of land, or the rent at which it is naturally meant that land should,
for the most part, be let. The rent of land, it may be thought, is frequently
no more than a reasonable proft or interest for the stock laid out by the
landlord upon its improvement. This, no doubt [207b], may be partly
the case upon some occasions; for [207c] it can scarce ever be more than
partly the case. The landlord demands a rent even for unimproved land,
and the supposed interest or proft upon the expense of improvement is
generally an addition to this original rent. Those improvements, besides
[207d], are not always made by the stock of the landlord, but sometimes
by that of the tenant. When the lease comes to be renewed, however
[207e], the landlord commonly demands the same augmentation of rent
(CLSMI1, The Wealth of Nations, Chapter XI, Part 1).
(b) Hence [207f] a greater rent becomes due to the landlord. It requires, too
[207g], a more attentive and skilful management. Hence [207h] a greater
proft becomes due to the farmer. The crop, too [207i], at least in the hop
and fruit garden, is more precarious. Its price, therefore [207j], besides
compensating all occasional losses, must afford something like the proft
of insurance (CLSMI1, The Wealth of Nations, Chapter XI, Part 1).
13.2. Sentence-initial collocations
The differences between these two passages, which are prototypical of their re-
spective period, are obvious at frst glance: in the passages from Chaucers Tale
of Melibee, all of the sentences start with explicit markers of textual cohesion.
In all cases, these explicit markers are not simply conjunctions or adverbial
connectors, but collocations of a conjunction (which functions as a rather loose
connective) and an adverbial connector. The adverbial connector may express
an additional semantic relation such as cause (and therefore), addition (and
also) or transition (but certes, for soothly).
Collocations like these are typical of Chaucers prose (Kerkhof 1982: 456,
459). Unfortunately, with the exception of the introductory passages of the Trea
Medial positions of adverbial connectors 235
tise on the Astrolabe (CMASTRO), all of Chaucers prose works are translations
from Latin or French (CMBOETH: Boece, CMCTPROS: Tale of Melibee; Par
sons Tale), so that we cannot rule out loan infuence in these collocations. Yet,
the predilection for collocations is also widely attested in prose texts which were
originally composed in Middle English, such as in Caxtons Prologues:
(208) For in the sayd boke they may see what this transitorie & mutable worlde
is And wherto euery mann liuyng in hit / ought to entende Thenne for
as moche as this sayd boke so translated is rare & not spred ne knowen
as it is digne and worthy For the erudicion and lernyng of suche as ben
Ignoraunt & not knowyng of it And furthermore I desire & require
you and therfore he ought eternelly to be remembrid. of whom the
body and corps lieth buried (CMCAXPRO, p.63).
It is also the preferred position of connectors in Early Modern English. In a
quantitative study of Jonathan Swifts prose style, Milic (1967) shows that
Swift starts a third of his sentences with collocations of a conjunction (as a
rather loose connective) and an adverbial connector. Milic also analyzes the
rhetorical and stylistic functions of Swifts use of connectors and fnds that
they lead to rhetorical persuasiveness rather than clarity. Swifts prose may not
be clear, but it induces the reader to believe in the clarity and simplicity of
what he has read (Milic 1967: 254).
13.3. Medial positions of adverbial connectors
Not a single one of these collocations is found in the passages from Adam
Smiths The Wealth of Nations (1776). There are sentence-initial connectors,
such as causal for (207c) and two instances of resultive hence (207f, 207h). All
of the other adverbial connectors are placed sentence-medially, such as reinforc-
ing besides (207d) and too (207g, 207i), contrastive however (207a, 207e), and
resultive therefore (207j). Most of these are placed in post-frst-position (207a,
207d, 207e, 207i, 207j), i. e. the position after the frst obligatory constituent.
Adverbials in post-frst-position have since Old English been used to
indicate that the element in frst position is in the (contrastive) focus (see above,
Chapter 5.4.2). Post-frst-position in accordance with its discourse functions
is mainly found with adverbial connectors signalling the semantic relations
result/inference or contrast/concession. In Middle English and Early
Modern English, post-frst-position of adverbial connectors is only rarely at-
tested (in only ca. thirty instances in my corpus for Middle English and ca.
236 Perspicuity and the New Rhetoric
twenty-fve instances for Early Modern English). Furthermore, it is predomi-
nantly restricted to resultive/inferential therefore and then:
1
(209) For in e loue of Ihesu ere schal be in help. Loue is soche a migt at it
maki alle ing comoun. Loue erfore Ihesu, (CMCLOUD, p.21).
For in the love of Jesus will be your help. Love is such a might that it
makes all things equal. Love therefore Jesus, ..
(210) Men that hunt so, be either ignorant persones, preuie stealers, or night
walkers. Learning therefore, ye wise fathers, and good bringing vp,
and not blinde & dangerous experience, is the next and readiest waie,
that must leede your Children, frst, to wisdom, and than to worthinesse
(CEEDUC1B; Ascham, The Scholemaster, p.215).
(211) is prayoure passe all our in worthynesse, for God hym-selfe made itt;
By is preyour, an, late ichon of vs onke and preye God as Crist
and Seynt Poule teche vs (CMROYAL, p.9).
This prayer surpasses all other prayers in spiritual worth, because God
himself made it, . By this prayer, then, let each one of us thank and
pray to God as Christ and Saint Paul teach us.
(212) Such a Schoole then as may be ft for the education of all sorts of child-
ren should be situated in a City or Town of great concourse and trad-
ing, whose inhabitants are generally addicted, and suffciently accom-
modated to entertain Tablers (CEEDUC3B; Hoole, p.221).
In the passages quoted from the Wealth of Nations above in (207), by contrast,
we also fnd additive, reinforcing connectors in post-frst-position. Moreover,
medial connectors are not restricted to this position after the frst obligatory
sentence constituent, but may also follow the verb (207g).
Another common position attested from the Late Modern English period
(increasingly from LModE2) onwards is the position after an operator
1 The only exception after the period ME1 are two instances of post-frst soothly in
the main part of Chaucers Treatise of the Astrolabe (as compared to four instances
of the collocation but soothly): The ascendent sothly, as wel in alle nativites as in
questions and eleccions of tymes, is a thing which that these astrologiens gretly ob-
serven. The ascendent sothly, to take it at the largest, is thilke degre that ascend-
ith at eny of these forseide tymes upon the est orisounte. such as (CMASTRO,
p.670).
Corpus fndings 237
(213) The natural sciences do not, however, stand on the same footing with
these instrument-knowledges (CLARN3, 127).
(214) The natural strength of such literature will, of course, be in the line of
its tendencies; in transparency, variety, and directness. (CLPAT3, 11).
or between a verb and a complement clause:
(215) It is time, however, to pass on to those precursors of Babism who were
(CLCHEY3).
(216) There is still some chance, however, that Sufsm may be a record of its
activity (CLCHEY3).
(217) The case is so marked, however, that illustrations easily overfow, and
there is no need of forcing doors that stand wide open (CLJAM3).
Medial adverbial connectors may also be doubled, so that we fnd collocations,
such as
(218) In speaking of the future of the novel we must of course, therefore, be
taken as limiting the inquiry to those types that have, for criticism, a
present and a past (CLJAM3; Henry James, Future of the Novel).
It has to be emphasized again that the post-frst-position has to be distinguished
from the other medial positions. Sentence-medial (just as sentence-initial and
sentence-fnal) adverbial connectors serve the function of combining their
connects. Adverbial connectors in post-frst-position do not only express this
function, but in addition single out one individual constituent of one of the
connects and put it in the focus.
13.4. Corpus fndings
13.4.1. Present Day English
The corpus fndings for the position of adverbial connectors in Present Day
English confrm this impression of a high number of sentence-medial connec-
tors in the written genres (see Biber et al. 1999: 889, Table 10.18; = 5 %; =
less than 2.5 %):
238 Perspicuity and the New Rhetoric
Table 131: Positions of linking adverbials in conversation and academic prose
(see Biber et al. 1999: 889, Table 10.18)
% in initial position % in medial position % in fnal position
CONV

ACAD

= 5 %; = less than 2.5 %
In the corpus of the Longman Grammar, medial adverbial connectors are pre-
dominantly attested for the written genres; in less then 2.5 per cent of all in-
stances are adverbial connectors placed sentence-medially in spoken registers.
academic prose clearly favours the initial (50 per cent) and medial positions
(more than 40 per cent). The placement of adverbial connectors in the middle
of a sentence is thus a shibboleth of written English.
13.4.2. Old English to Late Modern English
As has been illustrated by the examples above, medial positions apart from
post-frst-position have only been attested in higher frequencies since the
Late Modern English period. This assertion was put to the test in a corpus
study, using the smaller corpus of Treatises and Homilies (for the principles
of compilation and the texts, see Appendix C.2). The analysis focuses on the
relation of the coordinators and, but (OE ac) and for (all of which obligatorily
have to be placed sentence-initially) and adverbial connectors.
Figure 131: Conjunctions vs. adverbial connectors (including OE a) (all occur-
rences are given in normalized frequencies per 10,000 words).
Corpus fndings 239
Figure 13.1. shows that ME 1/3 and Late Modern English 2/3 mark shifts in
the use of connectors. In many respects, Middle English has in this study been
found to be a period of transition and experiment in the use of adverbial con-
nectors. The sharp decrease of adverbial connectors in Figure 13.1 from OE to
ME1 is mainly due to the fact that OE a, the shibboleth of Old English narra-
tive style (see above, Chapter 5.3.1), is only rarely used from that time onwards.
If we exclude the instances of OE a (see Figure 13.2), we see a more even
distribution of conjunctions and adverbial connectors over most of the periods
of English.
Figure 132: Conjunctions vs. adverbial connectors (without OE a)
Yet in the period Late Modern English 3, i. e. the English of the nineteenth cen-
tury (18501920), we see a sharp decrease in the number of the conjunctions.
This is mainly due to the decreasing frequencies of and and, in particular, for,
and leads to the differences between the spoken and the written mode in the
frequency of and. Figures 13.1 and 13.2 thus implicitly corroborate the fndings
on and by Culpeper and Kyt (2000), i. e. that there was not much difference
between the use of and in the spoken and written medium until the end of the
Early Modern English period (see above, Chapter 11.7).
The differences are, as the examples above illustrate, not only due to an
increasing use of different adverbial connectors, but to deliberate choices of
authors not to use sentence-initial collocations, but medial adverbial con-
nectors.
240 Perspicuity and the New Rhetoric
Figure 133: Collocations (conjunction + adverb) vs. adverbial connector, medial
position (without OE a)
Figure 13.3 shows that the impression we have gained from the examples above
is indeed supported by the quantitative analysis of the corpus texts (C.2).
We again see two periods emerge as decisive. First, the beginning of Mid-
dle English saw a rapid increase in the number of sentences which are in-
troduced by a conjunction (and, but, or for) in a collocation together with an
adverbial connector (see the examples from Chaucer in (206)). In Old English,
the number of medial connectors had been higher (even if we exclude OE a)
because Old English regularly allowed post-frst-position of adverbial connec-
tors. ME3 (13501420) again appears as a period of experiment and transi-
tion and shows the highest numbers of such collocations. This process can
be explained by the attempts at the evolution of an English prose style during
the fourteenth and ffteenth centuries (see Mueller 1984), when English had
not been used in the written medium in particular in offcial registers and
prose literature for almost two centuries, but had been supplanted by Latin
and French.
The second decisive period appears in the language of the late eighteenth
and early nineteenth century (LModE2 from 17801850), when we see a re-
versal of the distribution of sentence-initial collocations and medial adverbial
connectors. The difference, which leads to a specifc Modern English prose
style, is thus not that initial conjunctions or adverbial connectors are no longer
used in Present Day English academic prose: with more than 50 per cent, the
initial position can, even in academic prose, still be considered the unmarked
position for adverbial connectors (see Biber et al. 1999: 891). More than 40 per
cent of adverbial connectors, however, are now placed in medial position. This
Copia perspicuitas 241
differentiates the spoken from the written language, since in spoken English
less then 2.5 per cent of connectors are placed sentence-medially. The position-
ing of adverbial connectors has become one of the most specifc properties of
the English prose style of the nineteenth and twentieth centuries.
13.5. Copia perspicuitas
13.5.1. Perspicuity
This process did not start by chance in the eighteenth century, but was primari-
ly inspired by the fundamental changes in the perception of language in the
philosophy of the Enlightenment and, in particular, in the spirit of John Locke.
In her chapter on Literary Language in the Cambridge History of the Eng
lish Language covering the years 1476 to 1776, Adamson shows that there is a
profound stylistic gulf which separates Bacon from Locke (Adamson 1999:
541). Accordingly, Adamson (1999) divides her account into two overlapping
phases, which are governed by two distinct stylistic key concepts: copia copi-
ousness vs. perspicuitas perspicuity. While the paradigm of copia dictated
the stylistic norms of the Renaissance phase from 1500 to 1667, the concept of
perspicuity determined the stylistic ideas of the neo-classical phase from the
1640s onwards.
The concept of perspicuity is primarily indebted to attitudes on language
found in the philosophical investigations of Lockes Essay Concerning Human
Understanding. Locke emphasized that reason fundamentally consists in fnd-
ing connections between ideas, and he saw knowledge as the Perception of
the Connection or Agreement, or Disagreement and Repugnancy of any of our
Ideas (Locke [1690] 1975: IV, i, 2). With respect to language, Locke argued for
a common usage and propriety in language:
Mens Intentions in speaking are, or at least should be, to be understood; which can-
not be without frequent Explanations, Demands, and other the like incommodious
Interruptions, where Men do not follow the common use. Propriety of Speech is
that which gives our Thoughts entrance to other Mens Minds with the greatest Ease
and Advantage; and therefore deserves some part of our Care and Study (Locke
[1690] 1975: III, xi, 1)
This change in perspective has often been described as a rejection of rhetoric,
but it should rather be described with Adamson as a redirection of rhetoric
(Adamson 1999: 599). Like copia, perspicuity has its roots in classical anti-
quity, in particular in the writings of Quintilian (who saw perspicuity as a prac-
242 Perspicuity and the New Rhetoric
tical necessity for the orator), but it was turned into a key concept of style in
the late seventeenth and eighteenth century. Writers divert their energies away
from copia towards alternative goals, such as the clart that the French Acad-
emy prescribed as the frst virtue of literature (see Thomas and Turner 1994).
Similarly, the Royal Society rather harshly obliges their members to ban copia
from their writings:
They have therefore been most rigorous in putting into execution, the only Remedy,
that can be found for this extravagance: and that has been, a constant Resolution,
to reject all the amplifcations, digressions, and swellings of Style: to return back to
the primitive purity, the shortness, They have exacted from all their members,
a close, naked, natural way of speaking: positive expressions, clear senses; a native
easiness; bringing all things as near the mathematical plainness, as they can
(Sprat, The History of the Royal Society, [1667] 1959: 113).
13.5.2. Punctuation
The ideas of perspicuity have infuenced the use of adverbial connectors in
two fundamental ways. In his survey on the history of punctuation, Parkes
demonstrates that the attempts at a new kind of punctuation, representing the
balance between grammatical, logical and rhetorical analyses of discourse,
were to a large extent affected by attitudes towards language as promoted by
Locke (Parkes 1992: 8991). Increasingly, it was felt that a consensus should be
observed in the application of the diverse punctuation marks, and thus Lowths
grammar is the frst one containing a chapter on punctuation (Lowth [1762]
1969: 154172; see also Salmon 1988).
When adverbial connectors are placed in the middle of a sentence, fxed
rules in punctuation are essential for an unambiguous en- and decoding of the
semantic relations between two connects. While the placement of adverbial
connectors in post-frst-position is, because of its contrastive focus, also fea-
sible in the spoken medium and in periods without a fxed system of punctua-
tion, other positions in the middle of a sentence require a clear, unambiguous
system of a rhetorical punctuation.
In the majority of the instances of medial adverbial connectors in Late Mod-
ern English and Present Day English, these are distinguished from circum-
stance adverbials by commas. These commas are commonly explained as mir-
roring pauses in the spoken medium. This, however, is an argument which is
not at all substantiated by the actual language data: there are almost no medial
adverbial connectors in spoken discourse. In view of the ideas of perspicuity
and the attempts at mathematical plainness in the Royal Society (see the quote
Copia perspicuitas 243
from Sprat ([1667] 1959), above), they rather resemble mathematical or logical
symbols, which guide the reader in the interpretation of a formula or in the case
of natural language, a text.
13.5.3. The Scottish Rhetoricians: The New Rhetoric
It is also not accidental that example (207), which illustrates the use of medial
adverbial connectors, is taken from Adam Smiths The Wealth of Nations, since
Adam Smith is one of the main representatives of a group of scholars forming
the Scottish Enlightenment (among them also David Hume and Hugh Blair).
In this atmosphere, a new kind of rhetoric was developed which replaced the
old classical art of oratory (see Gneuss 1996: 3233). The leading exponents
of these Scottish Rhetoricians were George Campbell (The Philosophy of
Rhetoric, Campbell ([1776] 1963); over forty reprints in the nineteenth century)
and Hugh Blair (Lectures on Rhetoric and Belles Lettres, Blair ([1783] 1965);
at least 130 editions until 1911).
The high number of editions and reprints demonstrates that these books were
most infuential for the formation of stylistic ideas in the eighteenth and nine-
teenth centuries. The books propose following Lockes lines of thoughts a
new kind of rhetoric which places much less emphasis on tropes or rhetorical
fgures, but deals with syntactical, text-linguistic and pragmatic issues, such
as sentence structure and paragraphs, which had been almost completely neg-
lected before (see above, Chapter 3.5): the fundamental qualities of this new
style are precision and perspicuity (Blair [1783] 1965, Lecture X).
13.5.4. George Campbells The Philosophy of Rhetoric (1776)
Therefore it does not come as a surprise that George Campbells The Philoso
phy of Rhetoric is the frst metalinguistic study which explicitly deals with
conjunctions and adverbial connectors, and also with their respective func-
tions: He explains why he in order to conduce more to perspicuity (Camp-
bell [1776] 1963: 385) decided to devote a whole chapter each to connectives
working on the level of the phrase and sentence (Campbells Chapter IV) and
also to Connectives employed in combining the Sentences in a Discourse
(Campbells Chapter V).
In the connection of sentences of larger chunks of discourse, Campbell cat-
egorically rejects asyndetic constructions:
244 Perspicuity and the New Rhetoric
It will scarcely be doubted by any person of discernment, that as there should al-
ways be a natural connexion in the sentiments of a discourse, there should gener-
ally be corresponding to this, an artifcial connexion in the signs (Campbell [1776]
1963: 403).
After a general account of the diverse categories of conjunctions and adverbial
connectors, Campbell adds altogether seven remarks dealing with the use of
the conjunction and on its own or in collocations:
(Remark 1) Being a monosyllable, it [and] will ., if not used too often, serve to
smooth the current of discourse (Campbell [1776] 1963: 409).
(Remark 2) One of the best expedients for preventing the connexives from becom-
ing to conspicuous, is to avoid the frequent recurrence of the same particles, espe-
cially if they consist of more than one syllable. (Campbell [1776] 1963: 409).
(Remark 3) Another useful expedient for answering the same end is to vary the situ-
ation of the conjunction, wherever the idiom of the tongue and the harmony of the
sentence will permit the variation (Campbell [1776] 1963: 411).
(Remark 4) Though certain circumstances require that one connexive be immedi-
ately followed by another, the accumulating of these without necessity ought always
to be avoided (Campbell [1776] 1963: 411).
All of these remarks have, as my quantitative analyses have shown, proved to
be highly infuential in the evolution of a genuine English prose style.
2
Remark
(2) has, as has been shown above in Chapter 6.3, led to a steady increase in
the number of types of connectors used in different texts. Campbell concedes
that there may be cases when the repetition of a connector fulfls an expressive
function, comparable to other rhetorical repetitions. Yet he fully rejects the
repetition of connectors comprising more than one syllable:
The now, and, for, but, nay, nor, have this advantage from their brevity, that though
often repeated they pass unnoticed. But who, that hath any taste, can endure the
incessant quick returns of the alsos, and the likewises, and the moreovers and
howevers, and the notwithstandings? An excess in these is insupportable. (Camp-
bell [1776] 1963: 409)
With respect to the dramatic changes in constituent order demonstrated above,
Remarks (3) and (4) are clearly the most important ones, since they recom-
2 It is interesting to note that Campbell places English over the classical languages in
the summary of his Chapter Modern Languages compared with Greek and Latin,
particularly in regard to the Composition of Sentences: For this reason, I should
not hesitate to pronounce that English is considerably richer than Latin, and in
the main ftter for all the subtle disquisitions both of philosophy and of criticism
(Campbell [1776] 1963: 402403).
Copia perspicuitas 245
mend variation in the constituent order (Remark 3) and explicitly disapprove of
sentence-initial collocations of conjunction and adverbial connector (Remark
4). In agreement with iconic principles, Campbell regards the sentence-initial
position as the unmarked position for connectors.
The place where we should naturally expect to fnd it, when it connects two sen-
tences, is doubtless the beginning of the second (Campbell [1776] 1963: 410).
Yet Campbell rejects collocations of connective elements in all cases in which
they are not necessary to encode two different relations. Collocations such as
and if, for example, are not presented as problematic, because and serves as an
additive connector, while the subordinator if introduces the subordinate condi-
tional clause. And and if thus work on two different planes of discourse.
In view of the fact that Campbells book was used as one of the most import-
ant style guides in the nineteenth century, it is furthermore crucial that he
explicitly condemns collocations of three connectors, because they are incon-
sistent with the principles of perspicuity, of vivacity, or of elegance (Campbell
[1776] 1963: 413). By contrast, Campbell recommends variation in the position
of the adverbial connectors (conjunctions, of course, cannot be moved) because
a position later in the sentence serves as a cover to render it less observable:
In the beginning it stands by itself; whereas, placed in the manner now mentioned,
it may be said to stand in a crowd (Campbell [1776] 1963: 413).
He then goes on to describe the possibilities of variation in the English of his
time. Interestingly, Campbell feels that the number of syllables of the respec-
tive items is decisive:
With us in particular, no monosyllabic conjunction, except the illative then, can be
thus transposed (Campbell [1776] 1963: 410).
Polysyllabic adverbial connectors are, on the other hand, said to be much more
versatile:
Our language, however, hath been abundantly indulgent (where indulgence is of
greater consequence) in the power it gives us in the disposal of those which consist
of more than one syllable. Thus almost all the copulatives [reinforcing connectors,
U. L.] which come under this denomination, the disjunctives however and neverthe
less and the illative therefore, may be shifted to the second, the third, the fourth
place, and even further (Campbell [1776] 1963: 410).
This clearly descriptive account is very illuminating for a synchronic analysis
of the use of English adverbial connectors at the end of the eighteenth cen-
tury. Campbell himself, although very self-confdent in most of his evaluations,
seems somewhat at a loss in trying to give the reasons for this distribution.
246 Perspicuity and the New Rhetoric
It would be diffcult to assign a satisfactory reason for the difference that hath
been made in this respect, between conjunctions of one syllable and those of more.
(Campbell [1776] 1963: 410).
This is a clear indicator that there was a situation of ongoing language change
in the English of the late eighteenth century. The distinctions and restrictions
noted by Campbell are clearly intermediate steps, all of which have eventually
been given up.
Campbell, for instance, specifcally comments on the position of again in
a footnote, saying that the reinforcing again cannot conveniently be trans-
posed, as it would scarcely fail to occasion an ambiguity, and be mistaken for
the adverb signifying a second time (Campbell [1776] 1963: 410, n. 5). This,
however, had changed by the latest Late Modern English period (LModE3):
(219) He, again, who thirsts alone for power squanders his wealth, despises
pleasure, and thinks fame and rank alike worthless without power (CL-
BOETH3).
(220) There, again, is the constitutional shrinking, through a kind of meta-
physical prejudice, from the concrete that fear of the actual in this
case, of the Church of history; to which the admissions, which form so
large a part of these volumes, naturally lead (CLPAT3, 33).
Likewise, the constraints on monosyllables are no longer valid from the nine-
teenth or twentieth century onwards. In the period LModE3 (18501920), the
monosyllables here, frst, thus, and still are frst attested in adverbial connector
function. In Present Day English, all adverbial connectors may be placed sen-
tence-medially in the written mode. The proliferation of this position in written
English since the nineteenth century may thus be a consequence of Campbells
remarks, since his Philosophy of Rhetoric has seen so many reprints.
14. Conclusions
A close reasoner and a good writer in general may
be known by his pertinent use of connectives. Read
that page of Johnson; you cannot alter one conjunc-
tion without spoiling the sense. It is a linked strain
throughout. In your modern books, for the most part,
the sentences in a page have the same connection with
each other that marbles have in a bag; they touch with-
out adhering.
Samuel Taylor Coleridge, Table Talk
The present study set out to examine the changes in English adverbial connec-
tors, in particular in contrast to other connectors, such as coordinating and sub-
ordinating conjunctions. Connectors are here defned as linguistic items which
signal a two-place relation between segments of text (here called connects)
above the level of the phrase, i. e. between sentences or chunks of discourse.
The meaning of a connector is procedural, not conceptual: it does not change
the propositional content of any of the segments it relates. Connectors thus
reveal or make explicit the connections already operating in a text. They are
powerful clues about what commitment the speaker or writer makes regarding
the relationship between sentences or chunks of discourse. Adverbial connec-
tors in particular work as linguistic signposts, guiding the reader through a
text. Accordingly, they have in all periods of English been particularly fre-
quent in genres such as academic prose, which puts an emphasis on convey-
ing logical coherence and whose main communicative purpose is information,
argumentation and explanation for a specialist audience. Adverbial connectors
are a very apt means for these ends because they allow writers to mark the
development of their arguments overtly by relating one proposition to another
and by explicitly showing contrasts, restatements and conclusions.
Traditionally, linguists differentiate three kinds of connectors which are
distinguished predominantly by topological criteria: coordinating conjunc-
tions, subordinating conjunctions and adverbial connectors. The present
study, however, emphasizes that these connectors are by no means function-
ally equivalent, and sees issues of discourse deixis, textual information and
information processing as crucial for an understanding of the different or even
diverging developments in the history of coordinators, subordinators and of
adverbial connectors in English. In this view, adverbial connectors in particu-
lar are in contrast to subordinators, which work on a more local level (the
248 Conclusions
information presented in the subordinate clause is commonly not pursued in
the following discourse) very strong indicators of an equal illocutionary
weight of both of the connects. Since adverbial connectors are employed to
explicitly highlight or specify the relation already operating in the text (which
may already be signalled less explicitly or specifcally by a coordinator such
as and, but or or), they are themselves commonly very explicit in marking
the respective anaphoric (or cataphoric) relations, by, for example, so-called
pronominal connectors such as German des-wegen or Old English form
(because of this).
Since adverbial connectors have as yet only rarely been discussed in the lit-
erature in a more systematic way, it was frst of all necessary to compile inven-
tories of all the adverbs which have been employed as adverbial connectors in
the respective periods of English (see Appendix A.1 and the Index of Adverbial
Connectors, Appendix A.2). These inventories then allowed a frst comparison
with the diachrony of English subordinators as sketched in Kortmann (1997).
The origins of both of these connectors lie in Old English forms commonly
called ambiguous adverbs/conjunctions, i. e. linguistic items which are syn-
tactically and semantically polyfunctional. The very general line of develop-
ment is the same for subordinators and adverbial connectors: in the course of a
better form-to-function mapping after the Old English period, most of the am-
biguous adverbs/conjunctions are discarded, so that we see a corresponding
decrease in syntactic polyfunctionality and also in semantic polyfunctionality
in both now separate classes. Adverbial connectors and subordinators are
further similar in that the number of items (i. e. types) expressing these func-
tions has increased from Old English to Present Day English.
Yet we also observe striking differences in the developments of these two
types of connectors which ask for an explanation. With respect to subordina-
tors, Middle English emerges as the crucial period in which most of todays
subordinators were coined. The Middle English inventory of subordinators re-
sembles that of Modern English most closely. As regards adverbial connectors,
by contrast, there is no single period which could be said to have supplied the
bulk of the adverbial connectors of todays inventory. On closer inspection, we
fnd that Middle English emerges as a period of experiment and transition:
a large number of adverbial connectors were coined in the Middle English
period, but most of them were ephemeral and have not survived into Modern
English.
An examination of their etymological origin and morphological make-up
furthermore shows that the specifc characteristics in the history of adverbial
connectors are not typical of the general expansion of the English lexicon in
Middle and Early Modern English: although adverbial connectors are mainly
Conclusions 249
employed in written, specialized registers which commonly show a high per-
centage of Romance vocabulary, we, surprisingly, very rarely fnd French or
Latin loans among the new coinages. Most of the common innovations in
the feld are univerbated or at least lexicalized prepositional phrases. These
prepositional phrases became frequent at exactly the time when another
pattern, the pronominal connectors typical of Old English, were no longer
commonly used, let alone newly coined. Furthermore, we also see that these
prepositional phrases basically only employ two cognitive source domains:
time and space.
In the present study, the term shifting deictics is used to summarize these
processes, which mainly affect adverbial connectors expressing the impure
relations cause and contrast/concession. After the collapse of both of the
paradigms of the demonstratives at the end of the Middle English period,
pronominal connectors could only be formed by using the new demonstra-
tives that and this, which are, however, no longer infected for gender and
case. That, however, came to be the general subordinator in Middle English
and served as a complementizer, relativizer, and, more importantly, was often
also used with simple subordinators in order to highlight their subordinating
function. This means that an adverbial connector such as for that is not only
ambiguous, but misleading: readers would expect a pre-posed subordinate
clause introduced by for that and not an adverbial connector. Accordingly, we
see that pronominal connectors which were still coined in the (early) Middle
English period are only rarely used from Early Modern English onwards
(mainly in fully transparent prepositional phrases). The overtly deictic charac-
ter, which is for the reasons of information processing sketched above al-
most a prerequisite for adverbial connectors, is now expressed by lexicalized
phrases which signal the anaphoric relations by time or space deixis (such as
PDE after all or hence). space in particular becomes more important from
the later Middle English period onwards, when we see a spatialization of lan-
guage in literacy.
This shows that the history of adverbial connectors indeed mirrors over-
arching principles of long-term linguistic change in English: the re-structuring
of the system of adverbial connectors is at least partly a consequence of
typological changes (see the analogous developments in OE for m (e) to
for and of French par ce que to pas in French-based creoles). Similar fac-
tors of typological change may also have been the cause for the emergence of
sentence-fnal adverbial connectors, such as PDE sentence-fnal though, since
this retrospective marker of linkage fulfls similar functions as PDG obwohl
with main clause word order (instead of verb-fnal), a differentiation no longer
made in Present Day English.
250 Conclusions
The history of adverbial connectors also attests to the increasing impor-
tance of lexical bundles or clusters of words (Wortverband) instead of loans
from the Later Modern English period onwards, a phenomenon which can in
this case since it affects linguistic items which are mainly used in genres
such as academic prose hardly be accounted for as processes of colloqui-
alization.
As concerns cross-linguistic patterns of language change, we see the univer-
sal patterns in the formation of concessives in the use of originally temporal
phrases denoting simultaneity or concomitance (at the same time, meanwhile)
or the use of negatives (nevertheless). The peripheral group of transitional con-
nectors also attests clear instances of the adverbial cline which has been
established as one of the clines in a semantic-pragmatic approach to gram-
maticalization in the regularities of semantic change in OE solice, ME for
sooth or ME verily from adverbs used as circumstance adverbials (lit. truly)
to non-propositional transitional connectors.
Complex connectors such as nayls and swaeahhwere nevertheless
are typical examples of the general iconic principle that formal complexity cor-
responds to conceptual complexity. More characteristic for adverbial connec-
tors are (often pleonastic) compound adverbs expressing the relation addition,
such as furthermore, overmore or moreover. These do not signal cognitive
complexity, but are iconic in their length, in that they not only overtly signal,
but expressively highlight the semantic relation which is already operating in
the linear sequence of the text. This shows that the use of adverbial connectors
is also affected by rhetorical preferences and refections.
These infuences are most noteworthy in changes triggered by the idea of
perspicuitas (which replaces copia as the leading style paradigm in the eigh-
teenth century). Adverbial connectors, which make the speakers commitment
as to the relations of discourse segments explicit, are a very apt means of fos-
tering perspicuity in language. The spirit of perspicuity also led to a more sys-
tematic punctuation, which in turn allowed the medial positioning of adverbial
connectors in the written mode. The commas, which commonly signal the
status of an adverb as an adverbial connector (and not a circumstance adver-
bial), do, however, not primarily refect pauses in the same construction in the
spoken language. Since the medial positioning of connectors is rarely attested
in spoken language, these commas should rather be regarded as rhetorical or
logical (almost mathematical) signs signalling an authors attempt at perspi-
cuity. Since adverbial connectors are such an apt means for perspicuity, they
have from the end of the eighteenth century onwards become, at least in
genres such as academic prose, the principal connectors and are employed
even more frequently than conjunctions.
References
1. Corpora
CLMET The Corpus of Late Modern English Texts [see De Smet 2005]
<http://perswww.kuleuven.be/~u0044428/clmet.htm>
CME Corpus of Middle English Prose and Verse
<http://www.hti.umich.edu/c/cme/>
DOEC The Complete Corpus of Old English in Machine Readable Form, ed. An-
tonette di Paolo Healey, Dictionary of Old English Project, Toronto: Centre
for Medieval Studies, WWW-accessible version, University of Michigan
Humanities Text Initiative, 2000. < http://ets.umdl.umich.edu/o/oec/>
FLOB The FreiburgLOB Corpus of British English
<http://www.helsinki.f/varieng/CoRD/corpora/FLOB/
FROWN The FreiburgBrown corpus of American English
<http://www.helsinki.f/varieng/CoRD/corpora/FROWN>
grammis grammis das grammatische informationssystem des ids Mannheim:
Institut fr deutsche Sprache. <http://hypermedia.ids-mannheim.de/
grammis/>
HC The Helsinki Corpus of English Texts (1991) Helsinki: Department of Eng-
lish [see Kyt 1996] < http://helmer.aksis.uib.no/-icame/hc/index.htm>
LION Literature Online. <http://lion.chadwyck.com>
LLC The London Lund Corpus of Spoken English
<http://khnt.hit.uib.no/icame/manuals/LONDLUND/INDEX.HTM>
LOB The LancasterOslo/BergenCorpus of British English
<http://khnt.hit.uib.no/icame/manuals/lob/INDEX.HTM>
Project
Gutenberg <http://www.gutenberg.org>
2. Sources
Baker, Peter S. and Michael Lapidge (eds.)
1995 Byrthferths Enchiridion. (Early English Text Society. Supplementary
Series 15.) Oxford: Oxford University Press. [= ByrM]
Benson, Larry D. (ed.)
1987 The Riverside Chaucer. Oxford: Oxford University Press. [= Chaucer]
Blair, Hugh
[1783] 1965 Lectures on Rhetoric and Belles Lettres. Edited by Harold F. Hardings.
(Landmarks in Rhetoric and Public Address.) Carbondale: Southern Il-
linois University Press. First published London: Printed for W. Strahan,
T. Cadell and W. Creech in Edinburgh 1783.
252 References
Campbell, George
[1776] 1963 The Philosophy of Rhetoric. Edited by Lloyd F. Bitzer. Carbondale:
Southern Illinois University Press. First published London: Printed for
W. Strahan and T. Cadell and W. Creech at Edinburgh 1776.
Colville, George
[1556] 1897 Boethius: Boethius Consolation of Philosophy: Translated from the
Latin. Edited by Ernest B. Bax. (The Tudor Library 5.) London: David
Nutt.
Crotch, Walter J. B. (ed.)
1928 The Prologues and Epilogues of William Caxton. (Early English Text
Society. Original Series 176.) London: Oxford University Press.
Dedeck-Hry, Venceslas L.
1952 Boethius De Consolatione by Jean de Meun. Medieval Studies 14: 165
275.
Forshall, Josiah and Frederic Madden (eds.)
1850 The Holy Bible: Translated from the Latin Vulgate by John Wycliff and
his Followers. Oxford: Oxford University Press. Reprint: New York:
AMS Press, 1982.
Fox, Samuel
[1864] 2004 The Consolation of Philosophy: King Alfreds AngloSaxon Version of
Boethius Work London: H. G. Bohn. Reprint: Kessinger Publishings,
2004 (Book on Demand).
Furnivall, Frederick J. (ed.)
1922 Hali Meidenhad: An Alliterative Homily of the Thirteenth Century.
(Early English Text Society. Original Series 18.) Oxford: Oxford Uni-
versity Press.
Goolden, Peter (ed.)
1958 The Old English Apollonius of Tyre. (Oxford English Monographs.) Ox-
ford: Oxford University Press.
Grasmck, Ernst Ludwig (ed.)
1997 Boethius: Trost der Philosophie Zweisprachige Ausgabe. Aus dem
Lateinischen von Ernst Neitzke. Mit einem Vorwort von Ernst Ludwig
Grasmck. Frankfurt: Insel.
Green, Richard H. (ed. and transl.)
1962 The Consolation of Philosophy: Boethius. Mineola, New York: Dover
Publications.
Herrtage, Sidney J. H. (ed.)
1881 Catholicon Anglicum: An EnglishLatin Wordbook, dated 1483 From
the Ms no1968 in the Library of Lord Monson, collated with the ad
ditional Ms 15,562, British Museum: With Introduction and Notes
(Early English Text Society. Original Series 75.) London: Trbner.
Holthausen, Ferdinand (ed.)
1888 Vices and Virtues: Being a Souls Confession of its Sins with Reasons
Description of the Virtues: A MiddleEnglish Dialogue of About 1200
A D, Part I. Text and Translation. (Early English Text Society. Original
Series 89.) London: Oxford University Press.
Sources 253
Johnson, Samuel
[1747] 1970 The Plan of a Dictionary. (English Linguistics 15001800 223.) Men-
ston: Scolar Press. First published London: Printed for J. and P. Knap-
ton, T. Longman and T. Shewell, C. Hitch, A. Millar, and R. Dodsley
1747.
Johnson, Samuel
[1755] 1967 A Dictionary of the English Language: In which the Words are Deduced
from their Originals, and Illustrated in their Different Signifcations
by Examples from the Best Writers: To which are Prefxed a History of
the Language and an English Grammar. New York: AMS Press. First
published London: W. Strahan 1755.
Liuzza, Roy Michael (ed.)
1994 The Old English Version of the Gospels Volume 1: Text and Introduc
tion. (Early English Text Society. Original Series 304.) Oxford: Oxford
University Press.
Locke, John
[1690] 1975 An Essay Concerning Human Understanding. Edited and introduced by
Peter H. Nidditch. (The Clarendon Edition of the Works of John Locke.)
Oxford: Clarendon Press. First published London: Printed for Tho. Bas-
set and sold by Edw. Mory 1690.
Lowth, Robert
[1762] 1969 A Short Introduction to English Grammar. (English Linguistics 1500
1800 18.) Menston: Scolar Press. First published London: Printed by J.
Hughs for A. Millar and R. and J. Dodsley 1762.
Macaulay, George C. (ed.)
19001901 The English Works of John Gower. (Early English Text Society. Extra
Series 81, 82.) London: Oxford University Press. [= Gower]
Pemberton, Caroline (ed.)
1899 Queen Elizabeths Englishings of Boethius, De Consolatione Philoso
phiae A D 1593, Plutarch, De Curiositate, Horace, De Arte Poetica
(Part), A D 1598. (Early English Text Society. Original Series 113.)
London: Kegan Paul, Trench, Trbner.
Preston, Richard Lord Viscount
1695 Boethius: Anicius Manlius Severinus Boetius: Of the Consolation of
Philosophy: In Five Books: Made English and Illustrated with Notes.
London: Awnsham and John Churchill.
Scragg, Donald (ed.)
1992 The Vercelli Homilies and Related Texts. (Early English Text Society.
Original Series 300.) Oxford: Oxford University Press.
Sedgefeld, Walter J.
1899 King Alfreds Old English Version of Boethius De Consolatione Phi-
losophiae. Oxford: Clarendon Press.
Skeat, Walter H. (ed.)
1900 lfrics Lives of Saints: Being a Set of Sermons on Saints Days For
merly Observed by the English Church: From Manuscript Julius E VII
in the Cottonian Collection, with Various Readings from Other Manu
254 References
scripts, Volume 2. (Early English Text Society. Original Series 114.)
London: Oxford University Press.
Swanton, Michael (ed.)
1975 AngloSaxon Prose. (Everymans Library 809.) London: Dent.
Swift, Jonathan
1729 A Modest Proposal for Preventing the Children of Poor People from
Being a Burthen to their Parents or the Country, and for Making them
Benefcial to the Publick. Dublin: S. Harding (Project Gutenberg, E-Text
Nr.1080)
Thomson, David (ed.)
1984 An Edition of the Middle English Grammatical Texts. (Garland Medi-
eval Texts 8.) New York: Garland.
Watts, Victor E. (transl.)
1969 Boethius: The Consolation of Philosophy. (Penguin Classics.) London:
Penguin Books.
Zupitza, Julius (ed.)
2001 Aelfrics Grammatik und Glossar: Text und Varianten. Third reprint
with a new introduction by Helmut Gneuss. Hildesheim: Olms. First
published Berlin: Weidmann 1880.
3. Dictionaries
AND = Rothwell, William, Louise W. Stone and Thomas B. W. Reid (eds.)
1992 AngloNorman Dictionary. (Modern Humanities Research Association.
Publications 8.) London: Modern Humanities Research Association.
BT = Bosworth, Joseph and Thomas Northcote Toller (eds.)
1898 An AngloSaxon Dictionary: Based on the Manuscript Collections of
the Late Joseph Bosworth. Edited and enlarged by Thomas Northcote
Toller. Oxford: Clarendon.
CH = Clark Hall, John R.
1984 A Concise AngloSaxon Dictionary. Fourth edition with a supplement
by Herbert D. Meritt. Toronto: University of Toronto Press.
DOE = Dictionary of Old English in Electronic Form A F, Toronto: Centre for Medi-
eval Studies [CD ROM publication 2003].
EMEDD = Early Modern English Dictionaries Database, ed. Ian Lancashire. Toronto.
<[http://www.chass.utoronto.ca/english/emed/emedd.html> (accessed
2006; EMEDD has been replaced by Lexicons of Early Modern English
(LEME) at <leme.library.utoronto.ca>)
Hoad, Terry F. (ed.)
1986 The Concise Oxford Dictionary of English Etymology. Oxford: Claren-
don Press.
Holthausen, Ferdinand
1934 Altenglisches etymologisches Wrterbuch. (Germanische Bibliothek
4,7.) Heidelberg: Winter.
Secondary sources 255
Kluge/Seebold = Kluge, Friedrich
2002 Etymologisches Wrterbuch der deutschen Sprache. Bearbeitet von El-
mar Seebold, 24., durchgesehene und erweiterte Aufage. Berlin: Mou-
ton de Gruyter.
Mayhew, Anthony L. (ed.)
1908 The Promptorium Parvulorum: The First EnglishLatin Dictionary
Ed from the Manuscript in the Chapter Library at Winchester with
Introduction, Notes and Glossaries. (Early English Text Society. Extra
Series 102.) London: Kegan Paul, Trench, Trbner.
MED = Kurath, Hans, Sherman M. Kuhn, Robert E. Lewis et al (eds.)
19542001 Middle English Dictionary. Ann Arbor, Michigan: University of Michi-
gan Press. [online version: Middle English Compendium:
<http://ets.umdl.umich.edu/m/mec/>]
OED = A New English Dictionary on Historical Principles. Edited by James A. H.
Murray, H. Bradley, W. A. Craigie and C. T. Onions (Oxford 18881928).
Corrected Re-Reissue with an Introduction, Supplement and Bibliogra-
phy: The Oxford English Dictionary (1933). Second Edition, prepared
by John A. Simpson and Edmund S. C. Weiner, 20 vols. (Oxford: Oxford
University Press, 1989). Online version: <www.oed.com>.
Oxford DNB = Oxford Dictionary of National Biography
<http://www.oxforddnb.com>
4. Secondary sources
Adamson, Sylvia
1999 Literary language. In: Roger Lass (ed.), The Cambridge History of the
English Language. Volume III: 14761776, 539653. Cambridge: Cam-
bridge University Press.
Altenberg, Bengt
1984 Causal linking in spoken and written English. Studia Linguistica 38 (1):
2069.
Altenberg, Bengt
1986 Contrastive linking in spoken and written English. In: Gunnel Tottie and
Ingegerd Backland (eds.), English in Speech and Writing A Symposium,
1340. (Studia Anglistica Upsaliensia 60.) Stockholm: Almqvist and
Wiksell.
Altenberg, Bengt
1999 Adverbial connectors in English and Swedish: semantic and lexical cor-
respondences. In: Hilde Hasselgrd and Signe Oksefjell (eds.), Out of
Corpora: Studies in Honour of Stig Johansson, 249268. (Language
and Computers 26.) Amsterdam: Rodopi.
Andrew, Samuel O.
1940 Syntax and Style in Old English. New York: Russell & Russell.
Appenzeller-Gassmann, Verena
1961 Mittelenglische Bekrftigungsformeln. Zrich: Juris-Verlag.
256 References
Auwera, Johan van der (ed.)
1998 Adverbial Constructions in the Languages of Europe. (Empirical Ap-
proaches to Language Typology 20,3. Empirical Approaches to Lan-
guage Typology, EUROTYP 3.) Berlin/New York: Mouton de Gruyter.
Bailey, Richard W.
1996 NineteenthCentury English. Ann Arbor, Michigan: University of
Michigan Press.
Barber, Charles Laurence
1997 Early Modern English. Edinburgh: Edinburgh University Press.
Barth-Weingarten, Dagmar
2003 Concession in Spoken English: On the Realisation of a DiscoursePrag
matic Relation. (Language in Performance 28.) Tbingen: Narr.
Barth-Weingarten, Dagmar and Elizabeth Couper-Kuhlen
2002 On the development of fnal though: A case of grammaticalization? In:
Ilse Wischer and Gabriele Diewald (eds.), New Refections on Gram
maticalization, 345361. (Typological Studies in Language 49.) Amster-
dam: Benjamins.
Baugh, Albert Croll and Thomas Cable
2003 A History of the English Language. (ffth ed.) London: Routledge.
Beal, Joan C.
2004 English in Modern Times: 17001945. London: Arnold.
Bellert, Irena
1977 On semantic and distributional properties of sentential adverbs. Linguis
tic Inquiry 8: 337351.
Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad and Edward Finegan
1999 Longman Grammar of Spoken and Written English. Harlow: Longman.
Blake, Norman F.
1992 The Cambridge History of the English Language. Volume 2: 1066
1476. Cambridge: Cambridge University Press.
Blakemore, Diane
2002 Relevance and Linguistic Meaning: The Semantics and Pragmatics of
Discourse Markers. (Cambridge Studies in Linguistics 99.) Cambridge:
Cambridge University Press.
Blank, Andreas and Peter Koch (eds.)
1999 Historical Semantics and Cognition. (Cognitive Linguistics Research
13.) Berlin: Mouton de Gruyter.
Blockley, Mary
2001 Aspects of Old English Poetic Syntax: Where Clauses Begin. (Illinois
Medieval Studies.) Urbana: University of Illinois Press.
Bolkestein, Alide Machtelt
2000 Discourse organization and anaphora in Latin. In: Susan C. Herring,
Pieter van Reenen and Lene Schsler (eds.), Textual Parameters in Old
er Languages, 107137. (Amsterdam Studies in the Theory and History
of Linguistic Science Series IV. Current Issues in Linguistic Theory
195.) Amsterdam: Benjamins.
Secondary sources 257
Bowerman, Melissa
1986 First steps in acquiring conditionals: Some cognitive, semantic, formal
and pragmatic considerations. In: Elizabeth Closs Traugott, Alice ter
Meulen, Judy Snitzer Reilly and Charles A. Ferguson (eds.), On Condi
tionals, 258307. Cambridge: Cambridge University Press.
Breul, Carsten
1997 Grammatik und Bedeutung der kausalen Satzverbnde: Because, as,
since und for im schriftsprachlichen Englisch. (Linguistische Arbeiten
368.) Tbingen: Niemeyer.
Brinton, Laurel J.
1996 Pragmatic Markers in English: Grammaticalization and Discourse
Functions (Topics in English Linguistics 19.) Berlin/New York: Mou-
ton de Gruyter.
Brinton, Laurel J.
2006 Pathways in the development of pragmatic markers in English. In: Ans
van Kemenade and Bettelou Los (eds.), The Handbook of the History of
English, 307334. (Blackwell Handbooks in Linguistics.) Malden, Mas-
sachusetts: Blackwell.
Brinton, Laurel J. and Elizabeth Closs Traugott
2005 Lexicalization and Language Change. (Research Surveys in Linguis-
tics.) Cambridge: Cambridge University Press.
Brugmann, Karl
1904 Die Demonstrativpronomina der indogermanischen Sprachen: Eine
bedeutungsgeschichtliche Untersuchung. (Abhandlungen der Philolo-
gisch-Historischen Klasse der Kniglich Schsischen Gesellschaft der
Wissenschaften 22, 6.) Leipzig: Teubner.
Brunner, Karl
19601962 Die englische Sprache: Ihre geschichtliche Entwicklung. (Sammlung
kurzer Grammatiken germanischer Dialekte B, Ergnzungsreihe 6.) T-
bingen: Niemeyer.
Brunner, Karl
1965 Altenglische Grammatik: Nach der Angelschsischen Grammatik von
Eduard Sievers (Sammlung kurzer Grammatiken germanischer Dia-
lekte A, Hauptreihe 3.) Tbingen: Niemeyer.
Campbell, Alistair
1959 Old English Grammar. Oxford: Clarendon Press.
Chafe, Wallace
1982 Integration and involvement in speaking, writing and oral literature. In:
Deborah Tannen (ed.), Spoken and Written Language: Exploring Oral
ity and Literacy, 3553. (Advances in Discourse Processes 9.) Norwood,
New Jersey: Ablex Publishing Corporation.
Claridge, Claudia and Terry Walker
2002 Causal clauses in written and speech-related genres in Early Modern
English. ICAME Journal: Computers in English Linguistics 25: 3163.
Clemoes, Peter
1985 Language in context: Her in the 890 Anglo-Saxon Chronicle. Leeds
Studies in English 16: 2736.
258 References
Consten, Manfred
2004 Anaphorisch oder deiktisch? Zu einem integrativen Modell domnen
gebundener Referenz. (Linguistische Arbeiten 484.) Tbingen: Nie-
meyer.
Couper-Kuhlen, Elizabeth and Bernd Kortmann (eds.)
2000 Cause Condition Concession Contrast: Cognitive and Discourse
Perspectives. (Topics in English Linguistics 33.) Berlin: Mouton de
Gruyter.
Couper-Kuhlen, Elizabeth and Sandra A. Thompson
2000 Concessive patterns in conversation. In: Elizabeth Couper-Kuhlen and
Bernd Kortmann (eds.), Cause Condition Concession Contrast
Cognitive and Discourse Perspectives, 381410. (Topics in English Lin-
guistics 33.) Berlin: Mouton de Gruyter.
Croll, Morris William
1966 Style, Rhetoric, and Rhythm Essays by Morris W Croll. Edited by John
Max Patrick Princeton, New Jersey: Princeton University Press.
Culpeper, Jonathan and Merja Kyt
2000 The conjunction and in Early Modern English: Frequencies and uses
in speech-related writing and other texts. In: Ricardo Bermdez-Otero,
David Denison, Richard M. Hogg and Chris B. McCully (eds.), Genera
tive Theory and Corpus Studies A Dialogue from 10 ICEHL, 299326.
(Topics in English Linguistics 31.) Berlin: Mouton de Gruyter.
Davis, Norman, Douglas Gray, Patricia Ingham and Anne Wallace-Hadrill
1979 A Chaucer Glossary. Oxford: Clarendon Press.
De Smet, Hendrik
2005 A corpus of late Modern English texts. ICAME Journal Computers in
English Linguistics 29: 6982.
Denison, David
1993 English Historical Syntax Verbal Constructions. (Longman Linguis-
tics Library.) London: Longman.
Detges, Ulrich
1998 Echt die Wahrheit sagen: berlegungen zur Grammatikalisierung von
Adverbmarkern. Philologie im Netz 4: 125.
Detges, Ulrich and Richard Waltereit
2002 Grammaticalization vs. reanalysis: A semantic-pragmatic account of
functional change in grammar. Zeitschrift fr Sprachwissenschaft 21
(2): 151195.
Devriendt, Betty, Louis Goossens and Johan van der Auwera (eds.)
1996 Complex Structures A Functionalist Perspective. (Functional Gram-
mar Series 17.) Berlin: Mouton de Gruyter.
Diessel, Holger
1996 Processing factors of pre- and postposed adverbial clauses. In: Jan
Johnson, Matthew L. Juge and Jeri L. Moxley (eds.), Proceedings of
the TwentySecond Annual Meeting of the Berkeley Linguistics Society,
7182. Berkeley, California: Berkeley Linguistics Society.
Secondary sources 259
Diessel, Holger
2004 The Acquisition of Complex Sentences. (Cambridge Studies in Linguis-
tics 105.) Cambridge: Cambridge University Press.
Diessel, Holger
2005 Competing motivations for the ordering of main and adverbial clause.
Linguistics 43 (3): 449470.
Diewald, Gabriele
1997 Grammatikalisierung Eine Einfhrung in Form und Werden gramma
tischer Formen. (Germanistische Arbeitshefte 36.) Tbingen: Niemeyer.
Diewald, Gabriele
1999 Die Entwicklung der Modalpartikel aber: Ein typischer Grammatika-
lisierungsweg der Modalpartikeln. In: Hans Otto Spillmann and Ingo
Warnke (eds.), Internationale Tendenzen der Syntaktik, Semantik und
Pragmatik Akten des 32 Linguistischen Kolloquiums in Kassel 1997,
8391. (Linguistik International 1.) Frankfurt am Main: Lang.
Diller, Hans-Jrgen and Manfred Grlach (eds.)
2001 Towards a History of English as a History of Genres. (Anglistische
Forschungen 298.) Heidelberg: Winter.
Donner, Morton
1991 Adverb form in Middle English. English Studies 72 (1): 111.
Donoghue, Daniel and Bruce Mitchell
1992 Parataxis and hypotaxis: A review of some terms used for Old English
syntax. Neuphilologische Mitteilungen 93 (2): 163183.
Eitle, Hermann
1914 Die Satzverknpfung bei Chaucer. (Anglistische Forschungen 44.) Hei-
delberg: Winter.
Enkvist, Nils Erik and Brita Wrvik
1987 Old English a, temporal chains, and narrative structure. In: Anna Gia-
calone Ramat, Onofrio Carruba and Giuliano Bernini (eds.), Papers
from the Seventh International Conference on Historical Linguistics,
220237. (Amsterdam Studies in the Theory and History of Linguistic
Science 4, 48.) Amsterdam: Benjamins.
Fettig, Adolf
1934 Die Gradadverbien im Mittelenglischen. (Anglistische Forschungen
79.) Heidelberg: Winter.
Fischer, Olga
1992 Syntax. In: Norman Blake (ed.), The Cambridge History of the English
Language. Volume 2: 10661476, 207408. Cambridge: Cambridge
University Press.
Fischer, Olga, Ans van Kemenade, Willem Koopman and Wim van der Wurff
2000 The Syntax of Early English. (Cambridge Syntax Guides.) Cambridge:
Cambridge University Press.
Fischer, Olga
2007 Morphosyntactic Change Functional and Formal Perspectives. (Ox-
ford Surveys in Syntax and Morphology 2.) Oxford: Oxford University
Press.
260 References
Foucault, Michel
1966 Les Mots et les Choses Une Archologie des Sciences Humaines (Bib-
liothque des Sciences Humaines.) Paris: Gallimard [English: The Or
der of Things. New York: Vintage, 1973].
Franz, Wilhelm
1939 Die Sprache Shakespeares in Vers und Prosa: Unter Bercksichtigung
des Amerikanischen entwicklungsgeschichtlich dargestellt. Halle/Saa-
le: Niemeyer.
Fraser, Bruce
1999 What are discourse markers? Journal of Pragmatics 31 (7): 931952.
Fries, Udo
1994 Towards a description of text deixis in Old English. In: Klaus R. Grinda
and Claus-Dieter Wetzel (eds.), AngloSaxonica Beitrge zur Vor und
Frhgeschichte der englischen Sprache und zur altenglischen Litera
tur Festschrift fr Hans Schabram zum 65 Geburtstag, 527540. Mn-
chen: Wilhelm Fink.
Fugier, Huguette
1987 Les connecteurs de de cause en latin. Modles Linguistiques 9: 917.
Gil, Alberto
1995 Textadverbiale in den romanischen Sprachen: Eine integrale Studie zu
Konnektoren und Modalisatoren im Spanischen, Franzsischen und
Italienischen. (Bonner Romanistische Arbeiten 53.) Frankfurt: Lang.
Givn, Talmy
2001 Syntax An Introduction, Volume 2. Amsterdam: Benjamins.
Gneuss, Helmut
1996 English Language Scholarship A Survey and Bibliography from the
Beginnings to the End of the Nineteenth Century. (Medieval and Re-
naissance Texts and Studies 125.) Binghamton, New York: Center for
Medieval and Renaissance Texts and Studies.
Godden, Malcolm
2000 lfrics Catholic Homilies Introduction, Commentary and Glossary.
(Early English Text Society. Supplementary Series 18.) Oxford: Oxford
University Press.
Godden, Malcolm
2007 Did King Alfred write anything? Medium vum 76: 123.
Grlach, Manfred
1986 Middle English a creole? In: Dieter Kastovsky and Alexander Szwe-
dek (eds.), Linguistics across Historical and Geographical Boundaries
In Honour of Jacek Fisiak on the Occasion of his Fiftieth Birthday I:
Linguistic Theory and Historical Linguistics, 329344. (Trends in Lin-
guistics. Studies and Monographs 32, 1.) Berlin: Mouton de Gruyter.
Grlach, Manfred
1999 English in NineteenthCentury England An Introduction. Cambridge:
Cambridge University Press.
Grlach, Manfred
2001 EighteenthCentury English. (Sprachwissenschaftliche Studienbcher.)
Heidelberg: Winter.
Secondary sources 261
Grlach, Manfred
2004 Text Types and the History of English. (Trends in Linguistics. Studies
and Monographs 139.) Berlin: Mouton de Gruyter.
Gonzlez-Cruz, Ana I.
2007 On the subjectifcation of adverbial clause connectives: Semantic and
pragmatic considerations in the development of while-clauses. In: Ursu-
la Lenker and Anneli Meurman-Solin (eds.), Connectives in the History
of English, 145166. (Amsterdam Studies in the Theory and History of
Linguistic Science Series IV. Current Issues in Linguistic Theory 283.)
Amsterdam: Benjamins.
Greenbaum, Sidney
1969a Studies in English Adverbial Usage. London: Longman.
Greenbaum, Sidney
1969b The question of but. Folia Linguistica 3: 245254.
Grice, Herbert Paul
1989 Studies in the Ways of Words. Cambridge, Massachusetts: Harvard Uni-
versity Press.
Guimier, Claude
1985 On the origin of the suffx -ly. In: Jacek Fisiak (ed.), Historical Seman
tics Historical WordFormation, 155170. (Trends in Linguistics. Stud-
ies and Monographs 29.) Berlin: Mouton de Gruyter.
Guimier, Claude
1991 Peut-on dfnir ladverbe? In: Claude Guimier and Pierre Larcher (eds.),
Les tats de lAdverbe Travaux Linguistiques du CERLICO, 1134.
(Cercle Linguistique du Centre et de lOuest. Travaux linguistiques du
CERLICO 3.) Rennes: Presses Universitaires de Rennes.
Gnthner, Susanne
2000 From concessive connector to discourse marker: The use of obwohl in
everyday German interaction. In: Elizabeth Couper-Kuhlen and Bernd
Kortmann (eds.), Cause Condition Concession Contrast Cognitive
and Discourse Perspectives, 439468. (Topics in English Linguistics
33.) Berlin: Mouton de Gruyter.
Haiman, John and Sandra A. Thompson (eds.)
1988 Clause Combining in Grammar and Discourse. (Typological Studies in
Language 18.) Amsterdam: Benjamins.
Halliday, Michael A. K. and Ruqaiya Hasan
1976 Cohesion in English. (English Language Series 9.) London: Longman.
Halliday, Michael A. K.
1993 On the language of physical science. In: Michael A. K. Halliday and
James R. Martin (eds.), Writing Science: Literacy and Discursive
Power, 5468. (Critical Perspectives on Literacy and Education.) Lon-
don: Falmer.
Heine, Bernd, Ulrike Claudi and Friederike Hnnemeyer
1991 Grammaticalization A Conceptual Framework. Chicago: University of
Chicago Press.
262 References
Herman, Jszef
1963 La Formation du Systme Roman des Conjonctions de Subordination.
(Deutsche Akademie der Wissenschaften zu Berlin. Verffentlichungen
des Instituts fr Romanische Sprachwissenschaft 18.) Berlin: Akade-
mie-Verlag.
Heuer, Hermann
1932 Studien zur syntaktischen und stilistischen Funktion des Adverbs bei
Chaucer und im Rosenroman. (Anglistische Forschungen 75.) Heidel-
berg: Winter.
Higashiizumi, Yuko
2006 From a Subordinate Clause to an Independent Clause A History of
English because-clause and Japanese kara-clause. (Hituzi Linguistics
in English 2.) Tokyo: Hituzi Syobo Publishing.
Hogg, Richard M. and David Denison (eds.)
2006 A History of the English Language. Cambridge: Cambridge University
Press.
Hogg, Richard M.
1992 The Cambridge History of the English Language. Volume I: The Begin
nings to 1066. Cambridge: Cambridge University Press.
Hopper, Paul J. and Elizabeth Closs Traugott
2003 Grammaticalization. (Cambridge Textbooks in Linguistics.) Cam-
bridge: Cambridge University Press.
Huddleston, Rodney and Geoffrey K. Pullum
2002 The Cambridge Grammar of the English Language. Cambridge: Cam-
bridge University Press.
Hllen, Werner
1989 Their Manner of Discourse Nachdenken ber Sprache im Umkreis
der Royal Society. Tbingen: Narr.
Itani, Reiko
1992 Japanese conjunction kedo (but) in utterance-fnal use: A rele-
vance-based analysis. English Linguistics 9: 265283.
Jones, Charles
1989 A History of English Phonology. (Longman Linguistics Library.) Lon-
don: Longman.
Jones, Richard Foster
1951 The Seventeenth Century Essays by Richard Foster Jones and Others
Writing in his Honour. Palo Alto: Stanford University Press.
Jucker, Andreas
1991 Between hypotaxis and parataxis: Clauses of reason in Ancrene Wisse.
In: Dieter Kastovsky (ed.), Historical English Syntax, 203220. (Topics
in English Linguistics 2.) Berlin: Mouton de Gruyter.
Jucker, Andreas
1997 The discourse marker well in the history of English. English Language
and Linguistics 1 (1): 91110.
Keller, Rudi
1993 Das epistemische weil: Bedeutungswandel einer Konjunktion. In: Hans
Jrgen Heringer and Georg Sttzel (eds.), Sprachgeschichte und Sprach
Secondary sources 263
kritik Festschrift fr Peter von Polenz zum 65 Geburtstag, 219247.
Berlin: Mouton de Gruyter.
Kemenade, Ans van and Bettelou Los (eds.)
2006a The Handbook of the History of English. (Blackwell Handbooks in Lin-
guistics.) Malden, Massachusetts/Oxford: Blackwell.
Kemenade, Ans van and Bettelou Los
2006b Discourse adverbs and clausal syntax in Old and Middle English. In:
Ans van Kemenade and Bettelou Los (eds.), The Handbook of the His
tory of English, 224248. (Blackwell Handbooks in Linguistics.) Mal-
den, Massachusetts/Oxford: Blackwell.
Kerkhof, Jelle
1982 Studies in the Language of Geoffrey Chaucer. (Leidse germanistische
en anglistische reeks 5.) Leiden: Brill.
Kim, Taejin
1992 The Particle a in the West Saxon Gospels A Discourse Level Analysis.
(Europische Hochschulschriften 14, 249.) Bern: Lang.
Kisbye, Torben
197172 An Historical Outline of English Syntax, 2 Volumes. Aarhus: Aka-
demisk Boghandel.
Kivimaa, Kirsti
1966 e and at as Clause Connectives in Early Middle English with Es
pecial Consideration of the Emergence of Pleonastic at. (Commen-
tationes Humanarum Litterarum 39, 1.) Helsinki: Societas Scientiarum
Fennica.
Klare, Johannes
1958 Entstehung und Entwicklung der konzessiven Konjunktionen im Fran
zsischen. (Verffentlichungen, Institut fr Romanische Sprachwissen-
schaft, Berlin 13.) Berlin: Akademie-Verlag.
Koch, Peter and Wulf Oesterreicher
1985 Sprache der Nhe Sprache der Distanz: Mndlichkeit und Schriftlich-
keit im Spannungsfeld von Sprachtheorie und Sprachgeschichte. Roma
nistisches Jahrbuch 36: 1541.
Kohnen, Thomas
2004 Text Textsorte Sprachgeschichte: Englische Partizipial und Gerun
dialkonstruktionen 1100 bis 1700. (Buchreihe der Anglia 37.) Tbingen:
Niemeyer.
Kohnen, Thomas
2007 Connective profles in the history of English texts: Aspects of orality
and literacy. In: Ursula Lenker and Anneli Meurman-Solin (eds.), Con
nectives in the History of English, 289306. (Amsterdam Studies in the
Theory and History of Linguistic Science Series IV. Current Issues in
Linguistic Theory 283.) Amsterdam: Benjamins.
Knig, Ekkehard
1985a On the history of concessive connectives in English: Diachronic and
synchronic evidence. Lingua 66: 119.
264 References
Knig, Ekkehard
1985b Where do concessives come from? On the development of concessive
connectives. In: Jacek Fisiak (ed.), Historical Semantics and Historical
Word Formation, 263282. (Trends in Linguistics. Studies and Mono-
graphs 29.) Berlin: Mouton de Gruyter.
Knig, Ekkehard
1988 Concessive connectives and concessive sentences: Cross-linguistic regu-
larities and pragmatic principles. In: John A. Hawkins (ed.), Explaining
Language Universals, 145166. Oxford: Blackwell.
Knig, Ekkehard
2009 Review of Lenker and Meurman-Solin 2007. English Language and
Linguistics 13:147152.
Knig, Ekkehard and Elizabeth Traugott
1982 Divergence and apparent convergence in the development of yet and
still. In: Monica Macaulay and Orin D. Gensler et al. (eds.), Proceed
ings of the Eighth Annual Meeting of the Berkeley Linguistics Society,
170179. Berkeley: Berkeley Linguistics Society.
Kortmann, Bernd
1997 Adverbial Subordination A Typology and History of Adverbial Sub
ordinators Based on European Languages. (Empirical Approaches to
Language Typology 18.) Berlin: Mouton de Gruyter [Material in: Kort-
mann, Bernd 1997b A CrossLinguistic Dictionary of Adverbial
Subordinators. Unterschleiheim: Lincom].
Krefeld, Thomas
1989 Unterordnung Beiordnung Zuordnung: Was ist romanisch an franz-
sisch que? In: Wolfgang Raible (ed.), Romanistik, Sprachtypologie und
Universalienforschung: Beitrge zum Freiburger Romanistentag 1987,
1136. (Tbinger Beitrge zur Linguistik 332.) Tbingen: Narr.
Krefeld, Thomas
1999 Agens mit Leib und Seele: Zur Grammatikalisierung romanischer Ad-
verbbildungen. In: Jrgen Lang and Ingrid Neumann-Holzschuh (eds.),
Reanalyse und Grammatikalisierung in den romanischen Sprachen,
111127. (Linguistische Arbeiten 410.) Tbingen: Niemeyer.
Kroon, Caroline
1995 Discourse Particles in Latin A Study of nam, enim, autem, vero and at.
(Amsterdam Studies in Classical Philology 4.) Amsterdam: Gieben.
Kryk-Kastovsky, Barbara
1997 From temporal adverbs to discourse particles: An instance of cross-lin-
guistic grammaticalization? In: Terttu Nevalainen and Leena Kahlas-
Tarkka (eds.), To Explain the Present Studies in the Changing English
Language in Honour of Matti Rissanen, 319328. (Uusflologinen Yh-
distys: Mmoires de la Socit Nophilologique de Helsinki 52.) Hel-
sinki: Socit Nophilologique.
Khner, Raphael and Carl Stegmann
19121914 Ausfhrliche Grammatik der lateinischen Sprache II: Satzlehre, Band
2. Hannover: Hahnsche Buchhandlung.
Secondary sources 265
Kyt, Merja
1996 Manual to the Diachronic Part of the Helsinki Corpus of English
Texts Coding Conventions and Lists of Source Texts. (International
Computer Archive of Modern English. ICAME Collection of English
Language Corpora 5.) Helsinki: Department of English, University of
Helsinki.
Kyt, Merja and Terry M. Walker
2006 Guide to a Corpus of English Dialogues 15601760. (Acta Universitatis
Upsaliensis: Studia Anglistica Upsaliensia 130.) Uppsala: Acta Univer-
sitatis Upsaliensis.
Lakoff, George and Mark Johnson
1980 Metaphors we Live by. Chicago: University of Chicago Press.
Landau, Sidney I.
1984 Dictionaries The Art and Craft of Lexicography. New York: Scribner.
Lang, Ewald
2000 Adversative connectors on distinct levels of discourse: A re-examination
of Eve Sweetsers three-level approach. In: Elizabeth Couper-Kuhlen
and Bernd Kortmann (eds.), Cause Condition Concession Con
trast: Cognitive and Discourse Perspectives, 235256. (Topics in Eng-
lish Linguistics 33.) Berlin: Mouton de Gruyter.
Langacker, Ronald
1999 Losing control: Grammaticalization, subjectifcation and transparency.
In: Andreas Blank and Peter Koch (eds.), Historical Semantics and Cog
nition, 147175. (Cognitive Linguistics Research 13.) Berlin: Mouton de
Gruyter.
Lass, Roger (ed.)
1999 The Cambridge History of the English Language. Volume 3: 14761776.
Cambridge: Cambridge University Press.
Law, Vivien
1997 Grammar and Grammarians in the Early Middle Ages. (Longman Lin-
guistics Library.) London: Longman.
Lefvre, Michel
2000 Subordination in Syntax, Semantik und Textlinguistik. (Eurogermanistik
15.) Tbingen: Stauffenburg.
Lehmann, Christian [1982]
1995 Thoughts on Grammaticalization. Unterschleiheim: LINCOM Europa
(originally published as: Thoughts on Grammaticalization A Program
matic Sketch, vol. I. University of Cologne: Arbeiten des Klner Univer-
salienprojekts).
Lehmann, Christian
1988 Towards a typology of clause linkage. In: John Haiman and Sandra A.
Thompson (eds.), Clause Combining in Grammar and Discourse, 181
225. (Typological Studies in Language 18.) Amsterdam: Benjamins.
Lehmann, Christian
2002 New refections on grammaticalization and lexicalization. In: Ilse
Wischer and Gabriele Diewald (eds.), New Refections on Grammati
266 References
calization, 118. (Typological Studies in Language 49.) Amsterdam:
Benjamins.
Leisi, Ernst and Christian Mair
1999 Das heutige Englisch Wesenszge und Probleme. (8th ed.) (Sprachwis-
senschaftliche Studienbcher.) Heidelberg: Winter.
Lenk, Uta
1998 Marking Discourse Coherence: Functions of Discourse Markers in
Spoken English. (Language in Performance 15.) Tbingen: Narr.
Lenker, Ursula
1997 Die westschsische Evangelienversion und die Perikopenordnungen im
angelschsischen England. (Texte und Untersuchungen zur Englischen
Philologie 20.) Mnchen: Fink.
Lenker, Ursula
2000 Solice and witodlice: Discourse markers in Old English. In: Olga
Fischer, Anette Rosenbach and Dieter Stein (eds.), Pathways of Change
Grammaticalization in English, 229249. (Studies in Language Com-
panion Series 53.) Amsterdam: Benjamins.
Lenker, Ursula
2002 Is it, stylewise or otherwise, wise to use -wise? Domain adverbs and the
history of English -wise. In: Teresa Fanego, Mara Jos Lpez-Couso
and Javier Prez-Guerra (eds.), English Historical Syntax and Morpho
logy Selected Papers from the Eleventh International Conference on
English Historical Linguistics, Santiago de Compostela, 711 Septem
ber 2000, 157180. (Amsterdam Studies in the Theory and History of
Linguistic Science Series IV. Current Issues in Linguistic Theory 223.)
Amsterdam/Philadelphia: Benjamins.
Lenker, Ursula
2003 Forsooth, a source: Metalinguistic thought in Early English. In: Lucia
Kornexl and Ursula Lenker (eds.), Bookmarks from the Past Studies in
Early English Language and Literature in Honour of Helmut Gneuss,
262288. (Texte und Untersuchungen zur Englischen Philologie 30.)
Frankfurt: Lang.
Lenker, Ursula
2004 Review of Traugott and Dasher 2002. Word 55: 262265.
Lenker, Ursula
2007a Forhwi because: Shifting deictics in the history of English causal con-
nection. In: Ursula Lenker and Anneli Meurman-Solin (eds.), Connec
tives in the History of English, 193227. (Amsterdam Studies in the
Theory and History of Linguistic Science Series IV. Current Issues in
Linguistic Theory 283.) Amsterdam: Benjamins.
Lenker, Ursula
2007b Solice, forsoothe, truly Communicative principles and invited infer-
ences in the history of truth-intensifying adverbs in English. In: Susan
Fitzmaurice and Irma Taavitsainen (eds.), Methods in Historical Prag
matics, 81105. (Topics in English Linguistics 52.) Berlin: Mouton de
Gruyter.
Secondary sources 267
Lenker, Ursula
2008 Booster prefxes in Old English an alternative view of the roots of ME
forsooth. English Language and Linguistics 12: 245265.
Lenker, Ursula and Anneli Meurman-Solin (eds.)
2007 Connectives in the History of English (Amsterdam Studies in the The-
ory and History of Linguistic Science Series IV. Current Issues in Lin-
guistic Theory 283.) Amsterdam: Benjamins.
Lenz, Friedrich
1997 Diskursdeixis im Englischen: Sprachtheoretische berlegungen und
lexikogrammatische Analysen. (Linguistische Arbeiten 369.) Tbin-
gen: Niemeyer.
Levinson, Stephen
1983 Pragmatics. (Cambridge Textbooks in Linguistics.) Cambridge: Cam-
bridge University Press.
Levinson, Stephen
2000 Presumptive Meanings The Theory of Generalized Conversational Im
plicature. (Language, Speech, and Communication.) Cambridge, Mas-
sachusetts: MIT Press.
Levinson, Stephen
2003 Space in Language and Cognition: Explorations in Cognitive Diversity.
(Language, Culture and Cognition 5.) Cambridge: Cambridge Univer-
sity Press.
Liggins, Elizabeth M.
1955 The expression of causal relationship in Old English prose. Unpublished
dissertation, University of London.
Lutz, Angelika
2002 When did English begin? In: Teresa Fanego, Beln Mndez-Naya and
Elena Seoane (eds.), Sounds, Words, Texts and Change Selected Pa
pers from 11 ICEHL, Santiago de Compostela, 711 September 2000,
145171. (Amsterdam Studies in the Theory and History of Linguistic
Science Series IV. Current Issues in Linguistic Theory 224.) Amster-
dam: Benjamins.
Lutzky, Ursula
2006 Discourse markers? Well, delimiting the basic features of discourse
markers. Vienna English Working Papers 15: 324.
Machan, Tim William
1985 Techniques of Translation Chaucers Boece. Norman, Oklahoma: Pil-
grim Books.
Mahler, Andreas
1996 Die Materialitt der Transparenz: Sprache, Politik und Literatur in der
englischen Aufklrung. In: Klaus Garber and Heinz Wismann (eds.),
Europische Soziettsbewegung und demokratische Tradition Die
europischen Akademien der Frhen Neuzeit zwischen Frhrenais
sance und Sptaufklrung, 721754. Tbingen: Niemeyer.
Mahler, Andreas
1997 Nominalism and Literary Discourse New Perspectives, 251268. (Crit-
ical studies 10.) Amsterdam: Rodopi.
268 References
Marchand, Hans
1969 The Categories and Types of PresentDay English WordFormation A
SynchronicDiachronic Approach. (Handbcher fr das Studium der
Anglistik.) Mnchen: Beck.
Markus, Manfred
1996 Chaucers prose in the Canterbury Tales as parody. In: Jrgen Klein and
Dirk Vanderbeke (eds.), Anglistentag 1995 Greifswald Proceedings,
259272. (Anglistentag: Proceedings of the Conference of the German
Association of University Teachers of English 17.) Tbingen: Niemeyer.
Markus, Manfred
2000 Wherefore therefore: Causal connectives in Middle English prose as
opposed to Present-Day English. 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,
215232. (Language and Computers 33.) Amsterdam: Rodopi.
Mauranen, Anna
1993 Cultural Differences in Academic Rhetoric: A Textlinguistic Study.
(Nordeuropische Beitrge aus den Human- und Gesellschaftswissen-
schaften 4.) Frankfurt: Lang.
McIntosh, Angus
1991 Old English adjectives with derivative -lic partners: Some semantic
problems. Neuphilologische Mitteilungen 92 (3): 297310.
McIntosh, Carey
1998 The Evolution of English Prose, 17001800 Style, Politeness and Print
Culture. Cambridge: Cambridge University Press.
Meurman-Solin, Anneli
2004 Towards a variationist typology of clausal connectives: Methodologi-
cal considerations based on the corpus of Scottish correspondence. In:
Marina Dossena and Roger Lass (eds.), Methods and Data in English
Historical Dialectology, 171197. (Linguistic Insights 16.) Bern: Lang.
Mey, Jacob
2001 Pragmatics An Introduction. Oxford: Blackwell.
Michael, Ian
1970 English Grammatical Categories and the Tradition to 1800. Cam-
bridge: Cambridge University Press.
Michelsen, Martina
1993 Weg vom Wort zum Gedankenstrich Zur stilistischen Funktion eines
Satzzeichens in der englischen Literatur des 17 und 18 Jahrhunderts
(Mnchner Studien zur neueren englischen Literatur 6.) Mnchen: Fink.
Milic, Louis Tonko
1967 A Quantitative Approach to the Style of Jonathan Swift. (Studies in
English Literature 23.) The Hague: Mouton.
Mitchell, Bruce
1985 Old English Syntax, 2 Volumes. Oxford: Clarendon.
Secondary sources 269
Mitchell, Bruce and Fred C. Robinson
2001 A Guide to Old English. (6th ed.) Oxford: Blackwell.
Moessner, Lilo
2006 The subjunctive in Early Modern English adverbial clauses. In: Chris-
tian Mair, Reinhard Heuberger and Josef Wallmannsberger (eds.), Cor
pora and the History of English Papers Dedicated to Manfred Markus
on the Occasion of his SixtyFifth Birthday, 249263. (Anglistische
Forschungen 363.) Heidelberg: Winter.
Molencki, Rafal
1997a Albeit a conjunction, yet it is a clause: A counter-example to unidirec-
tionality hypothesis? Studia Anglica Posnaniensia An International
Review of English Studies 31: 163178.
Molencki, Rafal
1997b Concessive clauses in Chaucers prose. In: Jacek Fisiak (ed.), Studies
in Middle English Linguistics, 351371. (Trends in Linguistics. Studies
and Monographs 103.) Berlin: Mouton de Gruyter.
Morrow, Philipp R.
1989 Conjunct use in business news stories and academic journal articles: A
comparative study. English for Specifc Purposes 8: 239254.
Moss, Fernand
1986 Mittelenglische Kurzgrammatik Lautlehre, Formenlehre, Syntax.
(Hueber Hochschulreihe 11.) Mnchen: Hueber.
Mueller, Janel L.
1984 The Native Tongue and the Word Developments in English Prose Style
13801580. Chicago: University of Chicago Press.
Mugglestone, Lynda (ed.)
2006 The Oxford History of English. Oxford: Oxford University Press.
Mustanoja, Tauno F.
1960 A Middle English Syntax, Part I. Helsinki: Socit Nophilologique.
Nate, Richard
2001 Wissenschaft und Literatur im England der frhen Neuzeit. (Figuren 9.)
Mnchen: Fink.
Nevalainen, Terttu
1991 BUT, ONLY, JUST: Focusing Adverbial Change in Modern English
15001900. (Uusflologinen Yhdistys: Mmoires de la Socit Nophi-
lologique de Helsinki 5.) Helsinki: Socit Nophilologique.
Nevalainen, Terttu
1997 The process of adverb derivation in late Middle and Early Modern Eng-
lish. In: Matti Rissanen, Merja Kyt and Kirsi Heikkonen (eds.), Gram
maticalization at Work: Studies of LongTerm Developments in English,
145189. (Topics in English Linguistics 24.) Berlin: Mouton de Gruyter.
Ong, Walter
1982 Orality and Literacy The Technologizing of the World. (New Accents.)
London: Methuen.
sterman, Aune
1997 There compounds in the history of English. In: Matti Rissanen, Merja
Kyt and Kirsi Heikkonen (eds.), Grammaticalization at Work Stud
270 References
ies of LongTerm Developments in English, 191276. (Topics in English
Linguistics 24.) Berlin: Mouton de Gruyter.
Parkes, Malcolm B.
1992 Pause and Effect An Introduction to the History of Punctuation in the
West. Aldershot: Scolar Press.
Pasch, Renate, Ursula Braue, Eva Breindl and Ulrich Hermann Waner
2003 Handbuch der deutschen Konnektoren Linguistische Grundlagen der
Beschreibung und syntaktische Merkmale der deutschen Satzverknp
fer (Konjunktionen, Satzadverbien und Partikeln). (Schriften des Insti-
tuts fr Deutsche Sprache 9.) Berlin: Mouton de Gruyter.
Prez Quintero, Mara Jess
2002 Adverbial Subordination in English A Functional Approach. (Lan-
guage and Computers 41.) Amsterdam: Rodopi.
Peters, Pam
2006 The Cambridge Guide to English Usage. Cambridge: Cambridge Uni-
versity Press.
Pinkster, Harm
1972 On Latin Adverbs. (North-Holland Linguistic Series 6.) Amsterdam:
North-Holland Publishing Company.
Pittner, Karin
1999 Adverbiale im Deutschen Untersuchungen zu ihrer Stellung und Interpre
tation. (Studien zur deutschen Grammatik 60.) Tbingen: Stauffenburg.
Poldauf, Ivan
1948 On the History of Some Problems of English Grammar before 1800.
(Pspvky k djinm ei a literatury anglick Sv.7. Prce z vdeckch
stavu 55.) Prague: Prague University Press.
Porter, David W. (ed.)
2002 Excerptiones de Prisciano The Source for lfrics LatinOld Eng
lish Grammar. (Anglo-Saxon Texts 4.) Woodbridge: Brewer.
Quirk, Randolph
1954 The Concessive Relation in Old English Poetry. (Yale Studies in Eng-
lish 124.) New Haven: Yale University Press.
Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech and Jan Svartvik
1985 A Comprehensive Grammar of the English Language. London: Long-
man.
Quirk, Randolph and Charles L. Wrenn
1957 An Old English Grammar. (Methuens Old English Library.) New York:
Holt, Rinehart & Winston.
Raible, Wolfgang
1992 Junktion Eine Dimension der Sprache und ihre Realisierungsformen
zwischen Aggregation und Integration, vorgetragen am 4 Juli 1987.
(Sitzungsberichte der Heidelberger Akademie der Wissenschaften. Phi-
losophisch-Historische Klasse 1992/2.) Heidelberg: Winter.
Ramat, Paolo and Davide Ricca
1998 Sentence adverbs in the languages of Europe. In: Johan van der Auwera
in Collaboration with Dnall P. Baoill (eds.), Adverbial Constructions
Secondary sources 271
in the Languages of Europe, 187275. (Empirical Approaches to Lan-
guage Typology 20, 3. EUROTYP 3.) Berlin: Mouton de Gruyter.
Rissanen, Matti
1997 Towards an integrated view of the development of English: Notes on
causal linking. In: Jacek Fisiak and Marcin Krygier (eds.), Advances in
English Historical Linguistics (1996), 389406. (Trends in Linguistics.
Studies and Monographs 112.) Berlin/New York: Mouton de Gruyter.
Rissanen, Matti
1999a On the adverbialisation of rather: Surfng for historical data. In: Hilde
Hasselgrd and Signe Oksefjell (eds.), Out of Corpora Studies in Hon
our of Stig Johansson, 4959. (Language and Computers 26.) Amster-
dam: Rodopi.
Rissanen, Matti
1999b Syntax. In: Roger Lass (ed.), The Cambridge History of the English
Language. Volume 3: Early Modern English 14761776, 187331.
Cambridge: Cambridge University Press.
Rissanen, Matti
2006 Latin infuence on an Old English idiom: to Wit. In: John Walmsley
(ed.), Inside Old English Essays in Honour of Bruce Mitchell, 222241.
Malden, Massachusetts: Blackwell.
Rissanen, Matti, Merja Kyt and Minna Palander-Collin
1993 Early English in the Computer Age Explorations through the Helsinki
Corpus. (Topics in English Linguistics 11.) Berlin: Mouton de Gruyter.
Robinson, Ian
1998 The Establishment of Modern English Prose in the Reformation and the
Enlightenment. Cambridge: Cambridge University Press.
Robinson, Orrin W.
1997 Clause Subordination and Verb Placement in the Old High German
Isidor Translation. (Germanische Bibliothek/Neue Folge 3, 26.) Heidel-
berg: Winter.
Romaine, Suzanne
1998 The Cambridge History of the English Language. Volume 4: 17761997.
Cambridge: Cambridge University Press.
Rudolph, Elisabeth
1989 The role of conjunctions and particles for text connexity. In: Maria-Elisa-
beth Conte, Jnos S. Petf and Emel Szer (eds.), Text and Discourse
Connectedness Proceedings of the Conference on Connexity and Co
herence, Urbino, July 1621, 1984, 175191. (Studies in Language.
Companion Series 16.) Amsterdam: Benjamins.
Rudolph, Elisabeth
1996 Contrast Adversative and Concessive Relations and their Expressions
in English, German, Spanish, Portuguese on Sentence and Text Level.
(Research in Text Theory 23.) Berlin/New York: Mouton de Gruyter.
Salmon, Vivian
1988 English punctuation theory 15001800. Anglia 106 (3): 285314.
272 References
Samuels, Michael
1972 Linguistic Evolution with Special Reference to English (Cambridge
Studies in Linguistics 5.) Cambridge: Cambridge University Press.
Sauer, Hans
2006 Adverbs and adverbials in the earliest English text (pinal-Erfurt). In:
Andrew James Johnston, Ferdinand von Mengden and Stefan Thim
(eds.), Language and Text Current Perspectives on English and Ger
manic Historical Linguistics and Philology, 255268. (Anglistische
Forschungen 359.) Heidelberg: Winter.
Scheler, Manfred
1977 Der englische Wortschatz. (Grundlagen der Anglistik und Amerikanistik
9.) Berlin: Schmidt.
Schiffrin, Deborah
1987 Discourse Markers. (Studies in Interactional Sociolinguistics 5.) Cam-
bridge: Cambridge University Press.
Schleburg, Florian
2002 Altenglisch swa: Syntax und Semantik einer polyfunktionalen Partikel.
(Sprachwissenschaftliche Studienbcher.) Heidelberg: Winter.
Schmid, Hans-Jrg
2005 Englische Morphologie und Wortbildung: Eine Einfhrung. (Grundla-
gen der Anglistik und Amerikanistik 25.) Berlin: Schmidt.
Scragg, Donald
1974 A History of English Spelling. (Mont Follick Series 3.) Manchester:
Manchester University Press.
Sorva, Elina
2007 From an analytic phrase to an atomic connective: The grammatical-
ization of Albeit. In: Ursula Lenker and Anneli Meurman-Solin (eds.),
Connectives in the History of English, 115143. (Amsterdam Studies in
the Theory and History of Linguistic Science Series IV. Current Issues
in Linguistic Theory 283.) Amsterdam: Benjamins.
Sprat, Thomas
1959 History of the Royal Society. Edited with Critical Apparatus by Jackson
J. Cope and Harold W. Jones. London: Routledge. First published Lon-
don 1667.
Sperber, Dan and Deidre Wilson
1996 Relevance Communication and Cognition. Oxford: Blackwell.
Stanley, Eric G.
2000 Hwt. In: Jane Roberts and Janet Nelson (eds.), Essays on AngloSaxon
and Related Themes in Memory of Lynne Grundy, 525556. (Kings
College London Medieval Studies 17.) London: Kings College, Centre
for Late Antique and Medieval Studies.
Stempel, Wolf-Dieter
1964 Untersuchungen zur Satzverknpfung im Altfranzsischen. (Archiv fr
das Studium der neueren Sprachen und Literaturen: Beihefte 1.) Braun-
schweig: Westermann.
Secondary sources 273
Stockwell, Robert P. and Donka Minkova
1991 Subordination and word order change in the history of English. In: Die-
ter Kastovsky (ed.), Historical English Syntax, 367408. (Topics in Eng-
lish Linguistics 2.) Berlin: Mouton de Gruyter.
Stoffel, Cornelis
1901 Intensives and DownToners A Study in English Adverbs (Anglistische
Forschungen 1.) Heidelberg: Winter.
Swan, Toril
1988 The development of sentence adverbs in English. Studia Linguistica 42
(1): 117.
Swan, Toril
1988 Sentence Adverbials in English: A Synchronic and Diachronic Investi
gation. (Troms-studier i sprkvitenskap 10.) Oslo: Novus.
Sweetser, Eve
1990 From Etymology to Pragmatics Metaphorical and Cultural Aspects of
Semantic Structure. (Cambridge Studies in Linguistics 54.) Cambridge:
Cambridge University Press.
Thim, Stefan
2006 Phrasal verbs in everyday English: 15001700. In: Andrew James John-
ston, Ferdinand von Mengden and Stefan Thim (eds.), Language and
Text Current Perspectives on English and Germanic Historical Lin
guistics and Philology, 290306. (Anglistische Forschungen 359.) Hei-
delberg: Winter.
Thomas, Francis-Nol and Mark Turner
1994 Clear and Simple as the Truth Writing Classic Prose. Princeton:
Prince ton University Press.
Thompson, Sandra A.
1985 Grammar and written discourse: Initial vs. fnal purpose clauses in Eng-
lish. Text 5 (1): 5584.
Tomlin, Russell S. (ed.)
1987 Coherence and Grounding in Discourse: Outcome of a Symposium, Eu
gene, Oregon, June 1984. (Typological Studies in Language 11.) Am-
sterdam: Benjamins.
Traugott, Elizabeth Closs
1992 Syntax. In: Richard M. Hogg (ed.), The Cambridge History of the Eng
lish Language. Volume 1: The Beginnings to 1066, 168289. Cambridge:
Cambridge University Press.
Traugott, Elizabeth Closs
1995 The role of the development of discourse markers in a theory of gram-
maticalization. [Paper presented at ICHL XII; Manchester, 1995; (on-
line version accessed 12/11/1997; <http://www.stanford.edu/~traugott/
papers>)].
Traugott, Elizabeth Closs
1997 The discourse connective after all: A historical pragmatic account.
[Paper prepared for ICL, Paris, July 1997; (online version accessed
12/11/2003; <http://www.stanford.edu/~traugott/papers>)].
274 References
Traugott, Elizabeth Closs
1999 The rhetoric of counter-expectation in semantic change: A study in
subjectifcation. In: Andreas Blank and Peter Koch (eds.), Historical
Semantics and Cognition, 6189. (Cognitive Linguistics Research 13.)
Berlin/New York: Mouton de Gruyter.
Traugott, Elizabeth Closs
2004 Historical pragmatics. In: Laurence R. Horn and Gregory Ward (eds.),
The Handbook of Pragmatics, 538561. (Blackwell Handbooks in Lin-
guistics 16.) Oxford: Blackwell.
Traugott, Elizabeth and Richard B. Dasher
2002 Regularity in Semantic Change. (Cambridge Studies in Linguistics 97.)
Cambridge: Cambridge University Press.
Uhler, Karl
1926 Die Bedeutungsgleichheit der altenglischen Adjektiva und Adverbia mit
und ohne lic(lice). (Anglistische Forschungen 62.) Heidelberg: Winter.
Ungerer, Friedrich
1988 Syntax der englischen Adverbialien. (Linguistische Arbeiten 215.) T-
bingen: Niemeyer.
Valera, Salvador
1998 On subject-orientation in English lyadverbs. English Language and
Linguistics 2 (2): 263282.
Vandenbergen-Simon, Anne-Marie and Karin Aijmer
2001 The expectation marker of course in a cross-linguistic perspective. Lan
guages in Contrast 4 (1): 1343.
Vorlat, Emma
1975 The Development of English Grammatical Theory 15861737 With
Special Reference to the Theory of Parts of Speech. Leuven: Leuven
University Press.
Walker, James A.
1949 Gothic -leik and Germanic *lik in the Light of Gothic translations of
Greek originals. Philological Quarterly 28: 274293.
Wegener, Heide
2000 Koordination und Subordination Semantische und Pragmatische Unter-
schiede. In: Michel Lefvre (ed.), Subordination in Syntax, Semantik und
Textlinguistik, 3344. (Eurogermanistik 15.) Tbingen: Stauffenburg.
Westin, Ingrid
2002 Language Change in English Newspaper Editorials. (Language and
Computers 44.) Amsterdam: Rodopi.
Wiegand, Nancy
1982 From discourse to syntax: for in English causal clauses. In: Anders
Ahlqvist (ed.), Papers from the 5th International Conference on His
torical Linguistics, 385393. (Amsterdam Studies in the Theory and
History of Linguistic Science 4, 21.) Amsterdam: Benjamins.
Wischer, Ilse and Gabriele Diewald (eds.)
2002 New Refections on Grammaticalization. (Typological Studies in Lan-
guage 49.) Amsterdam: Benjamins.
Secondary sources 275
Workman, Samuel K.
1940 Fifteenth Century Translation as an Infuence on English Prose.
(Prince ton Studies in English 18.) Princeton/New York: Princeton Uni-
versity Press.
Appendix A
A.1. Adverbial connectors: items
The following list provides an inventory of all linguistic elements which have
served the function of an adverbial connector in any period of the history of
English, giving the date for their frst occurrence as an adverbial connector as
well as the periods in which they were used as adverbial connectors.
The dates for the frst occurrence refer to the frst attested use of the element
in question as a linking adverbial. This means that the lexeme itself can be
much older, but did not have a connective function on the sentence or text level
before or after this date (for the methodology and examples, see Chapter 6.4).
Items in square brackets were found in the OED or the MED, but are not
attested in the corpus material of the present study (for details on the corpus
texts, see Appendix C.1).
Italics are used to indicate that the items were found in an adverbial connec-
tor function in the corpus texts but that there is no such meaning given in the
OED. Dates are given in italics if the items are attested earlier in an adverbial
connector function than stated by the OED.
The inventory only comprises single word adverbs or lexicalized items such
as verbal (e. g. albeit, howbeit) or prepositional phrases (e. g. after all, on the
other hand, at the same time). It does not record fully transparent prepositional
phrases such as as a consequence or by contrast (for the criteria applied, see
Chapter 7.4).
The items are listed alphabetically; lexicalized, but not yet univerbated prep-
ositional phrases such as above all or at any rate are given under their respec-
tive nominal head element, i. e. all and rate.
Length marks are only given for the Old English items.
Symbols:
ambiguous adverb/conjunction (for details, see Chapter 5.3)
[ ] item not attested in the corpus texts (for this period)
italics item not listed as an adverbial connector in the OED
date in italics pre-dating of adverbial connector function (earlier than the frst
occurrence in the OED)
Adverbial connectors: items 277
O
l
d

E
n
g
l
i
s
h
(
O
E
)
1
1
0
0

1
3
5
0
(
M
E
1
/
2
)
1
3
5
0

1
4
2
0
(
M
E
3
)
1
4
2
0

1
5
0
0
(
M
E
4
)
1
5
0
0

1
5
7
0
(
E
M
o
d
E
1
)
1
5
7
0

1
6
4
0
(
E
M
o
d
E
2
)
1
6
4
0

1
7
1
0
(
E
M
o
d
E
3
)
1
7
1
0

1
7
8
0

(
L
M
o
d
E
1
)
1
7
8
0

1
8
5
0

(
L
M
o
d
E
2
)
1
8
5
0

1
9
2
0

(
L
M
o
d
E
3
)
1
9
2
0

(
P
D
E
)
a
c
c
o
r
d
i
n
g
l
y

1
7
1
1

a
c
c
o
r
d
i
n
g
l
y

a
c
c
o
r
d
i
n
g
l
y

a
c
c
o
r
d
i
n
g
l
y

r
e
s
t

f
r
s
t

a
f
t
e
r

1
4
0
0
/
1
4
5
0
a
g
a
i
n

1
5
3
3
a
g
a
i
n
a
g
a
i
n
a
g
a
i
n
a
g
a
i
n
a
g
a
i
n
a
g
a
i
n
a
l
b
e
i
t

1
5
3
1
a
l
b
e
i
t

a
l
g
a
t
e
s

1
3
8
6
[
a
l
g
a
t
e
s
]
[
a
l
g
a
t
e
s
]
a
l
s
o
,

s
e
e

e
a
l
l

s
w

a
b
o
v
e

a
l
l

1
6
2
5
a
b
o
v
e

a
l
l
a
b
o
v
e

a
l
l
a
b
o
v
e

a
l
l
a
b
o
v
e

a
l
l
a
b
o
v
e

a
l
l

[
a
f
t
e
r

a
l
l

1
7
1
3
]
a
f
t
e
r

a
l
l
a
f
t
e
r

a
l
l
a
f
t
e
r

a
l
l
a
f
t
e
r

a
l
l
[
a
l
l

i
n

a
l
l
]
[
a
l
l

t
h
e

s
a
m
e

1
8
7
8
]
[
a
l
l

t
h
e

s
a
m
e
]
[
a
l
t
o
g
e
t
h
e
r

1
8
1
7
]
[
a
l
t
o
g
e
t
h
e
r

1
8
1
7
]
[
a
l
t
o
g
e
t
h
e
r

1
8
1
7
]
[
a
n
y
h
o
w

1
8
2
5
]
[
a
n
y
h
o
w
]
[
a
n
y
h
o
w
]
[
a
n
y
w
a
y

1
8
5
9
]
[
a
n
y
w
a
y
]
b
e
s
i
d
e

1
5
9
2
b
e
s
i
d
e
s

1
5
9
6
b
e
s
i
d
e
s
b
e
s
i
d
e
s
b
e
s
i
d
e
s
b
e
s
i
d
e
s
b
e
s
i
d
e
s
i
n

a
n
y

c
a
s
e

1
9
2
0
i
n

a
n
y

c
a
s
e
[
c
e
r
t
a
i
n

1
3
3
0
]
c
e
r
t
a
i
n
c
e
r
t
a
i
n
[
c
e
r
t
a
i
n
l
y

1
3
0
0
]
c
e
r
t
a
i
n
l
y
[
c
e
r
t
a
i
n
l
y
]
[
c
e
r
t
a
i
n
l
y
]
[
c
e
r
t
a
i
n
l
y
]
[
c
e
r
t
a
i
n
l
y
]
[
c
e
r
t
a
i
n
l
y
]
c
e
r
t
a
i
n
l
y
c
e
r
t
a
i
n
l
y
c
e
r
t
a
i
n
l
y
[
c
e
r
t
e
s

1
3
0
0
]
c
e
r
t
e
s
c
e
r
t
e
s
c
e
r
t
e
s
c
e
r
t
e
s
c
e
r
t
e
s
[
c
o
n
s
e
q
u
e
n
t
l
y

1
5
3
3
]
[
c
o
n
s
e
q
u
e
n
t
l
y
]
c
o
n
s
e
q
u
e
n
t
l
y
c
o
n
s
e
q
u
e
n
t
l
y
c
o
n
s
e
q
u
e
n
t
l
y
c
o
n
s
e
q
u
e
n
t
l
y
c
o
n
s
e
q
u
e
n
t
l
y

c
o
n
t
r
a
r
y

1
5
9
3
c
o
n
t
r
a
r
i
l
y

1
5
4
0
[
c
o
n
t
r
a
r
i
l
y
]
[
c
o
n
t
r
a
r
i
l
y
]

[
c
o
n
t
r
a
r
i
l
y
]

[
c
o
n
t
r
a
r
i
l
y
]
c
o
n
t
r
a
r
i
l
y
[
c
o
n
t
r
a
r
i
l
y
]
c
o
n
t
r
a
r
i
w
i
s
e

1
4
5
0
c
o
n
t
a
r
i
w
i
s
e
c
o
n
t
r
a
r
i
w
i
s
e
[
c
o
n
t
r
a
r
i
w
i
s
e
]
[
c
o
n
t
r
a
r
i
w
i
s
e
]
[
c
o
n
t
r
a
r
i
w
i
s
e
]
c
o
n
t
r
a
r
i
w
i
s
e
[
c
o
n
t
r
a
r
i
w
i
s
e
]
[
c
o
n
t
r
a
r
i
w
a
y
s

1
5
8
8
]
c
o
n
v
e
r
s
e
l
y

1
8
0
6
[
c
o
n
v
e
r
s
e
l
y
]
[
c
o
n
v
e
r
s
e
l
y
]
c
o
r
r
e
s
p
o
n
d
i
n
g
l
y
o
f

c
o
u
r
s
e

1
8
1
1
o
f

c
o
u
r
s
e
o
f

c
o
u
r
s
e
[
c

c
e
]

a
c
e
e
k
e
e
k
278 Appendix A
O
l
d

E
n
g
l
i
s
h
(
O
E
)
1
1
0
0

1
3
5
0
(
M
E
1
/
2
)
1
3
5
0

1
4
2
0
(
M
E
3
)
1
4
2
0

1
5
0
0
(
M
E
4
)
1
5
0
0

1
5
7
0
(
E
M
o
d
E
1
)
1
5
7
0

1
6
4
0
(
E
M
o
d
E
2
)
1
6
4
0

1
7
1
0
(
E
M
o
d
E
3
)
1
7
1
0

1
7
8
0

(
L
M
o
d
E
1
)
1
7
8
0

1
8
5
0

(
L
M
o
d
E
2
)
1
8
5
0

1
9
2
0

(
L
M
o
d
E
3
)
1
9
2
0

(
P
D
E
)

a
c

s
w
y
l
c
e
e
a
c

s
w
l
y
c
e

e
a
l
l

s
w

a
l
s
w
a
a
l
s
o
a
l
s
o
a
l
s
o
a
l
s
o
a
l
s
o
[
a
l
s
o
]
a
l
s
o
a
l
s
o
a
l
s
o
e
f
t
e
f
t
[
e
f
t
s

n
a
/

e
s
]
e
f
t
s
o
n
a
/

e
s
e
l
l
e
s
a
n
d
/
o
r

e
l
l
e
s
a
n
d
/
o
r

e
l
l
e
s
a
n
d
/
o
r

e
l
s
e
o
r

e
l
s
e
o
r

e
l
s
e
o
r

e
l
s
e
o
r

e
l
s
e
o
r

e
l
s
e
o
r

e
l
s
e
o
r

e
l
s
e
e
o
r
n
o
s
t
l

c
e
[
e
r
g
o

1
4
0
0
]
i
n

a
l
l

e
v
e
n
t
s

1
8
1
1
i
n

a
l
l

e
v
e
n
t
s
i
n

a
l
l

e
v
e
n
t
s
i
n

f
a
c
t

1
8
1
5
i
n

f
a
c
t
i
n

f
a
c
t
f
n
a
l
l
y

1
4
2
5
f
n
a
l
l
y
[
f
n
a
l
l
y
]
[
f
n
a
l
l
y
]
[
f
n
a
l
l
y
]
[
f
n
a
l
l
y
]
f
n
a
l
l
y
f
n
a
l
l
y
[
i
n

f
n
e

1
5
5
0
]
[
i
n

f
n
e
]
[
i
n

f
n
e
]
[
i
n

f
n
e
]
i
n

f
n
e
i
n

f
n
e
[
f
r
s
t

1
3
0
0
]
f
r
s
t
f
r
s
t
f
r
s
t
f
r
s
t
f
r
s
t
f
r
s
t
f
r
s
t
f
r
s
t
f
r
s
t

f
o
r

f
o
r

f
o
r

f
o
r

f
o
r

f
o
r

f
o
r

f
o
r

f
o
r

f
o
r

f
o
r
h
w
i

1
2
2
5
f
o
r
h
w
i
f
o
r
h
w
i
[
f
o
r
s
o
o
t
h
e

1
3
2
0
]
f
o
r
s
o
o
t
h
e
f
o
r
s
o
o
t
h
e

f
o
r

f
o
r

o
n

f
o
r

o
n

f
o
r

o
n

f
o
r

t
h
a
t
f
o
r

t
h
a
t
f
o
r

t
h
a
t

f
o
r

f
o
r

i
f
o
r

i
f
o
r

i
f
o
r

w
h
i
c
h
f
o
r

w
h
i
c
h
f
o
r

w
h
i
c
h
f
o
r

t
h
e

w
h
i
c
h
[
f
u
r
t
h
e
r

1
4
0
5
]
f
u
r
t
h
e
r

[
f
a
r
t
h
e
r

1
5
6
2
]
f
u
r
t
h
e
r
[
f
u
r
t
h
e
r
]
[
f
u
r
t
h
e
r
]
[
f
u
r
t
h
e
r
]
f
u
r
t
h
e
r
f
u
r
t
h
e
r
f
e
r
t
h
e
r
m
o
r
e

1
3
8
2
f
u
r
t
h
e
r
m
o
r
e

1
3
9
0
f
u
r
t
h
e
r
m
o
r
e
f
u
r
t
h
e
r
m
o
r
e
f
u
r
t
h
e
r
m
o
r
e
f
u
r
t
h
e
r
m
o
r
e
[
f
u
r
t
h
e
r
m
o
r
e
]
[
f
u
r
t
h
e
r
m
o
r
e
]
[
f
u
r
t
h
e
r
m
o
r
e
]
f
u
r
t
h
e
r
m
o
r
e
[
f
a
r
t
h
e
r
o
v
e
r
]
,

f
u
r
t
h
e
r
o
v
e
r

1
3
9
0
[
f
u
r
t
h
e
r
o
v
e
r
]
o
n

o

e
r

h
a
l
f
o
n

t
h
e

o
t
h
e
r

h
a
n
d

o
n

t
h
e

o
t
h
e
r

h
a
n
d
o
n

t
h
e

o
n
e
/
o
t
h
e
r

h
a
n
d
o
n

t
h
e

o
n
e
/
o
t
h
e
r

h
a
n
d
o
n

t
h
e

o
n
e
/

o
t
h
e
r

h
a
n
d
h
e
n
c
e

1
6
6
0
h
e
n
c
e
h
e
n
c
e
h
e
n
c
e
h
e
n
c
e
h

r
h
e
r
e
h
e
r
e
h
e
r
e
[
h
e
r
e
]
[
h
e
r
e
]
[
h
e
r
e
]
h
e
r
e
h
e
r
e
h
e
r
e
h
e
r
e
h
e
r
e
f
o
r
e

1
2
0
0
h
e
r
e
f
o
r
e
[
h
e
r
e
f
o
r
e
]
[
h
e
r
e
f
o
r
e
]
h
e
r
e
f
o
r
e
h
o
w
b
e
i
t

1
4
6
5
h
o
w
b
e
i
t
h
o
w
b
e
i
t
h
o
w
e
v
e
r

1
6
1
3
h
o
w
e
v
e
r
h
o
w
e
v
e
r
h
o
w
e
v
e
r
h
o
w
e
v
e
r
h
o
w
e
v
e
r
Adverbial connectors: items 279
O
l
d

E
n
g
l
i
s
h
(
O
E
)
1
1
0
0

1
3
5
0
(
M
E
1
/
2
)
1
3
5
0

1
4
2
0
(
M
E
3
)
1
4
2
0

1
5
0
0
(
M
E
4
)
1
5
0
0

1
5
7
0
(
E
M
o
d
E
1
)
1
5
7
0

1
6
4
0
(
E
M
o
d
E
2
)
1
6
4
0

1
7
1
0
(
E
M
o
d
E
3
)
1
7
1
0

1
7
8
0

(
L
M
o
d
E
1
)
1
7
8
0

1
8
5
0

(
L
M
o
d
E
2
)
1
8
5
0

1
9
2
0

(
L
M
o
d
E
3
)
1
9
2
0

(
P
D
E
)

h
o
w
s
o
e
v
e
r

1
5
8
6

h
o
w
s
o
e
v
e
r
[
h
o
w
s
o
m
e
v
e
r

1
5
6
2
]
[
h
o
w
s
o
m
e
v
e
r
]
[
h
o
w
s
o
m
e
v
e
r
]
[
h
o
w
s
o
m
e
v
e
r
]
[
h
o
w
s
o
m
e
v
e
r
]
[
h
o
w
s
o
m
e
v
e
r
]
h

r
u
h

r
u

i
n
g
a
h
w

t
h
w

t
h
w

h
w

h
w

(
e
)
r
e
i
n
c
i
d
e
n
t
a
l
l
y

1
9
2
5
i
n
d
e
e
d

1
6
0
0
i
n
d
e
e
d
i
n
d
e
e
d
i
n
d
e
e
d
i
n
d
e
e
d
i
n
d
e
e
d
i
n
s
t
e
a
d
i
t
e
m

1
3
9
8
i
t
e
m
i
t
e
m
i
w
i
s

1
1
7
5
a
t

(
t
h
e
)

l
a
s
t

1
2
7
5
a
l
a
s
t

1
2
3
0
a
t

l
a
s
t
a
t

l
a
s
t

a
t

l
a
s
t
a
t

l
a
s
t
[
a
t

l
a
s
t
]
[
a
t

l
a
s
t
]
[
a
t

l
a
s
t
]
a
t

l
a
s
t
a
t

l
a
s
t
l
a
s
t
l
y

1
5
8
6
l
a
s
t
l
y
l
a
s
t
l
y
l
a
s
t
l
y
[
l
a
s
t
l
y
]
[
l
a
s
t
l
y
]
a
t

l
e
a
s
t

1
6
0
5
a
t

l
e
a
s
t
a
t

l
e
a
s
t
a
t

l
e
a
s
t
a
t

l
e
a
s
t
a
t

l
e
a
s
t
l
i
k
e
w
i
s
e

1
5
0
9
l
i
k
e
w
i
s
e
l
i
k
e
w
i
s
e
l
i
k
e
w
i
s
e
l
i
k
e
w
i
s
e
l
i
k
e
w
i
s
e
l
i
k
e
w
i
s
e
[
l
i
k
e
w
a
y
s

1
5
5
1
]
[
l
i
k
e
w
a
y
s
]
[
l
i
k
e
w
a
y
s
]
[
l
i
k
e
w
a
y
s
]
i
n

t
h
e

m
e
a
n
t
i
m
e

1
6
3
1
m
e
a
n
t
i
m
e
[
i
n

t
h
e

m
e
a
n

t
i
m
e
]
[
i
n

t
h
e

m
e
a
n

t
i
m
e
]
i
n

t
h
e

m
e
a
n
t
i
m
e
[
i
n

t
h
e

m
e
a
n

t
i
m
e
]
[
i
n

t
h
e

m
e
a
n

t
i
m
e
]
(
i
n

t
h
e
)

m
e
a
n

w
h
i
l
e

1
5
9
7
[
m
e
a
n
w
h
i
l
e
]
[
m
e
a
n
w
h
i
l
e
]
[
m
e
a
n
w
h
i
l
e
]
[
m
e
a
n
w
h
i
l
e
]
[
m
e
a
n
w
h
i
l
e
]
[
m
o
r
e

1
2
3
0
]
[
m
o
r
e
]
[
m
o
r
e
]
[
m
o
r
e
]
[
m
o
r
e
]
[
m
o
r
e
]
[
m
o
r
e
]
[
m
o
r
e
]
[
m
o
r
e
]
[
m
o
r
e
]
m
o
r
e
o
v
e
r

1
3
8
5
m
o
r
e
o
v
e
r
m
o
r
e
o
v
e
r
m
o
r
e
o
v
e
r
m
o
r
e
o
v
e
r
m
o
r
e
o
v
e
r
m
o
r
e
o
v
e
r
m
o
r
e
o
v
e
r
m
o
r
e
o
v
e
r
n

s
n
a
t
h
e
l
e
s

n
a
t
h
e
l
e
s
n
a
t
h
e
l
e
s
[
n
e
v
e
r
t
h
e
l
a
t
t
e
r

1
2
2
5
]
[
n
e
v
e
r
t
h
e
l
a
t
t
e
r
]
n
e
v
e
r
t
h
e
l
a
t
t
e
r
n
e
v
e
r
t
h
e
l
e
s
s

1
3
8
2
n
e
v
e
r
t
h
e
l
e
s
s
[
n
e
v
e
r
t
h
e
l
e
s
s
]
[
n
e
v
e
r
t
h
e
l
e
s
s
]
[
n
e
v
e
r
t
h
e
l
e
s
s
]
[
n
e
v
e
r
t
h
e
l
e
s
s
]
[
n
e
v
e
r
t
h
e
l
e
s
s
]
[
n
e
v
e
r
t
h
e
l
e
s
s
]
[
n
e
v
e
r
t
h
e
l
e
s
s
]
[
n
o
n
e
t
h
e
l
e
s
s
]

1
5
3
3
[
n
o
n
e
t
h
e
l
e
s
s
]
[
n
o
n
e
t
h
e
l
e
s
s
]
[
n
o
n
e
t
h
e
l
e
s
s
]
[
n
o
n
e
t
h
e
l
e
s
s
]
[
n
o
n
e
t
h
e
l
e
s
s
]
n
o
n
e
t
h
e
l
e
s
s
280 Appendix A
O
l
d

E
n
g
l
i
s
h
(
O
E
)
1
1
0
0

1
3
5
0
(
M
E
1
/
2
)
1
3
5
0

1
4
2
0
(
M
E
3
)
1
4
2
0

1
5
0
0
(
M
E
4
)
1
5
0
0

1
5
7
0
(
E
M
o
d
E
1
)
1
5
7
0

1
6
4
0
(
E
M
o
d
E
2
)
1
6
4
0

1
7
1
0
(
E
M
o
d
E
3
)
1
7
1
0

1
7
8
0

(
L
M
o
d
E
1
)
1
7
8
0

1
8
5
0

(
L
M
o
d
E
2
)
1
8
5
0

1
9
2
0

(
L
M
o
d
E
3
)
1
9
2
0

(
P
D
E
)
n
o
t
w
i
t
h
s
t
a
n
d
i
n
g

1
4
4
0
n
o
t
w
i
t
h
s
t
a
n
d
i
n
g
n
o
t
w
i
t
h
s
t
a
n
d
i
n
g
n
o
t
w
i
t
h
s
t
a
n
d
i
n
g
n
o
t
w
i
t
h
s
t
a
n
d
i
n
g
[
n
o
u
g
h
t

f
o
r

t
h
a
n
]
[
n
o
u
g
h
t

f
o
r

t
h
a
n
]
[
n
o
u
g
h
t

f
o
r

t
h
a
n
]
[
n
o
u
g
h
t

f
o
r

t
h
a
t
]

[
n
o
u
g
h
t

f
o
r

t
h
a
t
]
[
n
o
u
g
h
t

f
o
r

t
h
i
]

[
n
o
u
g
h
t

f
o
r

t
h
i
]

n
u
n
u
n
o
w
n
o
w
n
o
w
n
o
w
n
o
w
n
o
w
n
o
w
n
o
w
o
t
h
e
r
w
a
y
s

1
3
9
0
[
o
t
h
e
r
w
a
y
s
]
[
o
t
h
e
r
w
a
y
s
]
[
o
t
h
e
r
w
a
y
s
]
[
o
t
h
e
r
w
a
y
s
]
[
o
t
h
e
r
w
a
y
s
]
[
o
t
h
e
r
w
a
y
s
]
[
o
t
h
e
r
w
i
s
e

1
3
9
3
]
[
o
t
h
e
r
w
i
s
e
]
[
o
t
h
e
r
w
i
s
e
]
o
t
h
e
r
w
i
s
e
o
t
h
e
r
w
i
s
e
[
o
t
h
e
r
w
i
s
e
]
[
o
t
h
e
r
w
i
s
e
]
[
o
t
h
e
r
w
i
s
e
]
[
o
t
h
e
r
w
i
s
e
]
o
v
e
r

t
h
a
t

1
3
9
3
o
v
e
r

t
h
a
t
o
v
e
r

t
h
a
t
[
o
v
e
r

t
h
i
s

1
3
9
3
]
o
v
e
r

t
h
i
s
o
v
e
r

t
h
i
s
o
v
e
r

a
n
d

a
b
o
v
e

1
5
8
8
[
o
v
e
r

a
n
d

a
b
o
v
e
]
[
o
v
e
r

a
n
d

a
b
o
v
e
]
[
o
v
e
r

a
n
d

a
b
o
v
e
]
o
v
e
r

a
n
d

a
b
o
v
e
[
o
v
e
r

a
n
d

b
e
s
i
d
e
s
]

1
5
3
3
o
v
e
r

a
n
d

b
e
s
i
d
e
s
[
o
v
e
r
a
l
l
]

1
2
2
5
o
v
e
r
m
o
r
e

1
3
8
2
o
v
e
r
m
o
r
e
[
p
l
u
s

1
9
6
8
]
a
t

a
n
y

r
a
t
e

1
8
1
1
[
a
t

a
n
y

r
a
t
e
]
r
a
t
h
e
r
s
e
k
i
r
l
y
s
e
k
i
r
l
y
s
e
m
b
l
a
b
l
y

1
5
3
1
[
s
e
m
b
l
a
b
l
y
]
o

e
r

s
i
d
e

1
2
2
5
o
n

t
h
e

o
t
h
e
r

s
i
d
e
o
n

t
h
e

o
t
h
e
r

s
i
d
e
o
n

t
h
e

o
t
h
e
r

s
i
d
e
o
n

t
h
e

o
t
h
e
r

s
i
d
e
s
i
m
i
l
a
r
l
y
s
o

s
e
e

s
w

c
e
s
o
o
t
h
l
y
s
o
o
t
h
l
y
s
o
o
t
h
l
y
s
t
i
l
l

1
7
0
7
s
t
i
l
l

s
t
i
l
l
s
t
i
l
l
s
t
i
l
l
s
u
i
n
g
l
y

1
3
8
0
s
u
i
n
g
l
y
i
n

s
u
m
[
i
n

s
u
m
]
[
i
n

s
u
m
]
[
i
n

s
u
m
]
[
i
n

s
u
m
]
[
i
n

s
u
m
]
s
u
r
e
s
u
r
e
s
u
r
e
s
u
r
e
[
s
u
r
e
]
[
s
u
r
e
]
s
u
r
e
l
y

1
5
5
6
s
u
r
e
l
y
s
u
r
e
l
y
s
u
r
e
l
y
s
u
r
e
l
y
s
u
r
e
l
y
[
s
u
r
e
l
y
]

s
w

s
w
a
s
o
s
o
s
o
s
o
s
o
s
o
s
o
s
o
s
o
Adverbial connectors: items 281
O
l
d

E
n
g
l
i
s
h
(
O
E
)
1
1
0
0

1
3
5
0
(
M
E
1
/
2
)
1
3
5
0

1
4
2
0
(
M
E
3
)
1
4
2
0

1
5
0
0
(
M
E
4
)
1
5
0
0

1
5
7
0
(
E
M
o
d
E
1
)
1
5
7
0

1
6
4
0
(
E
M
o
d
E
2
)
1
6
4
0

1
7
1
0
(
E
M
o
d
E
3
)
1
7
1
0

1
7
8
0

(
L
M
o
d
E
1
)
1
7
8
0

1
8
5
0

(
L
M
o
d
E
2
)
1
8
5
0

1
9
2
0

(
L
M
o
d
E
3
)
1
9
2
0

(
P
D
E
)
s
w

a
h
s
w
a

e
a
h
s
w

a
h
h
w

r
e

a
[

e
t
]

r
t

a
c
a
n
[
t
o
e
k
e
n
]

a
h

e
a
h

e
a
h

t
h
o

t
h
o
u
g
h

t
h
o
u
g
h

a
h
h
w

e
r
e

e
a
h
h
w

e
r
e

o
n
n
e

t
h
a
n
e
/
t
h
e
n
n
e
t
h
e
n
t
h
e
n
t
h
e
n
t
h
e
n
t
h
e
n
t
h
e
n
t
h
e
n
t
h
e
n
t
h
e
n
t
h
e
n
c
e

1
6
5
2

o
n
n
e
,

s
e
e

t
h
e
n

i
t
h
e
r
e
f
o
r
e

1
2
0
0
t
h
e
r
e
f
o
r
e
t
h
e
r
e
f
o
r
e
t
h
e
r
e
f
o
r
e
t
h
e
r
e
f
o
r
e
t
h
e
r
e
f
o
r
e
t
h
e
r
e
f
o
r
e
t
h
e
r
e
f
o
r
e
t
h
e
r
e
f
o
r
e
t
h
e
r
e
f
o
r
e
t
h
u
s

1
3
8
0
t
h
u
s
t
h
u
s
t
h
u
s
t
h
u
s
t
h
u
s
t
h
u
s
t
h
u
s
t
h
u
s
a
t

t
h
e

s
a
m
e

t
i
m
e

1
7
3
9
a
t

t
h
e

s
a
m
e

t
i
m
e
a
t

t
h
e

s
a
m
e

t
i
m
e
a
t

t
h
e

s
a
m
e

t
i
m
e
t
o
o
t
o
o
t
o
o
t
o
o
t
r
u
l
y

1
4
1
7
t
r
u
l
y
t
r
u
l
y
t
r
u
l
y
t
r
u
l
y
t
r
u
l
y
t
r
u
l
y
t
r
u
l
y
i
n

t
r
u
t
h

1
7
7
0
i
n

t
r
u
t
h
[
v
e
r
i
l
y

1
3
8
2
]
v
e
r
i
l
y
v
e
r
i
l
y
v
e
r
i
l
y
v
e
r
i
l
y
v
e
r
i
l
y
v
i
d
e
l
i
c
e
t
i
n

t
h
e

s
a
m
e

w
a
y
i
n

t
h
e

s
a
m
e

w
a
y
w
e
l
l
w
e
l
l
w
e
l
l
w
e
l
l
w
e
l
l
w
e
l
l
w
e
l
l
w
h
e
n
c
e

1
6
6
5
w
h
e
n
c
e

w
h
e
r
e
f
o
r
e

1
2
7
5
w
h
e
r
e
f
o
r
e
w
h
e
r
e
f
o
r
e
w
h
e
r
e
f
o
r
e
w
h
e
r
e
f
o
r
e
w
h
e
r
e
f
o
r
e
w
h
e
r
e
f
o
r
e
w
h
e
r
e
f
o
r
e
(
g
e
)
w

s
l

c
e
w
i
t
o
d
l

c
e
w
i
t
o
d
l
i
c
e
y
e
t

(
g
i
e
t
)

1
2
0
5
y
e
t
y
e
t
y
e
t
y
e
t
y
e
t
y
e
t
y
e
t
y
e
t
y
e
t
282 Appendix A
A.2. Alphabetical index of adverbial connectors
Adverbial Connector Semantics Listed in
Appendix
First
Occurrence
(in Corpus Texts)
Last
Occurrence
(in Corpus Texts)
rest frst listing/enumerative B.5.1 OE OE
accordingly result B.3.1 LModE1 PDE
after moreover additive/reinforcing B.1.2 ME4 ME4
again
(but yet) again
additive/reinforcing
contrast
B.1.2 EModE1
(EModE2)
PDE
albeit contrast/concession B.4.1 EModE1 EModE2
algates contrast/concession B.4.1 ME3 EModE1
(ME3)
all in all additive/reinforcing PDE PDE
above all additive/reinforcing B.1.2 EModE2 PDE
after all contrast/concession B.4.1 EModE3
(LModE1)
PDE
all the same contrast/concession LModE3 PDE
also (OE eallsw) additive/reinforcing B.1.2 OE PDE
altogether summative LModE2 PDE
anyhow contrast/concession/
dismissive transition
LModE2 PDE
anyway contrast/concession
dismissive transition
LModE3 PDE
beside additive/reinforcing B.1.2 EModE2 EModE2
besides additive/reinforcing B.1.2 EModE2
(EModE3)
PDE
in any case contrast/concession
dismissive
B.4.1 LModE3 PDE
certain ?transition B.5.2 ME1/2
(ME3)
ME4
certainly ?transition B.5.2 ME1/2 PDE
certes transition B.5.2 ME1/2 LModE2
consequently result B.3.1 EModE1
(LModE1)
PDE
contrary contrast/antithetic B.4.2 EModE2 EModE2
contrarily contrast/antithetic B.4.2 EModE1 PDE
contrariways contrast/antithetic EModE2
contrariwise contrast/antithetic B.4.2 ME4
(EModE1)
LModE3
conversely contrast/antithetic B.4.2 LModE2
(LModE3)
PDE
correspondingly additive/equative B.1.3 PDE PDE
of course contrast/concession
transition
B.4.1
B.5.2
LModE2 PDE
clce certainly result/transition OE OE
ac also additive/reinforcing B.1.2 OE PDE
ac swylce/swylce ac also additive/reinforcing B.1.2 OE ME1
eallsw see also
eft also additive/reinforcing B.1.2 OE ME2
Adverbial connectors: items 283
Adverbial Connector Semantics Listed in
Appendix
First
Occurrence
(in Corpus Texts)
Last
Occurrence
(in Corpus Texts)
eftsna/eftsnes also additive/reinforcing B.1.2 OE
(ME1)
ME1
else (OE elles) additive/reinforcing
contrast/reformulatory
B.4.3 OE PDE
eornostlce earnestly transition B.5.2 OE OE
ergo result ME3 LModE1
at/in all events contrast/concession
dismissive
B.4.1 LModE2 PDE
in fact contrast/avowal
transition
B.5.2 LModE2 PDE
fnally listing/enumerative B.1.1 ME4
(EModE1)
PDE
in fne summative B.2 EModE1
(LModE2)
LModE3
frst listing/enumerative B.1.1 ME1/2
(ME3)
PDE
for cause B.3.3 ME1/2 PDE
(LModE3)
forhwi result B.3.1 ME1/2 ME3/4
forsooth transition B.5.2 ME1/2
(ME3)
ME4
for that result B.3.1 ME1/2 ME4
form/foron result B.3.1 OE ME2
for result B.3.1 OE ME4
for which result B.3.1 ME3 EModE1
further additive/reinforcing B.1.2 ME3
(EModE1)
PDE
furthermore/furthermore additive/reinforcing B.1.2 ME3 PDE
furtherover/fartherover additive/reinforcing B.1.2 ME3 EModE2
(ME3)
on oer half contrast/antithetic B.4.2 ME1/2 ME1/2
on the other hand contrast/
antithetic/replacive
B.4.2 EModE3 PDE
hence result B.3.1 EModE3
(LModE1)
PDE
here (OE hr) transition B.5.1 OE PDE
herefore result B.3.1 ME1/2 EModE2
how ?transition B.5.4 EModE2 EModE2
howbeit contrast/concession B.4.1 ME4 EModE2
however contrast/concession B.4.1 EModE2
(EModE3)
PDE
howsoever contrast/concession B.4.1 EModE1 EModE3
howsomever contrast/concession EModE1 LModE3
hru nevertheless contrast/concession B.4.1 OE OE
hruinga nevertheless contrast/concession B.4.1 OE OE
hw(e)re contrast/concession B.4.1 OE OE
hwt what transition B.5.4 OE ME1/2
hwt what then transition B.5.4 OE ME1/2
incidentally transition B.5.2 PDE PDE
284 Appendix A
Adverbial Connector Semantics Listed in
Appendix
First
Occurrence
(in Corpus Texts)
Last
Occurrence
(in Corpus Texts)
indeed transition B.5.2 EModE2 PDE
instead contrast/antithetic B.4.2 PDE PDE
item additive/reinforcing B.1.2 ME3 EModE1
iwis transition ME1/2
at (the) last listing/enumerative B.1.1 ME2 PDE
lastly listing/enumerative B.1.1 EModE2 PDE
(LModE3)
at least contrast/concession B.4.1 EModE2
(EModE3)
PDE
likeways additive/equative B.1.3 EModE1 LModE1
likewise additive/equative B.1.3 EModE1 PDE
(in the) meantime contrast/concession
time
B.4.1 EModE2
(LModE2)
PDE
(in the) meanwhile contrast/concession EModE2 PDE
more additive/reinforcing ME1/2 PDE
moreover additive/reinforcing B.1.2 ME3
(ME4)
PDE
nls nevertheless contrast/concession B.4.1 OE
(ME1/2)
ME4
neverthelatter contrast/concession B.4.1 ME1/2
(ME4)
ME4
nevertheless contrast/concession B.4.1 ME3 PDE
next ?listing/enumerative B.1.1 ME3 LModE3
nonetheless contrast/concession B.4.1 EModE1 PDE
notwithstanding contrast/concession B.4.1 ME4 LModE1
noughtforthan nevertheless contrast/concession ME2 ME4
noughtforthat nevertheless contrast/concession ME3 ME4
noughtforthi nevertheless contrast/concession ME3 ME4
now (OE n) transition B.5.1 OE PDE
otherways contrast/antithetic ME3 LModE2
otherwise contrast/antithetic B.4.2 EModE2 PDE
over this/that additive/reinforcing B.1.2 ME3 EModE1
over and above additive/reinforcing B.1.2 EModE2 LModE3
over and besides additive/reinforcing B.1.2 EModE1
(EModE2)
EModE2
overmore additive/reinforcing B.1.2 ME3 ME4
overall additive/reinforcing ME1/2 PDE
peradventure ?transition B.5.3 ME3 EModE1
perchance ?transition B.5.3 EModE1 EModE1
perhaps ?transition B.5.3 LModE2 PDE
plus additive/reinforcing PDE PDE
at any rate contrast/concession
dismissive
B.4.1 LModE3 PDE
rather contrast/reformulatory B.4.3 PDE PDE
sekirly transition B.5.2 ME3 ME4
semblably additive/equative B.1.3 EModE1 EModE2
(EModE1)
oer side contrast/antithetic B.4.2 ME1/2 (LModE3)
Adverbial connectors: items 285
Adverbial Connector Semantics Listed in
Appendix
First
Occurrence
(in Corpus Texts)
Last
Occurrence
(in Corpus Texts)
similarly additive/equative B.1.3 PDE PDE
slce truly transition B.5.2 OE ME4
still contrast/concession B.4.1 EModE3 PDE
in sum summative B.2 EModE2 PDE
suingly consequently result B.3.1 ME3
(ME4)
ME4
sure transition B.5.2 EModE2
(EModE3)
PDE
surely transition B.5.2 EModE1 PDE
sw so result B.3.1 OE PDE
swah nevertheless contrast/concession B.4.1 OE ME1/2
swahhwere nevertheless contrast/concession B.4.1 OE OE
then transition B.5.1 OE ME1
get also additive/reinforcing OE OE
(r)tacan also additive/reinforcing B.1.2 OE ME1
s result OE OE
ah though contrast/concession B.4.1 OE PDE
ahhwere nevertheless contrast/concession B.4.1 OE ME1/2
then (OE onne) result/interference
time
B.3.2 OE PDE
thence result B.3.1 EModE3 EModE3
therefore result B.3.1 ME1/2 PDE
thus result B.3.1 ME3 PDE
result B.3.1 OE ME1/2
at the same time contrast/concession
time
B.4.1 LModE1 PDE
too additive/reinforcing B.1.2 LModE1
(LModE2)
PDE
truly transition B.5.2 ME3 LModE3
in truth transition B.5.2 LModE1 LModE3
verily transition B.5.2 ME3
(ME4)
EModE3
videlicet additive/equative B.1.3 EModE3 EModE3
in the same way additive/equative B.1.3 LModE3 PDE
well transition B.5.5 EModE1 PDE
whence result B.3.1 EModE3 EModE3
wherefore result B.3.1 ME1/2 LModE3
why ?transition B.5.4 EModE1 LModE1
gewslce certainly transition B.5.4 OE OE
witodlce certainly transition B.5.2 OE ME1/2
yet contrast/concession B.4.1 ME1/2 PDE
English
above all 11, 40, 81, 91, 98, 108, 112,
216, 276277, 282; B.1.2.
ac 1718, 76, 192, 238; B.4.4.
actually 127, 170171
accordingly 81, 9091, 93, 96, 140,
151, 153, 155, 167, 277, 282; B.3.1.
rest 58, 81, 92, 107, 109, 277, 282;
B.1.1.
after 81, 107, 112, 216, 282; B.1.2.
after all 3, 1011, 4546, 81, 91, 98,
104, 108111, 151, 154, 167, 180182,
190, 249, 276277, 282; B.4.1.
again 3940, 81, 9394, 108109, 112,
193, 216217, 246, 277, 282; B.1.2.
alas B.5.5.
albeit 4, 10, 25, 32, 79, 81, 97, 104, 173,
179, 182, 194, 276277, 282; B.4.1.
algates 81, 93, 96, 107, 112, 179, 182,
277, 282; B.4.1.
all (OE eall) 11, 170, 174, 181182,
190, 276
all in all 40, 81, 91, 98, 277, 282
all the same 81, 98, 171, 180, 182, 190,
277, 282
also (OE eall-swa) 19, 151, 215219,
277, 282; B.1.2.
although 9, 27, 110, 169173, 186, 190
191, 204, 207210, 226; see also
though, eah
altogether 81, 98, 277, 282
and 9, 14, 17, 61, 7677, 214215, 224
225, 238240, 244; B.1.5.
any 103, 170171, 181182
anyhow 40, 81, 91, 98, 103, 170171,
180, 182, 190, 277, 282
anyway 38, 40, 45, 81, 91, 98, 103, 108,
112, 180, 182, 190, 198, 204, 277, 282
as 6, 8, 3031, 164165
at (the) last 81, 107, 109, 222, 231, 279,
284; B.1.1.
at any rate 11, 81, 98, 170171, 276,
280, 284; B.4.1.
at least 81, 98, 170171, 180, 190, 279,
284; B.4.1.
at the same time 81, 98, 108109, 171,
180181, 190, 250, 276, 281, 285;
B.4.1.
at/in all events 11, 81, 98, 180, 182,
190, 278, 283; B.4.1.
because 6, 8, 22, 24, 28, 3031, 100,
104, 132140, 151, 153, 162, 164167,
185; B.3.3.
beside(s) 23, 3940, 81, 9397, 100,
108, 112, 216218, 220, 235, 277, 280,
282; B.1.2.
but 9, 2526, 77, 79, 203207, 238
240; B.4.4.
by/in consequence B.3.
by/in contrast B.4.2.
certain 79, 81, 8990, 93, 107, 114115,
228, 277, 282; B.5.2.
certainly 79, 81, 93, 95, 107, 114115,
228, 277, 282; B.5.2.
certes 79, 81, 8990, 93, 107, 114, 162
163, 227228, 231, 234, 277, 282;
B.5.2.
consequently 32, 38, 4041, 81, 9091,
93, 96, 106, 139140, 151, 153, 155,
167, 277, 282; B.3.1.
contrarily 81, 176177, 277, 282; B.4.2.
contrariways 81, 91, 95, 176177, 277,
282
contrariwise 81, 9091, 95, 97, 176
177, 277, 282; B.4.2.
contrary 81, 176177, 277, 282; B.4.2.
(on the) contrary 10, 40, 77, 96, 170
171, 177
conversely 40, 81, 9091, 93, 96, 177,
277, 282; B.4.2.
Word index
Word index 287
correspondingly 40, 81, 9091, 93, 221,
277, 282; B.1.3.
culice 81, 92, 95, 107, 155, 277, 282
eac 18, 5859, 64, 81, 92, 174, 216, 277,
282; B.1.2.
eac swylce/swylce eac 59, 81, 97, 216,
278, 282; B.1.2.
eala B.5.5.
eall-swa 151, 215, 277; see also also
efne B.5.5.
eft 1718, 5861, 64, 81, 92, 94, 107,
109, 216, 278, 282; B.1.2.
eftsona/eftsones 1718, 58, 81, 92, 94,
107, 109, 216, 278, 283; B.1.2.
else 23, 58, 8081, 92, 96, 176177,
216218, 278, 283; B.1.2., B.4.3.
eornostlice 58, 81, 92, 95, 107, 114,
228, 231, 278, 283; B.5.2.
ergo 81, 8990, 93, 155, 157, 278, 283
es 9596
ever 168, 182
farther see further
fnally 81, 9091, 93, 107, 109, 222,
278, 283; B.1.1.
frst(ly) 41, 81, 107, 112, 222224, 231,
246, 278, 283; B.1.1.
for 5, 89, 77, 152154, 156, 161164,
225226, 239240, 249, 278, 283;
B.3.3.
for as much as 132, 154, 164166
for a start B.1.1.
for example B.1.4.
forhwi 7981, 97, 155, 157158, 162,
278, 283; B.3.1.
for instance B.1.4.
forsooth 3, 39, 5456, 8081, 97, 107,
114, 122124, 228, 232, 250, 283;
B.5.2.
for that/for this 6, 27, 79, 81, 97, 101
102, 132, 155, 158161, 166, 249, 278,
283; B.3.1.
form/foron/fory 5, 24, 30, 59, 63
64, 73, 75, 81, 97, 99102, 132,
140157, 159160, 162, 164, 166, 185,
187, 214, 248, 278, 283; B.3.1.
for which 81, 97, 155, 157, 278, 283;
B.3.1.
further/farther 3840, 81, 9394, 107,
112, 216218, 220, 278, 283; B.1.2.
furthermore/farthermore 14, 16, 40, 81,
9394, 107, 112, 215219, 231, 250,
278, 283; B.1.2.
furtherover/fartherover 81, 9394, 107,
112, 216, 219, 278, 283; B.1.2.
gan 13
gewislice 52, 59, 81, 92, 95, 107, 114,
285; B.5.2.
giet see yet
hence 8, 23, 27, 3940, 8182, 93, 95,
96, 108, 112113, 132, 134, 140, 151,
153, 155, 167, 235, 249, 278, 283; B.3.1.
here 38, 5861, 8081, 92, 107108,
112, 158, 227, 230, 246, 278, 283; B.5.1.
herefore 79, 81, 97, 107, 112113, 155,
157, 278, 283; B.3.1.
how 46, 183, 229, 283; B.5.4.
howbeit 10, 25, 81, 97, 104, 173, 179,
194, 276, 278, 283; B.4.1.
however 9, 2526, 3032, 37, 3940,
44, 81, 93, 108109, 136, 170171,
180, 182, 194197, 200, 204, 213, 219,
226, 235, 245, 278, 283; B.4.1.
howsoever 81, 93, 108109, 178, 180,
182, 279, 283; B.4.1.
howsomever 81, 93, 108109, 179, 182,
279, 283
huru 58, 81, 92, 107, 114, 179, 279, 283;
B.4.1.
huruinga 59, 81, 92, 107, 114, 179,
279, 283; B.4.1.
hwere 58, 81, 97, 181, 186187;
B.4.1.
hwt 3, 97, 99, 227, 229230, 279, 283;
B.5.4.
hwt a 81, 97, 99, 227, 229, 279, 283;
B.5.4.
288 Word index
in addition B.1.2.
in any case 40, 81, 171, 180, 182, 190,
277, 282; B.4.1.
in conclusion B.2.
in fact 3, 39, 4546, 81, 98, 105, 108,
114, 124, 127129, 151, 154, 170171,
180, 190, 209, 227231, 278, 283;
B.4.1.; B.5.2.
in fne 81, 98, 108109, 221, 278, 283;
B.2.
in other words B.1.4.
in sum 40, 81, 98, 221, 280, 285; B.2.
in the same way 81, 98, 108, 112, 221,
281, 285; B.1.3.
in truth 81, 98, 108, 114, 228, 281, 285;
B.5.2.
incidentally 40, 81, 9091, 93, 279,
283; B.5.2.
indeed 3, 10, 39, 4546, 81, 98, 103,
105, 108, 114, 124, 127128, 151, 153
154, 162, 227228, 230, 279, 284;
B.5.2.
instead 3, 40, 81, 98, 100, 108, 112,
170171, 176177, 279, 284; B.4.2.
item 18, 81, 8990, 93, 216, 219, 279,
284; B.1.2.
it is true B.5.2.
iwis 81, 93, 107, 114, 228, 231, 279,
284; B.5.2.
la/loo B.5.5.
lastly 81, 108109, 223, 279, 284; B.1.1.
likeways 81, 93, 221, 279, 284
likewise 35, 3940, 81, 93, 95, 217, 221,
279, 284; B.1.3.
ly 33, 9091, 93, 9596
(in the) meantime 81, 98, 108109,
180181, 184, 190, 279, 284; B.4.1.
(in the) meanwhile 40, 81, 98, 108110,
180181, 184, 190, 250, 279, 284
more 9395, 250
moreover 3941, 81, 9394, 107, 112,
215217, 219, 250, 279, 284; B.1.2.
nayls 5859, 81, 97, 99, 174, 178
179, 183, 186187, 220, 250, 279
neverthelatter 79, 81, 97, 107, 109, 179,
279, 284; B.4.1.
nevertheless 14, 16, 3941, 79, 81, 97,
101, 107, 109, 171173, 178179, 181,
183, 193, 245, 250, 279, 284; B.4.1.
next 284; B.1.1.
no/nay B.5.5.
no doubt B.5.2.
nonetheless 38, 45, 81, 97, 101, 172173,
179, 183, 186, 208, 279, 284; B.4.1.
notwithstanding 10, 81, 91, 97, 104,
107, 111112, 173, 179, 183, 280, 284;
B.4.1.
nought-for-than 81, 97, 179, 181, 183,
280, 284
nought-for-that 81, 97, 100, 179, 181,
183, 280, 284
nought-for-thi 81, 97, 100, 181, 183,
280, 284
now 3840, 74, 80, 102, 108, 116, 121,
137, 140, 151, 153, 162, 227, 230, 244,
280, 284; B.5.1.; see also nu
nu 38, 5864, 81, 92, 107109, 144146,
152, 158, 185, 192, 280, 284; B.5.1.; see
also now
other side 7981, 97, 107, 112, 176,
280, 284; B.4.2.
of course 81, 98, 104, 108109, 115,
151, 154, 180, 190, 227, 231, 277, 282;
B.4.1., B.5.2.
oh B.5.5.
on other half 7981, 97, 107, 112, 176,
278, 283; B.4.2.
on the one hand B.4.2.
on the other hand 40, 81, 96, 98, 108,
112, 171, 176177, 276, 278, 283; B.4.2.
or 76; B.4.4.
otherways 81, 97, 107, 112, 176177,
280, 284
otherwise 81, 97, 176177, 280, 284;
B.4.2.
oe 76; B.4.4.
Word index 289
over and above 10, 81, 98, 104, 108,
112, 216, 220, 280, 284; B.1.2.
over and besides 81, 97, 108, 112, 216,
220, 280, 284; B.1.2.
over that/this 6, 27, 81, 97, 100101,
107, 112, 216, 218, 280, 284; B.1.2.
overall 40, 79, 81, 97, 107, 112, 280, 284
overmore 17, 79, 81, 9394, 107, 112,
216, 219220, 250, 280, 284; B.1.2.
peradventure 47, 228229, 284; B.5.3.
perchance 47, 228229, 284; B.5.3.
perhaps 24, 47, 228229, 284; B.5.3.
plus 81, 8990, 93, 216, 219, 280, 284
rather 3, 81, 93, 95, 108, 170171, 176
177, 280, 284; B.4.3.
really B.5.2.
second(ly) B.1.1.
sekirly 81, 93, 95, 107, 114, 228, 280,
284; B.5.2.
semblably 81, 90, 93, 221, 280, 284;
B.1.3.
similarly 40, 81, 9091, 93, 106, 221,
280, 285; B.1.3.
since 4, 6, 8, 3031, 110, 133134,
136137, 140, 151, 153, 164165
soothly 48, 117123, 126127, 129,
231232, 234, 236, 280
solice 3, 39, 46, 48, 5153, 59, 81, 92, 95,
107, 114, 116123, 125, 129, 216, 227
228, 231232, 250, 280, 285; B.5.2.
still 31, 40, 79, 81, 9394, 108110,
171172, 180181, 183184, 191194,
204, 246, 280, 285; B.4.1.
suingly 81, 280, 285; B.3.1.
sure 81, 93, 108, 114115, 228, 230
231, 280, 285; B.5.2.
surely 81, 91, 93, 95, 108, 114115, 130,
162, 228, 231, 280, 285; B.5.2.
swa 3, 5, 59, 66, 81, 92, 97, 99100, 151
152, 155, 178, 215, 277, 280, 285; B.3.1.
swaeah 5, 58, 6667, 71, 81, 97, 100,
178179, 186187, 189, 281, 285; B.4.1.
swaeahhwere 5, 58, 6667, 100,
174, 178, 181, 186187, 189, 220, 250,
285; B.4.1.
a 3, 46, 5860, 62, 6466, 68, 7072,
81, 92, 107109, 151152, 227, 230,
238239, 281, 285; B.5.1.
a giet 79, 81, 97, 107, 109, 112, 216,
281, 285
m 5, 24, 27, 7374, 8485, 99, 101,
140, 143, 147149, 151, 156, 159, 249
(r)toeacan 5860, 81, 94, 107, 109,
112, 174, 216, 219220, 281; B.1.2.
s 59, 73, 81, 9799, 155, 281, 285
e 5, 27, 7475, 102, 142144, 147, 152,
160161, 178, 183, 187189
eah 1920, 59, 63, 6667, 7172, 81,
92, 179, 184192, 214, 281, 285; see
also though
eahhwere see swaeahhwere
on 5, 24, 73, 99, 140, 147149, 151,
156, 159
onne 5859, 68, 7072, 81, 92, 107
109, 144145, 151152, 155, 281, 285
y 5, 24, 59, 73, 81, 97101, 140, 144
145, 147149, 151152, 156, 159, 178,
183, 187, 281, 285; B.3.1.
that 5, 6, 9192, 100102, 249
then (OE onne) 1314, 31, 3841, 44
45, 72, 80, 108109, 111, 131, 140, 151,
153, 155, 162, 164, 167, 198, 201, 212
213, 218, 236, 245, 281, 285; B.3.2.
thence 81, 93, 9596, 108, 112113,
155, 281, 285; B.3.1.
there 112113, 158
therefore 3, 8, 14, 24, 2728, 30, 38
40, 72, 79, 81, 90, 97, 100, 107, 112
113, 133140, 151155, 157158, 164,
166, 231, 234236, 245, 281, 285;
B.3.1.
this 6, 100102, 249
though 9, 1920, 2526, 3032, 40,
4445, 67, 78, 80, 171172, 174175,
179, 184213, 249, 281, 285; B.4.1.
290 Word index
thus 8, 27, 38, 40, 55, 8182, 100, 140,
151, 154155, 157, 162, 246, 281, 285;
B.3.1.
too 3940, 81, 9394, 174, 200, 213,
216217, 219, 235, 281, 285; B.1.2.
truly 48, 81, 93, 95, 107, 114, 124127,
162, 228, 230231, 281, 285; B.5.2.
verily 81, 91, 93, 95, 107, 114, 129, 228,
230, 232, 250, 281, 285; B.5.2.
videlicet 81, 8990, 93, 221, 281, 285;
B.1.3.
well 3, 81, 93, 122, 281, 285; B.5.5.
what 229230; B.5.4.; see also hwt
what is more B.1.2.
whence 81, 93, 9596, 108, 112113,
155, 281, 285; B.3.1.
where 112113, 158
wherefore 3, 14, 56, 79, 81, 97, 107,
112113, 152155, 157158, 166, 281,
285; B.3.1.
while 4, 150, 171172, 204
why 22, 46, 229230, 285; B.5.4.
wise 90, 9596, 177
(to) wit B.1.4.
witodlice 3, 46, 5253, 59, 81, 92, 95,
107, 114, 116117, 125, 162, 227228,
231, 281, 285; B.5.2.
yes/yea B.5.5.
yet (OE giet) 16, 2526, 28, 31, 39, 79,
81, 93, 107, 109110, 171172, 179,
183184, 191194, 196, 204, 226, 231,
281, 285; B.4.1.
French
car 137, 162
non obstant 104, 183
parce que 7475, 137, 156, 166, 249
puisque 75, 137, 218
que 7475, 156, 166, 174, 249
tout soit il que 104, 173
German
auerdem 199
da 3031, 137, 150, 181
daher 150, 151
darum 150, 199
dass 74, 150
demnach 24, 27, 73, 100, 149, 150
denn 27, 3032, 137, 140, 150, 164
deshalb 24, 27, 32, 73, 139, 150151,
199
deswegen 2628, 73, 100, 132, 135,
149, 150, 151, 199
falls 31
infolgedessen 100, 150
nmlich 2627, 31, 132, 135, 137,
139140, 150, 164
obwohl 9, 3132, 202203, 212213,
249
somit 150151
whrend 110
weil 2628, 3032, 110, 135, 137, 150,
212
weshalb 150
weswegen 150
Latin
autem 5256, 123124
cum 164165
enim 27, 5256, 71, 123124, 132, 137,
140, 161164
ergo 53, 154
igitur 27, 53, 164
itaque 53
nam 27, 5253, 71, 132, 137, 140, 161
164
quare 164
quia 137, 160, 164
quidem 52, 55, 123124
quod 74, 137
vero 5255, 123124
addition 4041, 45, 5859, 77, 9495,
109, 111112, 131, 154, 162, 214226,
231, 234, 250, 282285
adverb, classifcation/defnition of 33
39, 4957
adverb, formation of 33, 53, 9096
adverb/conjunction, ambiguous 5, 20,
5960, 6466, 7880, 8485, 96, 99,
104, 113, 132, 143, 152, 157, 181, 185,
248
adverbial cline 4748, 117, 122, 124,
127130, 182, 227, 231232, 250
adverbial connectors, semantic catego-
ries 3941, 4445
adverbial, linking 4, 10, 22, 25, 3539,
4246, 50, 53, 87, 182, 201, 224, 238
lfric 5153, 142
ambiguous adverb/conjunction see
adverb/conjunction, ambiguous
Blair, Hugh 243
Campbell, George 2, 8, 39, 57, 233,
243246
cause/result 13, 16, 21, 27, 3032,
4041, 45, 5859, 73, 80, 85, 96, 99,
100, 104, 106, 109, 111, 131168, 170
171, 181, 186, 198, 214, 219, 225227,
234, 249, 282285
cccc 21, 41, 78, 80, 82, 96, 106, 131,
167169, 198, 214, 227, 282285; see
also cause; condition; contrast/
concession
CLMET 1112
collocation(s) 9, 1617, 2628, 43, 61
62, 67, 74, 7980, 90, 108, 121123,
125126, 135, 163, 165, 177, 187, 192
193, 195, 212, 215, 218, 232237, 239
240, 244246
complexity, cognitive 106, 168, 174,
214, 220, 250
complexity, morphological 77, 174,
220, 250
compound 3, 9295, 113, 154, 174, 177,
219220, 250
concession see contrast/concession
condition 2932, 41, 106, 131, 167
169, 174, 282285
conjunct 4, 25, 3339, 5253
conjunction 9, 17, 20, 2528, 30, 4953,
5759, 6162, 64, 66, 70, 7677, 79,
102, 125, 140149, 159, 187, 192, 200,
212, 234235, 238241, 244245; see
also adverb/conjunction, ambiguous
connect 2328, 247
connector, defnition of 12, 2328
connector, fnal position of 16, 25, 43
44, 72, 78, 185186, 195, 197213, 238
connector, impure 41, 78, 82, 96, 131,
198, 200, 214, 249
connector, initial position of 16, 25, 27,
4344, 47, 136, 195, 197198, 238
241, 245
connector, medial position of 9, 16, 20,
25, 27, 4344, 72, 78, 188189, 195,
197, 219, 233243, 250
connector, pronominal 56, 21, 24, 32,
5859, 7275, 84, 92, 96103, 106, 112,
132, 139, 143, 147151, 157159, 166,
175, 177178, 181, 199, 218, 248249
connector, pure 41, 45, 52, 78, 9596,
131, 198, 214, 220
constituent order 2628, 63, 66, 70, 78,
142143, 213, 244245; see also word
order
construction, correlative 28, 60, 62,
6466, 74, 99, 103, 135, 139, 141142,
145, 148, 159, 161, 167, 172, 186, 188,
191, 193, 196, 208
contrast/concession 3, 3032, 4041,
58, 6667, 77, 109110, 112, 114116,
168213, 225, 249, 282285
Subject and name index
292 Subject and name index
coordination 46, 810, 21, 2432,
7677, 83, 8586, 133, 136, 142143,
163, 247248
copia 3, 241242
creole 5, 7375, 156157, 166, 249
deixis 56, 98, 102103, 109, 132,
147152, 157158, 166167, 178181,
247, 249
demonstrative 56, 24, 5859, 7374,
81, 84, 97101, 112, 140, 145, 147, 149,
156160, 166167, 178, 181, 183, 249
discourse marker 1, 3, 4648, 64, 116,
182, 201, 212, 227
Early Modern English 5, 910, 18, 20,
49, 5657, 77, 82, 84, 8689, 91, 102,
109, 111, 113, 132, 178, 181, 217218,
226, 235, 239, 248249
enumeration 37, 4041, 45, 77, 222
224, 282285
fact 21, 39, 107108, 114130, 228,
230232
fiction see text type(s)
French 5, 15, 56, 7375, 8891, 104,
111, 137, 156, 165166, 173175, 183,
217218, 221, 226, 240, 242, 249
genre(s) see text type(s)
German 2, 5, 9, 2328, 3032, 44, 63,
6869, 71, 7374, 100101, 115, 132
133, 135, 137, 139140, 149151, 164,
167, 174175, 179, 181, 198203, 212,
248
grammaticalization 3, 5, 2021, 33, 47,
52, 106, 117, 119, 131, 168, 182, 184
185, 201, 213, 250
Grice, Herbert Paul 117, 128130
Helsinki Corpus 1115, 46, 60, 65, 133,
144
Hume, David 243
iconicity 29, 174, 189, 219220, 245,
250
implicature, conversational 114116,
158, 192, 230
indicative 66, 70, 117, 165, 173
information processing 10, 2832, 45,
132, 138140, 166, 171173, 198, 204
205, 207208, 247249
interrogatives 79, 99, 157158, 229230
Kortmann, Bernd 4, 7685, 106, 111,
154, 248
Late Modern English 911, 68, 72, 77,
84, 9192, 166, 191, 215, 217, 226,
230, 236, 238246
Latin 15, 23, 4956, 71, 7374, 8892,
101, 111, 123124, 132, 137, 149, 154,
160164, 174, 217, 219, 221, 226, 240,
249
Lehmann, Christian 4, 22, 185
lexical bundle see Wortverband
lexicalization 911, 27, 33, 43, 45, 77,
80, 9091, 9698, 100, 102105, 157,
159, 177, 185, 218, 221222, 228, 230,
249
literacy 105, 109112, 158, 249
Locke, John 12, 8, 2021, 57, 214,
233, 241243
Longman Grammar 13, 20, 3536, 40,
4246, 197198, 201, 224, 238
Lowth, Robert 57, 242
main clause 9, 22, 2829, 31, 63, 73,
135136, 138140, 172, 202203,
212213, 249
Mauranen, Anna 68
Nacherstposition 16, 27, 43, 56, 58,
6772, 144, 185, 188, 191, 197, 235
238, 240, 242
New Rhetoric 57, 233246
news see text type(s)
particle 12, 23, 50, 57, 68, 99
Subject and name index 293
particle, discourse see discourse marker
particle, subordinating 5, 74, 91, 102,
142, 160, 166, 188189
perspicuitas 3, 233, 241246, 250
phrase, lexicalized 102105; see also
lexicalization
phrase, prepositional see prepositional
phrase
place see space
post-frst-position see Nacherstposition
pragmatic marker see discourse
marker
preposition 2, 5, 2324, 35, 4951, 89,
101, 154, 159, 173, 183, 228
prepositional phrase 5, 10, 22, 33, 40
41, 43, 45, 47, 77, 80, 9192, 94,
9798, 102105, 122, 127, 134, 140,
144, 147149, 151, 154, 165, 177, 183,
185, 228, 249, 276
punctuation 125, 143, 163, 220, 233,
242243, 250
quantifer, universal 170, 174, 181183,
190
Raible, Wolfgang 5, 2223, 7375, 98,
103, 147148, 156
result see cause/result
rhetoric 69, 21, 57, 78, 131, 136137,
163165, 212, 218, 220, 223, 226,
233246, 250
Royal Society 3, 233, 242243
Scottish Rhetoritians 2021, 57, 242
246
semantic relations see addition;
cause; contrast/concession;
condition; space; time; transition
Smith, Adam 233235, 243
space 21, 3839, 60, 64, 103, 105,
107113, 130, 150152, 158, 166167,
174, 181, 219, 227, 230, 249, 282285
Sprat, Thomas 242243
subjunctive 63, 66, 74, 78, 99, 130, 165,
173, 188, 194
subordinate clause 2732, 6263, 74,
102103, 138139, 157, 172173, 182,
187188, 191, 202, 207213, 248249
subordination 16, 2021, 2532,
7677, 8385, 142143, 173, 187189,
204205, 207210, 245, 247248
text type(s) 1216, 18, 4246, 105, 131
132, 139, 200201, 215, 219, 224226,
229, 233, 237238, 247, 250
text type, academic prose 4246, 82,
92, 105, 131, 197198, 224, 226, 238,
240, 247, 250
text type, argumentative 15, 4246, 65,
121, 133, 219, 224
text type, conversation 4246, 139,
197198, 200201, 207, 238
text type, fiction 13, 15, 4246, 139
text type, narrative 1315, 46, 6566,
119, 239
text type, news 13, 4246
time 21, 33, 3839, 45, 6064, 94, 105,
107110, 130, 150152, 158, 166167,
249, 282285
transition 16, 21, 4041, 4648,
5859, 62, 64, 81, 95, 108109, 112,
114, 116117, 131, 154, 162, 214,
227232, 282285
Traugott, Elizabeth Closs 3, 33, 41,
4647, 52, 117, 119, 122, 127128,
142, 158, 227, 231232
truth 21, 39, 4748, 107108,
114130, 162, 228, 230232, 282285
uncertainty/doubt 228229
univerbation 10, 33, 43, 94, 122, 149,
156, 159, 222, 230, 249, 276
word order see also constituent order;
connector, initial/fnal/medial position
word order, verb-fnal 26, 28, 3032,
6364, 135, 142, 202, 213, 249
word order, verb-second 26, 32, 63,
135136
Wortverband 105, 250