Professional Documents
Culture Documents
1. SUMMARY
The reception of interactional studies of natural, routine language use in social
activities from within linguistics has more than often been hesitant and far from
warm. In linguistics it is often claimed that conversation analysis (CA) is trivial,
obvious, unhelpful and theoretically unsound. From adherents of CA, these
criticisms are usually met with the reply that CA has been misunderstood and
misappropriated. They claim that CA is a sociological enterprise, not a linguistic
one, and thus it is concerned with the organisation of social action and not
language structure and function. We will examine this debate to determine to
what extent conversation analysis and linguistics are mutually relevant.
Reference will be made briefly to two areas of study: computational linguistics
and sign language research.
2. INTRODUCTION
It is increasingly recognised that for linguistics “an understanding of the
structures and processes of conversation is . . . fundamental; it is essential to the
understanding of language” (Schiffrin 1990b, p. 10). It is undeniable that
conversation is the primary domain for socialisation and conversation is speech
activity in which all members of a community routinely participate. Also, many
syntactic changes and processes of grammaticalisation have been shown to be
communicatively motivated. For these and other good reasons an interest in the
nature and organisation of conversation is a healthy one. Indeed, this has been
the prime focus of the non-linguistic enterprise known as conversation analysis
for the last twenty years. However, the reception of interactional studies of
natural, routine language use in social activities from within linguistics has more
than often been hesitant and far from warm. Much unnecessary confusion has
arisen in the research literature about the methodology of conversation analysis
and the status of its claims. We feel that a review of the criticisms and
counter-criticisms of conversation analysis is in order and that some perspective
on the mutual relevance of conversation analysis and linguistics be drafted. This
264 MCILVENNY & RAUDASKOSKI
paper undertakes the former and attempts the latter by reformulating the claims
of both disciplines in the light of recent theories of structure and action.
We will assume that the field of linguistics needs no introduction within the
context of these proceedings. However, it will rewarding to give a brief
introduction to the principles and practices of conversation analysis before we
make comparisons.
The early work in the 1960’s has been taken up increasingly in Europe. Indeed,
Scandinavian languages have been the focus of research by Hakulinen (1989)
and Juvonen (1989), among others. Useful summaries of conversation analysis
can be found in Levinson (1983), Atkinson & Heritage (1984), Heritage (1984,
ch. 8), Button & Lee (1987) and Nofsinger (1991). Important recent collections
include those of Psathas (1990), Boden & Zimmerman (1991) and Watson &
Seiler (1992).
Sharrock & Anderson (1987) also argue that conversation analysis has been
misunderstood and misappropriated. They claim that conversation analysis is a
sociological enterprise, not a linguistic one, and thus it is concerned with the
organisation of social action and not language structure and function.
Specific replies to the charges of Power & Dal Martello (1986) and Brown &
Yule (1983) take the following forms. First, the findings of conversation
analysis may seem innocuous and uninteresting for theories concerned with
THE MUTUAL RELEVANCE OF CA AND LINGUISTICS 267
At this point it is relevant to ask why linguistics has shown increasing interest
in conversation analysis, which conversation analysis has seen fit to reject by
maintaining a purity argument. The influence and promise of Sacks’ writings on
the ‘technology of conversation’ may be one such reason. Indeed, we have seen
the progressive formalisation of conversation analysis to such an extent that
Heritage (1989) almost puts forward a rules and units approach. Button (1990a)
has argued against such a position with respect to the nature of computation,
rules and conversation. And Bilmes (1988) has argued that there is only a
268 MCILVENNY & RAUDASKOSKI
These ideas are echoed in the work of Anderson & Garrod (1987). They
demonstrate experimentally that participants in dialogue negotiate referential
meaning through the dialogue itself and can arrive at competing and divergent
meanings.
away from situated use of language (eg. grammatical studies). We will discuss
the first and second possibilities next. A third position - that linguistics cannot
deal with conversational phenomena - will not be mentioned here.
As a warning about the danger of linguists usurping the claims of CA, Button
argues, against Auer (1990), in reference to phonetics that
It is maybe appropriate to start with a quote from John Lyons, who states in
1977:
“reflects a part of the struggle which results from trying to fruitfully combine
the two perspectives on language and language use: grammatical theory and
pragmatics”. (Bolkestein 1991, p. 107)
and,
Fox & Thompson (1990) have looked at interaction and relative clauses, and
claim to show that:
Lerner would not use traditional grammatical terms for syntactic structures, as
he finds them inadequate to explain what is happening in collaborative turn
completions. What he would want to use is the participants’ syntax - an example
of the participant’s interactional orientation to fleeting structures of talk, eg. the
turn, as well as grammatical structures. He claims that:
other: if the criticisms raised earlier are indicative of a general malaise then
more radical change may be necessary. For example, the following quotations
suggest something of a turmoil in the foundations of linguistics.
Recently, Thompson (1990) and Schiffrin (1988) have reopened the inquiry into
how interactive discourse may shape language structure.
Importantly, Schiffrin (1990a) also claims that what we are witnessing now is
the breakdown of the integrity of the notion of ‘language’ and of shared rules
and lexical units. Schiffrin suggests that
We would like to draw together this discussion and note the following major
points that must be taken into account if some sort of alignment or
accommodation of conversation analysis and linguistics is to be undertaken:
ii. language structure can no longer be seen in isolation from situated interaction
and conversation.
iii. the traditional notion of rule and the intersubjective assumption should be
abandoned.
It has also been argued that CA does not provide an adequate detailed account
of how coherence and sequential organisation in discourse is produced and
understood. Computational linguistics has found the systematics of turn-taking
and the organisation of repair problematic if a formal specification for
274 MCILVENNY & RAUDASKOSKI
6 . CONCLUSION
Not only will the study of signed and spoken interaction benefit the field of
linguistics, but also linguistics itself may have to reflect on its own methods and
results as a consequence of interactional findings. If we accept that the
interactional pressures of routine language use in everyday social activity
reproduce, shape and transform talk and thus language, then what new tools
could be used to study the ‘structures of a language’? We conclude that one
must not only study the structure of a language, but also how it is constituted,
maintained and used in real, practical settings that participants routinely
encounter and actively construct in their community. Ultimately, conversation
analysis and linguistics - the investigations of action and structure - are mutually
relevant if they are studied in the primary context of real knowledgeable
participants in situated, routine human activity.
THE MUTUAL RELEVANCE OF CA AND LINGUISTICS 275
REFERENCES