Relevance and conversation 1

This paper presents a general argument for a relevance-based approach to the problem of connectivity in conversation. First, the potential relevance of conversation for pragmatic theory is discussed, as examples in pragmatic theory generally consist of small conversations. A typical problem for conversation analysis and pragmatic analysis is the description of uses of pragmatic connectives. The French connective parce que is described using first a model of coherence (Ducrot’s theory of argumentation) and second an inferential approach. With a view to improving these models, conversational uses of parce que are presented, which show some difficulties which coherence-based approaches face in attempting to make the correct predictions about common uses of connectives in conversation.This is explained in terms of, first, an implicit theory of conversation underlying Ducrot’s theory of argumentation and second, relevance theory. The less discourse oriented approach (relevance theory) is better-equipped to account for conversation in general, and to solve the sequencing problem and the interpretive problem in conversation in particular. 1 Special thanks are due to Billy Clark and Deirdre Wilson for rewriting my French-English pidgin and for their relevant suggestions ! Many thanks to Neil Smith and Deirdre Wilson who endured a spoken version of this paper at UCL and have invited me to prepare a written version of it.