How Many Pragmatic Systems Are There ?

An important component of recent pragmatic theorising is the view that the linguistically encoded meaning of an utterance often falls far short of determining the proposition that a speaker explicitly communicates and that the gap between the two is bridged by highly context-sensitive pragmatic processes. Crucially, processes of ‘free pragmatic enrichment’, that is, processes that are not dictated by elements of linguistic form, mediate the transition from linguistic meaning to explicit propositional content. This ‘contextualist’ position is vigorously defended by Francois Recanati (1989, 1993, 2002a, 2004a) and by relevance theorists (Sperber & Wilson 1986/95; Carston 1988/91, 2002b), among others. In this respect, Recanati and Relevance Theory (RT hereafter) differ from Paul Grice (at least on most people’s reading of Grice), who appears to recognise only a minimal pragmatic contribution in the determination of ‘what is said’: the provision of an occasion-specific value for referring expressions and the selection of a single sense in cases of ambiguity. In his brief discussion of these processes, Grice (1975) indicates that his conversational principles (hence considerations of speaker m-intentions) play no role in them. Recanati calls the pragmatic processes that contribute to determining ‘what is said’ (or ‘explicature’ in RT terms) ‘primary’ pragmatic processes and those that are responsible for the derivation of conversational implicatures ‘secondary’ pragmatic processes. These are useful labels, which I have adopted in my own RT work, but for Recanati they are much more than that: the two kinds of pragmatic processes are of quite distinct kinds, only the secondary ones being properly inferential, the primary ones being more akin to associative processes. Hence, on Recanati’s view, linguistic communication is not fundamentally inferential, that is, addressees can, and standardly do, recover the explicit propositional content of utterances