Lexical pragmatics and negotiation of meaning

Abstract

The purpose of this doctoral thesis is to contribute to the study of lexical pragmatics and its instantiation in interactive discourse. In particular, I put forward two new lexical-pragmatic processes that result (1) from the merging of two simple concepts into one concept whose meaning is derived non-compositionally and (2) from the analysis of dialogical discourse. Up to date, the relevance-theoretic account of conceptual meaning has focussed on the domain of lexically encoded concepts and the pragmatically derived concepts that they are taken to communicate in specific contexts. Therefore, it is assumed, along with traditional lexical semantics, that the meaning of lexically-complex concepts, if linguistically encoded, is determined, at least in part, on compositional grounds. However, Pustejovsky's (1991, 1995) generative lexical semantics has shown that the combination of words has predictable semantic effects that can be formalised, and that compositional processes can operate below the level of lexical items. The present thesis adopts a pragmatic perspective to the treatment of lexically-complex concepts, but the influence of Pustejovsky's account pervades most of the theoretical apparatus. While Relevance Theory distinguishes some varieties of concept narrowing and broadening of lexically encoded concepts, nothing has been said about whether narrowing and broadening are the only pragmatic processes involved beyond the scope of lexically complex concepts. The main contribution of the present work to the study of lexical pragmatics is that further processes arise when the pragmatic scope is broadened so as to encompass syntactically and semantically complex concepts, on the one hand, and to see what occurs when these processes are put to work in real conversation, on the other. As it is shown, the picture that emerges is richer and has some far-reaching implications for lexical pragmatics and for the negotiation of lexical meaning in discourse. The description of what I call concept merging as a variety of ad hoc concept construction does not only reflect the fact that the communicated meaning goes far beyond linguistic stipulation, but it also intends to set concept merging as a possible origin for syntactically constrained cases of lexical semantic change, traditionally accounted for by means of generalised implicatures (Hopper & Traugott 1993). What is inherent to concept merging, distinctly from narrowing and broadening, is that there is a shift from two linguistically-specified denotations to only one communicated denotation.