Anne Bezuidenhout

Literal meaning, minimal propositions, and pragmatic processing

Abstract Relevance theorists and others have argued that what a speaker ‘says’ is determined by contextual factors to a greater extent than allowed for by Grice. This dispute has consequences for what Grice called generalized conversational implicatures. Contextualists have argued that the propositions that Grice thinks of as implicated in a generalized way should be treated as part of what is said. Griceans, on the other hand, by insisting that these are conversationally implicated, are committed to postulating more minimal propositions for the role of what is said. This paper examines whether such minimal propositions play a role in utterance understanding. Three pragmatic processing models, which embody some of the assumptions of the rival philosophical views, are postulated. The results of a series of experiments designed to test the predictions of these processing models are reported. Our results lend some initial support to a model in which minimal and contextually enriched interpretations are constructed in parallel, with a bias towards the enriched interpretations. However, the final section of the paper suggests a number of ways in which the issue of the role of minimal propositions still remains open for empirical exploration.

Implicature, Relevance and Default Pragmatic Inference

Grice distinguished between generalized and particularized conversational implicatures. The latter he described as ‘cases in which an implicature is carried by saying that p on a particular occasion in virtue of special features of the context’. The former he characterized as cases in which the ‘use of a certain form of words … would normally (in the absence of special circumstances) carry such-and-such an implicature or type of implicature’ (Grice, 1989, p. 37). Grice did not develop the notion of a generalized conversational implicature (GCI) to any great extent. When he introduces the terminology in his paper ‘Logic and conversation’ he gives a few examples of the following sort:1 (1) A man came to my office yesterday afternoon. (2) Max found a turtle in a garden. (3) Robert broke a finger last night. In the case of (1) the hearer would be surprised to discover that the man was the speaker’s husband, for the use of the indefinite noun phrase ‘a man’ implicates that the speaker is not intimately related to the man. Similarly, in (2) we assume that neither the turtle nor the garden was Max’s own, for if they were, the speaker would surely have used the expressions ‘his turtle’ and ‘his garden’. On the other hand, the use of an indefinite noun phrase does not always implicate the lack of an intimate relation between the subject and the thing indicated by the noun phrase. In the case of (3) there is an implicature that it was Robert‖s own finger that Robert broke.2

Semantics–Pragmatics Boundary

The Gricean distinction between saying and implicating suggests a clear division of labor between semantics and pragmatics. The standard view that a semantic theory delivers truth-conditions for every well-formed sentence of a language has been grafted onto a Gricean view of the semantics-pragmatics divide. Consequently, many believe that truth-conditions can be specified in a way that is essentially free from pragmatic considerations. This view has been challenged by those who argue for pragmatic intrusion into truth-conditional content. Others have argued in favor of preserving a pragmatically untainted conception of semantics, but for a more fine-grained conception of pragmatics. This debate has led to different proposals as to how to draw the boundary between semantics and pragmatics. This philosophical debate has been conducted largely independently of the debate in linguistics about the interfaces between the various subsystems of the language faculty in the mind/brain. However, there are interesting connections between these two debates, and a convergence on the idea of pragmatic intrusion into semantics.

What properly belongs to grammar? A response to Lepore and Stone

Abstract Lepore and Stone devote Part I of their book to setting out a number of views that act as foils for their own positive ‘disambiguation’ view of interpretation developed in Part II. They divide their opposition into three camps: The Gricean rationalists, the neo-Gricean lexicalists, and the empirical psychologists. I try to show why a ‘disambiguation’ view of such phenomena is unappealing and why Relevance Theory provides a better account of these phenomena. I end with some brief remarks about what all of this tells us about the interface between semantics and pragmatics.

The (in)Significance of the Referential-Attributive Distinction

This paper tackles the question as to whether or not the referential-attributive (RA) distinction has any information-structural significance. That is, does this distinction mark a contrast between different ways a speaker might package the informational content of utterances that use definite descriptions? Furthermore, are there any overt markers of this distinction? In this paper I focus on referential and attributive uses of definite descriptions, and leave the issue of indefinites for another occasion. I answer both of the above questions negatively. While definite descriptions have an information-structural role to play, the RA distinction does not. Moreover, after examining several potential candidates for markers of the RA distinction, I conclude that there are no such overt markers. This is in fact to be expected if one accepts my proposal to see referential and attributive uses of definite descriptions as having the same general function, namely to single out something as a center of interest. The difference is simply that referential uses focus on role bearers, whereas attributive uses focus on role properties. Which of these the speaker intends will depend on context and hearers will have to rely on contextual assumptions to recover the intended message.