Speaker-Meaning, Conversational Implicature and Calculability

Abstract

It is widely believed that H.P. Grice’s notion of speaker-meaning divides exhaustively into what is said and what is implicated.² Call this ‘Speaker Meaning Exhaustiveness’. And yet, as I have argued elsewhere (Saul 2002) this seems to be in clear conflict with Grice’s famous three-clause characterization of conversational implicature (1989: 30–1), which is also widely accepted. Legitimate worries can be raised, however, about the correct interpretation of the three-clause characterization, ones that may seem to cast doubt on the conflict just described. However, as I will argue here, there is also a conflict between Grice’s Calculability Criterion and Speaker Meaning Exhaustiveness. And the worries that arise for the three-clause characterization do not apply to the Calculability Criterion. We are, then, still left with a conflict between Speaker Meaning Exhaustiveness and key Gricean claims about conversational implicature. At the end of this chapter, I explore ways of resolving this conflict.

I am very grateful to many people for discussion of this paper and related issues, including especially: Kent Bach, David Braun, Wayne Davis, Ray Drainville, Chris Hookway, Larry Horn, and Marina Sbisa. I am also very grateful for my Philip Leverhulme Prize, which helped fund my work on this contribution.