The implicature theory: a case study

Abstract

Several attempts have been made by direct reference theorists to accommodate the intuitive datum of referential opacity—the failure of co-referential proper names to substitute salva veritate for one another within the embedded ‘that’-clauses of attitude ascription sentences. The theory advocated by Nathan Salmon in his 1986 book Frege’s Puzzle is probably the best worked out version of what is referred to below as ‘The Implicature Theory’. Salmon claims that referential opacity is an illusion brought about by our failure to distinguish the semantic content of attitude ascriptions from their ‘pragmatic impartations’, as Salmon calls them. Regrettably, his work leaves it entirely mysterious how the relevant pragmatic impartations are routinely conveyed. Salmon vaguely suggests that Gricean conversational implicatures are involved. My central claim in this paper is that Salmon is mistaken, since the pragmatic impartations required by his theory do not satisfy the criterion of cancelability, which is met whenever a genuine conversational implicature occurs. The argument in question is, to the best of my knowledge, original with me.