the logic of conventional implicatures（传统的逻辑含义）.pdfVIP

To appear in Ramchand, G. and C. Reiss, eds., The Oxford Handbook of Linguistic Interfaces. 1 Conventional implicatures, a distinguished class of meanings∗ Christopher Potts 1 A fresh look at an old definition The history of conventional implicatures is rocky, their current status uncertain. It seems wise to return to their source and start fresh, with an open-minded reading of the original definition (Grice 1975) and an eye open for novel support. Suppose the textbook examples (therefore, even, but and synonyms) disappeared. Where would conventional implicatures be then? This paper argues that they would still be widely attested, and moreover that if we move a few years forward from their genesis, we find in Karttunen and Peters’ (1979) multidimensional semantics the basis for an ideal description logic. Conventional implicatures were born into neglect. Early in ‘Logic and con- versation’, Grice advances the term and a definition, but primarily to set such meanings aside. The pragmatic theory of Grice 1975 takes the form of an over- arching cooperative principle and a set of maxims. Together, these help to shape both linguistic and nonlinguistic social interactions. The theory is thus tailored to describing conversational implicatures, a class of nonlexical meanings whose presence and nature are contextually determined and negotiable. In contrast, con- ventional implicatures trace back to individual lexical items and have the force of entailments. General principles of cooperative social interaction are of little help with them. So it is no surprise that Grice closes his passage on conventional ∗My thanks to Ash Asudeh, James Isaacs, Bill Ladusaw, Line Mikkelsen, Geoff Pullum, and Gillian Ramchand for valuable advice on earlier drafts. The work benefited also from discussion with Lyn Frazier, James Higginbotham, Angelik