Plans in the common ground: Toward a generative account of conversational implicature

The focus of this dissertation is the development of a general framework which can serve as the basis for a formal pragmatic (as opposed to computational or processing) theory of one variety of pragmatic data: conversational implicature. From a larger perspective, this framework may also be seen as contributing toward a theory of conversational competence: a theory of the rules and mechanisms necessary to model what people know that allows them to use language in conversation.; Conversational implicature is one of a number of linguistic phenomena initially discussed and categorized in a ground-breaking, informal account by Grice (1975), in order to delineate what kind of data must be accounted for in a semantic theory and what kind could be addressed as a part of pragmatics. The framework developed in this dissertation may serve as the foundation for a formal account of conversational implicature. In it, I provide a more explicit version of Grice's cooperative principle, which includes a formalized "purpose of the exchange" in terms of the plans and goals of interlocutors.; This framework incorporates a representation of utterance context that includes representations of the mutually recognized plans of interlocutors (that is, plans that are believed by all interlocutors to be mutually recognized). In constructing these representations, I have drawn on recent work in Discourse Representation Theory/File Change Semantics, which supports a formal representation of utterance context, and recent work in planning theory, which supports a formal account of plans and their role in communication. In addition, I develop two principles--a Revised Cooperative Principle and a Principle of Cooperative Inferencing--which interact to make predictions that are borne out by linguistic data. Two types of conversational implicature--base and comparative--are formally generated by the Revised Cooperative Principle.; The explicitness of the framework is especially significant in the treatment of conversational implicature, an area in which linguists have been looking for a more explanatory account for some time. However, further work must be done in a number of areas before the framework can be considered to be fully explanatory.