Using Situations to Reason about the Interpretation of Speech Events

Abstract

Situation semantics in its first incarnation in Barwise and Perry (1983) emphasized the importance of reasoning about the subtle interactions between speech events, various informational resources which may contribute contextual information and the content of an utterance. The idea is that the notion of situation in situation theory can be used not only to illuminate semantic notions of content but also grammatical and pragmatic aspects of speech events and context. In the implemented grammar based on situation theory that Massimo Poesies and I have been developing in connection with the FraCaS project,¹ we have attempted to blend this philosophy with the kind of compositionality mechanisms which are associated with Montague’s semantics and which have become standard in many approaches to computational semantics. This chapter discusses some of the issues involved in achieving this interaction between compositionality and reasoning about speech events, contextual information resources and the content of utterances. We will discuss first the way in which the grammar we are developing handles the interaction using a fairly rich variety of situation theoretic objects. We will then sketch an alternative view that would off-load some of the richness of the situation-theoretic universe by exploiting a type-theoretic approach to records.