Abstract

Discourses are dynamic things - new information gets communicated, affecting the state of the conversation and the states of minds of the conversational participants. This work explores the question of how much of these discourse dynamics should be accounted for by semantics and how much by pragmatics. There are some philosophers and linguists who claim that the dynamic nature of discourse is good reason for abandoning traditional truth-conditional semantics and adopting instead a notion of semantics that focuses on the level of discourse, treating the semantics of sentences as their contribution to the discourse as a whole, or as their potential effect on the conversational context. I argue that such a semantic explanation is the wrong sort of explanation; we can maintain a traditional, static semantics and explain changes to the context by appealing to pragmatics - broadly speaking, by appealing to the fact that conversations are rational, co-operative activities. In chapter 1, I examine some purported differences between dynamic and static semantics, arguing that the central difference between the two views is whether changes to the conversational context are encoded in the semantics or explained by pragmatics. In chapters 2-4, I look at an extended case study of discourses containing indefinite descriptions and cross-sentential anaphora, such as: A woman walked in. She ordered lunch. The first sentence in this sort of discourse seems to introduce a new object under discussion that the pronoun in the second sentence picks up on. In chapter 2, I argue that there are good reasons to think a dynamic semantic account is not the right sort of approach, and that a pragmatic approach better explains the data. In chapter 3, I extend my pragmatic account to cases of embedded indefinites such as: A wolf might walk in, Mary doesn’t own a car, and If a farmer owns a donkey, he beats it. In chapter 4, I argue against a rival approach that tries to explain the same phenomena by appealing to the referential intentions of speakers.