Course Descriptions

The course will be divided into three modules addressing different but related subject matters. Module 1 (lead: Jeff) will be devoted to the topic of deferred reference. Though there is some disagreement about whether various uses of expressions are instances of deferred reference, everyone agrees that some cases are cases of deferred reference. Here is such a case. We are in a book store and there is a copy of Jane Austen’s Pride and Prejudice sitting on a table immediately in front of us. I point at the book and say ‘She is such a great author.’ Despite the fact that I point at the book, the (deferred) referent of ‘she’ is Austen. This is what makes this a case of deferred reference. The seminal works on deferred reference are a pair of papers by Geoffrey Nunberg: (Nunberg 1993, 2004). Because Nunberg is primarily concerned with deferred reference involving indexicals, we will start the course by reading David Kaplan’s Demonstratives, which outlines the orthodox view of the semantics of indexicals. We will then read Nunberg 1993. Following that, we will look a (unpublished) paper by King on a semantics for ‘today’ and some chapters of a new book manuscript of King’s on deferred reference.

Module 2 (lead: Ernie) will focus on Donnellan’s distinction between referential and attributive uses of definite descriptions. Kripke famously argues that Donnellan’s distinction between attributive and referential uses of expressions commits him to an ambiguity thesis. In “Reference and Definite Descriptions" Donnellan develops a distinction between the "referential use" and the "attributive use" of a definite description. When a person uses a description such as "Smith's murderer" attributively, they mean to pick out the individual that fits that description, whoever it is. The referential use, on the other hand, functions to pick out what a speaker is talking about, so something can be said about it. Kripke rejects the thesis and argues that only one of these readings is semantically encoded, the attributive one. Referential uses, on the other hand, are a matter of speaker reference, not semantic reference; speaker reference is determined by speaker referential intentions, and so, to recover it, we need to uncover the speaker’s referential intentions. Further, according to Kripke, to do so requires us to read the speaker’s mind in any way we can. This suggests referential readings are a matter of psychology rather than of linguistics.

Module 3 (lead: Una) will focus on reference under conditions of contextual uncertainty, ignorance, or underspecification. Joint inquiry requires agents to exchange public content about a target domain, which in turn requires them to track which content a linguistic form contributes to a conversation. But, often, the inquiry delivers a necessary truth. For example, if we are inquiring whether that bird we see over there, Tweety, is a woodpecker, and discover that it is, then our inquiry concluding in this fact would conclude in a necessity, and the form “That (Tweety) is a woodpecker” expresses this necessary truth. Still, whether Tweety is a woodpecker seems a perfectly legitimate, empirical object of study, and the answers we accrue can be informative. But the dominant model of communication--in the tradition of Stalnaker (1975)—treats this situation as linguistically deviant, and diagnoses our ignorance and subsequent discovery as metalinguistic: we were ignorant, and ultimately discovered something, about the meaning of our terms. In this module, we’ll explore an alternative account of inquiry and communication, one that aims to capture how agents can learn something informative about the world—and not merely language—even when inquiry concerns such necessary facts. The key will be to provide a formal model that tracks how moves in discourse contribute public content onto the conversational record, and, crucially, how those moves are connected to one another and to real-world situations they are about. This will allow us to capture that utterances contribute determinate, public content, while representing the information states of the interlocutors who may have only partial access to the evidence and content of the conversation, without making their ignorance metalinguistic. Moreover, we’ll see how this model extends not just to cases of partial ignorance, but also cases of seeming underspecification—cases where facts about the common ground and interlocutors' mental states seemingly fail to uniquely determine a unique content of an utterance.