Abstract

Definite descriptions, when used referentially, are members in good standing of the class of singular terms. Hence, an adequate semantic account of reference in natural language should accommodate referential uses of these expressions. I begin by examining Keith Donnellan's work on this topic; with this examination as background, I develop a view of the distinction that is grounded in communication plans. This view assumes that in accounting for the significance of natural language, one must pay close attention to how it is used to communicate. There are those who criticize views of this type, arguing that they do not establish the relevance of the referential use of definite descriptions to the semantics of natural language. The version of this criticism I consider is due to Gareth Evans; this version is particularly compelling because it also assumes that communication and linguistic understanding provide the foundation for the semantics of natural language, as opposed, say, to the pragmatics of natural language. But even so, this criticism fails to undermine the planning view because it assumes an overly simplistic picture of linguistic understanding. When this picture is replaced by a more adequate one, there is no longer any reason to exclude referential uses of definite descriptions from communication-based semantic theories