Abstract

A number of philosophers continue to argue, inthe spirit of Keith Donnellans classic paperReference and Definite Descriptions, thatthere is more to the semantics of definitedescriptions than Russells theory predicts. If their arguments are correct, then a completesemantic theory for sentences that containdefinite descriptions will have to provide morethan one set of truth conditions. A unitaryRussellian analysis of sentences of the form`the F is G would not suffice. In this paper,I examine a recent line of argument for thisanti-Russellian conclusion.Unlike earlier Donnellan-style arguments, thenew argument does not rely upon the mereexistence of referential uses of definitedescriptions or the possibility of conveyinginformation about an object by misdescribingit. I argue, however, that we ought to rejectthis new line of reasoning and rest contentwith a Russellian theory of definitedescriptions.