Abstract

I discuss the role played by speaker intentions in determining the reference of demonstratives and definite descriptions. ;In the first section I examine contextual, intentional, and quasi-intentional views of demonstrative reference. According to the first, the demonstratum of a demonstrative expression is determined entirely by certain publicly accessible features of the context. According to the second, intentions play a "criterial" role in the determination of the demonstratum; contextual features have no more than a pragmatic significance. According to the third, intentions play a limited role in the securing of the demonstrata. I argue that the third view is the only one of the three with any plausibility. ;In the second section I look at the certain theory of the definite descriptions: the "maximal salience theory." According to this view an expression of the form "the F" denotes the most salient of the F's in the contextually delimited domain of discourse. After motivating the maximal salience theory, I draw attention to certain counter-examples to that theory. I then present Lewis' recently revised version of the maximal salience theory. I argue that the revised version fails in the end, as it fails to take in consideration the semantic significance of referential intentions. ;In the third section I look at Kaplan's attempt to extend his recently advocated intentional view of demonstrative reference, to the so-called "referential use" of definite descriptions. Kaplan suggests that referentially used expressions of the form "the F" be parsed at "that, the F," where the appositive description functions as a kind of demonstration, and is thus of merely pragmatic significance. The referential intention of the speaker is what does the semantic work. I concede that such a view has some initial plausibility when applied to cases where the description is only slightly off-target. However, I go on to argue that the view has no plausibility when applied to cases where the description is very wide of the intended mark. I then sketch two alternative accounts of referentially used descriptions, neither of which assigns a "criterial" role to referential intentions