Are Lexical Meaning and Context Sufficient for Determining Indexical Reference?

David Kaplan’s ([Kaplan, 1989b], [Kaplan, 1989a]) philosophical work on indexicals has been influential not only with philosophers, but also with psychologists and linguists (e.g., Miller and Nunberg). Nevertheless, Kaplan’s account of indexicals has several problems. In this paper, I discuss his view that lexical meaning and context are sufficient for determining the reference of what he calls ”pure indexicals” (e.g., ’I,’ ’here,’ and ’now’). He developed this view in his ([Kaplan, 1989b]) early paper, ”Demonstratives,” and retained it in his ([Kaplan, 1989a]) later paper, ”Afterthoughts,” even though the later one contains a substantial change in his view of the role of demonstrations in determining reference.1 In the present paper, I propose to clarify what constitutes the purity of pure indexicals, and to argue that several factors, in addition to lexical meaning and context, are needed for determining indexical reference. I also clarify and criticize Kaplan’s account of demonstrations, and suggest an alternative account (from my [Roberts, 1993a]) of how demonstrations, descriptions, and context function in determining indexical reference. I will not take up Kaplan’s direct reference theory in this paper, because I ([Roberts, 1993b]) have discussed it elsewhere, and because it concerns the contribution of indexicals to propositions rather than the mechanisms by which they determine reference.