Perry argues that certain sorts of indexicals are 'essential', in the sense that they cannot be eliminated in favor of descriptions. This paper also introduces the influential idea that certain sorts of indexicals play a special role in thought, and have a special connection to action.

Demonstratives seem to have posed a severe difficulty for Frege’s philosophy of language, to which his doctrine of incommunicable senses was a reaction. In “The Thought,” Frege briefly discusses sentences containing such demonstratives as “today,” “here,” and “yesterday,” and then turns to certain questions that he says are raised by the occurrence of “I” in sentences (T, 24-26). He is led to say that, when one thinks about oneself, one grasps thoughts that others cannot grasp, that cannot be communicated. However, (...) nothing could be more out of the spirit of Frege’s account of sense and thought than an incommunicable, private thought. In the first part of the paper, I explain the problem demonstratives pose for Frege, and explore three ways he might have dealt with it. I argue that none of these ways provides Frege with a solution to his problem consistent with his philosophy of language. The first two are plausible as solutions, but contradict his identification of the sense expressed by a sentence with a thought. The third preserves the identification, but is implausible. In the second part, I suggest that Frege was led to his doctrine of incommunicable senses as a result of some appreciation of the difficulties his account of demonstratives faces, for these come quickly to the surface when we think about “I.” I argue that incommunicable senses won’t help. I end by trying to identify the central problem with Frege’s approach, and sketching an alternative. (shrink)

This anthology of essays on the work of David Kaplan, a leading contemporary philosopher of language, sprang from a conference, "Themes from Kaplan," organized by the Center for the Study of Language and Information at Stanford University.

In this paper, I shall defend Russell's view that Mont Blanc, with all of its snow elds, is a component part" or constituent of what is actually asserted when one utters Mont Blanc is more than 4000 meters high," and of what one believes, when one believes that Mont Blanc is 4000 meters high. I also claim, however, that a proposition that does not have Mont Blanc as a constituent plays an important role in the assertion and the belief that (...) Mont Blanc is more than 4000 meters high. Taken somewhat out of context, the quotes from Frege and Russell express insights that pull in di erent directions in the contemporary philosophy of language. Behind Frege's remark is the insight that reference is not direct but mediated. When we think and talk about things, our thoughts and words are not in direct contact with the things thought about and talked about. The meanings of our words, and the cognitive roles of our ideas, do not in and of themselves determine their reference. They get at objects via some aspects of.. (shrink)