INDEXICALS, SEMANTICS, AND PRAGMATICS

The paper presents commentaries on three claims put forward in Kozyreva's paper under discussion. (1) Kozyreva claims that Kaplan's semantics defines communicatively relevant contexts for situations. I challenge this claim by showing that concepts like communicative situation and communicatively relevant context are semantically irrelevant. Thus the question "What is the communicatively relevant context in a given situation?" lies beyond the scope of Kaplan's semantics. (2) According to Kozyreva, Kaplan's semantics should be modified because it provides an incorrect answer to the question above. I agree with her in that a modification of Kaplan's semantics is needed, but I suggest this need is caused by another reason, namely by the fact that Kaplan's concept of context is too narrow to reflect all contexts involved in everyday communication. In Kaplan, the set {a, t, l, w} (a is an agent, t is a time, l is a location, and w is a possible world) is a context only if a is in l at t in w. Predelli suggests dropping this restriction so that any set containing an agent, a time, a location, and a possible world should be considered a context. In my view, this solution allows us to accommodate contexts involved in answering machine scenario and other situations examined by Kozyreva. (3) Finally, in the light of the received distinction between semantics and pragmatics, the concept of a communicatively relevant context belongs to pragmatics. Kozyreva disagrees: she claims that the received distinction should be revised. In my view, this claim (as well as Romdenh-Romluc's claim cited by Kozyreva) is erroneous because it is based on a confusion of two questions: "How do we understand sentences?" and "How do we understand utterances of sentences?".