Abstract

A complete cognitive science will include generalizations explanatory of human behavior which refer to certain internal states of human agents. We investigate various issues in the foundations of cognitive science arising from this observation. In particular, it is argued that the taxonomic descriptions of behavior which occur in generalizations over behavioral types are intentional, i.e. such descriptions of behavior must respect the semantic contents of the mental states which produce behavior. This principle provides the basis for an argument for the ineliminability of a semantic component from a completed psychological theory. The concept of intentional explanation is examined and it is argued that though behavioral explanation must be cast in the intentional format, intentional explanation ought not to be constrained by normative rationality assumptions. Since intentional explanation, in many cases, requires reference to the mental states of the agents of behavior, it is essential to understand how mental states are to be individuated. We argue that functional criteria are inadequate for the individuation of mental states such as propositional attitudes, e.g. the belief that P, the belief that Q, the desire that R, et cetera. Widely known criticisms of the possibility of a concept of the semantic equivalence of beliefs are examined and rejected and semantic criteria for the individuation of beliefs are offered