Abstract

Chapter one examines four characteristics of ordinary explanations of actions : their causal character, their independence of general laws, their ability to justify actions, and their use of sentences expressing psychological attitudes. I maintain that these characteristics should either be preserved in scientific explanations of behavior or some account of their presence in ordinary explanations should be given. In chapter two I present a composite physicalist approach to these characteristics which, because of a commitment to the unity of science, recommends reexpressing rational explanations as deductive nomological explanations and maintains that such explanations will eventually be reduced to the explanations of some basic science. In chapter three I show why this approach failed. I also discuss and reject the eliminative materialist's treatment of rational explanations. In chapter four I argue that token physicalism provides a way to reject both the unity of science thesis and the previous attempts to understand rational explanations while avoiding a return to metaphysical dualism. I explain and defend Davidson's version of token physicalism, showing why his interpretation is preferable to interpretations which treat token physicalism as a scientific hypothesis. In chapter five I return to the topic of rational explanations and discuss two perspectives on the significance of attributions of psychological attitudes. Daniel Dennett considers such attributions to be part of a pragmatic strategy for predicting behavior called 'the intentional stance'. Jerry Fodor believes that these attributions specify internal representations of a subject's psychological attitudes which constitute the causal antecedents of that subject's behavior. In chapter six I show that these two perspectives give rise to two accounts of rational explanations: strong instrumentalism and strong realism. After discussing how each deals with the four characteristics of rational explanations, I explore some considerations which establish that each is inadequate. I conclude by comparing Dennett's and Fodor's most recent, more moderate positions, and showing that each allows a place for rational explanations in scientific theories of behavior