Abstract

"Internalism" is the thesis that a subject's internal physical structure determines which beliefs and desires are properly attributable to him. Internalist theories of intentionality purport to solve several philosophical problems, most notably, how explanation in terms of belief and desire is compatible with subsumption of the subject under physical law. This dissertation argues that internalism is false. First it is argued that an internalistic construal of belief would make it impossible to understand a community's language-mediated division of epistemic labor. A second, more fundamental argument begins with an extensive account of how a subject's beliefs and desires might be inferred from his behavior. Here a number of methodological principles are formulated that must be respected if our attributions of belief and desire are best to serve prediction of the subject's behavior. These principles are used to show that there are no prospects for an internalistic psychological theory and that two subjects who are micro-structurally identical but embedded in physically different sorts of environment might be best interpreted as having different beliefs and desires. Finally, these general arguments are supported by independent criticisms of the internalist theories of Jerry Fodor and Daniel Dennett.