Abstract

Philosophers and laypersons typically presuppose an 'Integrationist' conception of cognition, whereby to believe is to hold-to-be-true and a person's cognitive endowment constitutes a single unified, coherent outlook. The folk-psychological world, however, presents many familiar phenomena that sit uneasily with this conception: personas that do not incorporate a person's complete personality, and even conflict; partial viewpoints that are utilized even if deemed unacceptable or false from other viewpoints; and temporary or occasion-based forgetting. Some philosophers have made piecemeal accommodations of these data, but the entire range warrants a thoroughgoing revision of the standard model of cognition---one that recognizes that a person's endowment of representational resources is divided into numerous subsets , which are individually brought to bear according to the perceived immediate pragmatic situation or task. The revised model explains certain kinds of common cognitive error, as well as the introduction and maintenance of inconsistent or contradictory cognitive resources. It also suggests certain impediments to formation of an 'all-in' judgment on any given topic and, moreover, implies a fundamental dispersion of cognitive authority : a person is a composite of differing, and possibly clashing, outlooks, without any single privileged 'I' controlling the whole. Except in a very qualified sense, there is no single authentic self, nor one autonomous intellect that can be elicited by stepping back in earnest reflection. Yet talk about beliefs is I preserved on this model, by revising our conception of belief to relax its tie to truth and by regarding belief talk as summarizing complicated profiles of sentence acceptance across cognizing contexts. Though on this account information-bearing mental representations do exist and exert causal influence in thought, behavior, feeling, and perception, they are not by virtue of this beliefs; beliefhood consists not in some inherent or local property of representations, but in certain culturally recognized patterns of use to which only some of them conform. One consequence is that cognition is conducted at times with sentences that are not held to be true, or are even held upon reflection---and maybe repeatedly---to be false