Abstract

In John Perry's 'messy shopper' example, a man in a supermarket sees a trail of sugar on the floor and follows it, intending to inform the mess-maker. When he later discovers that he in fact is making the mess, this discovery seems intuitively to involve a change in belief. The discovery also seems to bring about certain actions, and belief ascriptions made to the shopper can be seen to contain referential opacities. All of these should be accounted for, in a proper treatment of the 'shopper problem', I argue. However, the classical Russellian and Fregean theories of belief seem unable to account for the discovery by assigning different beliefs to the shopper before and after his discovery, in a way which explains his actions. I also survey and criticize attempts by Perry, Chisholm, and Kaplan to modify one of the classical theories of belief in order to treat the puzzle. The approach of Evans, however, and a similar recent approach by Perry and Crimmens seem to offer hope. ;The Crimmens-Perry theory employs ways of thinking of individuals and properties as constituents of attitudes. I describe the advantages these ways of thinking have over those offered by the earlier approaches . I argue that the way an agent has of thinking of herself, her self-notion, plays a pivotal role in connecting her cognitive system to action. Then I apply this to the shopper problem, to yield an account of the structure of the belief which makes the shopper act, and how it does so. I then turn to the opacities generated by belief ascriptions made to the shopper, using the earlier findings to analyze them. Finally, I apply my findings to attempt to give an account of certain 'subject opacities', or cases in which substitution of the noun phrase of a belief attribution appears to affect its truth value, as in Richard's 'phone booth' case