Abstract

This dissertation offers a theory of rational belief, or acceptance. It is based on the premise that a person's choices or decisions are rational insofar as they may be expected to advance his ends or goals. I argue that the only goals to which a person's beliefs are relevant are his purely cognitive goals. I defend the view that these purely cognitive goals are representable by a utility function, and that the most rational choice for a person to make is one which maximizes the expected value of this utility function. ;The principle of maximizing expected utility has been employed before in theories of rational belief. But the present theory differs from its predecessors in deriving the principle of maximizing expected utility from certain plausible conditions on rational preferences regarding cognitive options. Another difference is that previous theories have assumed that the consequences of holding a given belief depend only on the truth value of the belief, whereas the present theory allows that the consequences of holding a false belief may depend on how "close" this belief is to the truth. ;These two features together provide for a new approach to understanding cognitive aims. It has often been held that the aim of inquiry is to approach the truth, but the difficulty with that has always been to give clear meaning to the notion of distance being employed. I show how the notion of distance from the truth may be defined in terms of the utility function that has been derived from conditions on preference. The result is a subjective theory of verisimilitude