Throughout his career, Isaac Levi has advocated a unity of reason thesis according to which practical deliberation and theoretical inquiry, while differing as regards values and goals, are nonetheless similar from a structural point of view. In support of this thesis he has argued, first and foremost, that problems of induction can be seen as decision problems analogous to practical ones; both kinds of reasoning can be represented within the framework of Bayesian decision theory. In the practical case, the decision maker's options are practical actions. In the theoretical setting, they are, Levi thinks, acts of accepting potential answers to the inquirer's question. In this chapter I argue that his latter view exposes him to criticism unnecessarily.

LEVI'S COGNITIVE DECISION THEORY

Suppose that the agent is in an initial state of full belief K. There is a question that the agent wants to answer, and so he has identified a set of relevant propositions exclusive and exhaustive relative to K in the manner proposed by Levi in his first book from 1967 and defended in later works. This set is his ultimate partition U. Given U, a set of potential expansions relevant to the inquirer's demands for information can be identified. These potential expansions or, as Levi also calls them, cognitive options are formed by adding to K an element of U or a disjunction of such elements. The result is then closed under logical consequence. Levi assumes that adding some belief contravening proposition is also a potential expansion.