Probability logic and conversation

Abstract

The Conversational Hypothesis is examined and it is suggested that a primary application of this hypothesis is to solve the problem of indicative conditionals. It is argued that this application fails, and the extent of this failure is documented. The problem is that the Conversational Hypothesis is predicated upon a notion of truth, and conditionals are not truth-bearing objects. A probability alternative is sketched and its advantages underscored. The alternative, while not problem-free, is shown to have conceptual and theoretical strengths which recommend its adoption over the truth-based paradigm.