Conditional Degree of Belief

It is a commonplace in epistemology that credences should equal known chances. It is less clear, however, that conditional credences should do so, too. Following Ramsey, this paper proposes a counterfactual interpretation of conditional probability which provides a justification for this equality without relying on the Principal Principle. As a result, we obtain a refined view of Bayesian inference where both learning and supposing have a place