Indeterminate Probability and Change of View

This paper argues that if changes in confirmational commitment (or changes in prior probabilities) are to be defended without begging the question, indeterminacy in probability judgment is required. But no other modificiation of strict Bayesian doctrine is needed. The resulting qualified version of the Bayesian approach is compared with the way indeterminacy is deployed in the theory of H.E. Kyburg who uses a fixed confirmational commitment, but nonethless, invokes indeterminacy.