Quitting certainties: a Bayesian framework modeling degrees of belief

Quitting Certainties is an extremely ambitious treatise on Bayesian formal epistemology. The centrepiece is a modelling framework that can be applied to non-ideal epistemic agents as well as to ideal agents. Unlike many trad-itional approaches to Bayesian epistemology, the framework allows for agents to go from certainty to less-than-certainty in a claim; Titelbaum accordingly calls it the ‘certainty-loss framework ’ (CLF). It is designed to model agents’ degrees of belief in context-sensitive claims (such as the claim that it is now Monday), to allow for the forgetting of factual information (such as informa-tion about what was had for breakfast several years ago), and to allow for ignorance of logical truths (such as complicated mathematical theorems). The content of CLF is inextricably bound up with the methodology behind its application. Throughout the book, Titelbaum stresses the importance of taking an instrumentalist, modelling-based approach to the Bayesian probabil-ity calculus. This approach has points of contact with the attitudes of de Finetti, Jeffrey, Garber, and others, but Titelbaum explores it in unprecedented detail