Staking the Claim

My aim in this book has been to argue that the Bayesian approach to epistemology – an approach that seeks to derive epistemological insight from considerations concerning rational preference – makes an important contribution to epistemology. It turns out to be an aim whose realization requires some doing and some undoing. The proponents of the Bayesian approach to epistemology have, in their enthusiasm, burdened it with pretensions it cannot possibly bear. My strategy has been to identify and jettison these unsupportable pretensions and to show that, from what is left, we can still derive results of profound consequence for the way in which we are accustomed to think about (and conduct) the enterprise of inquiry, criticism and justification.

The pages above bear ample evidence of the negative aspect of my strategy – of my dissatisfaction with orthodox Bayesian doctrine. I have maintained that it is a mistake to suggest, as orthodox Bayesians do, that rational agents ought to assign precise degrees of confidence to the hypotheses they encounter in the course of their inquiries. I have argued that Bayesianism's claim to have provided a measure of the degree to which one hypothesis can be regarded as evidence for another is simply false. I have shown why (at least some Bayesians to the contrary) decision theory does not, of itself, impose any diachronic constraints on our states of confidence.