Accuracy and Doxastic Bribes

Brian Weatherson (University of Michigan, Ann Arbor)

Brian Weatherson (Ann Arbor) will present "Accuracy and Doxastic Bribes" at 11 in Arts West North Wing 353 on 22 March. 

Abstract: Recently several authors have used the idea that credal states should be accurate, or at least should not have provably avoidable inaccuracy, to motivate constraints on credences. In a classical setting, this mostly amounts to providing a new justification for a well known view - Probabilism. But in a non-classical setting, it helps resolve various questions that rival approaches leave open. This approach, however, is often thought to be vulnerable to an objection from doxastic bribery. It is wrong, say the critics, to deliberately acquire a false belief even if this will lead to many other true beliefs. And, say the critics, the accuracy approach to credences implies this is not wrong. We show this last claim is false. Given some weak assumptions, accepting the bribe makes one overall much less accurate. This is joint work with James Joyce.