A new paper by Jim Joyce, “How Probabilities Reflect Evidence” discusses more clearly some of the issues I mentioned in a previous post, where I suggested that Henry Kyburg opposed subjective probabilities because he misunderstood some of the ideas of Bayesian epistemology. What I was talking about there looks just like the contrast between what Joyce calls the “balance” and the “weight” of evidence. He also mentions “specificity” of evidence, and shows that Bayesian epistemology can deal with this as well. All three of these distinctions are made in examples where the agent has hypotheses about objective chance processes, but I’m sure that some of this can eventually be generalized beyond those circumstances. Anyway, it’s quite an interesting paper – I was especially intrigued by the mention of some attempts to avoid something like Bertrand’s paradox for the Indifference Principle (which Joyce calls “The Principle of Insufficient Reason”) about randomly generated squares.
(Thanks to Brian Weatherson and Jon Kvanvig for pointing me to the relevant issue of Philosophical Perspectives. I might as well also mention now that I’ll be on a panel about philosophy blogging with both of them, as well as Gillian Russell, at the Pacific APA in Portland in March.)