Gogoa 8 (1):138 (
2008)
Copy
BIBTEX
Abstract
In introducing the papers of the symposiasts, I distinguish between statistical, physical, and evidential probability. The axioms of the probability calculus and so Bayes’s theorem can be expressed in terms of any of these kinds of probability. Sober questions the general utility of the theorem. Howson, Dawid, and Earman agree that it applies to the fields they discuss--statistics, assessment of guilt by juries, and miracles. Dawid and Earman consider that prior probabilities need to be supplied by empirical evidence, while Howson considers that there are no objective constraints on prior probabilities. I argue that simplicity is a crucial determinant of prior probability. Miller discussed how Bayes’s theorem can be interpreted so as to apply to physical probability.