ABSTRACT
RC: Pn(p) = a iff Pn-I(p given e) = a, where e is the evidence on the basis of which the new probability function is formed.50 The evidence e generated the new probability function by an application of Bayes' theorem involving the old probability function, Pn.l . The instance of the theorem used for this purpose might have been:
Pn-I (p given e) = Pn-I (p)Pn-1 (e given p)
As I noted above, this strategy gives only a partial answer to the question asked: it shows us how we can assign a probability value to a statement if other probability values have been assigned. This information is not very helpful in the present context because it does not tell us how basic probability values are determined. Without basic probability values, we cannot even employ the probability calculus. One might suppose, therefore, that the probability calculus cannot possibly help us resolve our basic philosophical problems about experimental inference.
