1: \begin{abstract}

2: To the frequentist that computes posteriors, not all priors are 

3: useful asymptotically: in this paper Schwartz's Kullback-Leibler

4: condition is generalized to prove four simple and fully general frequentist

5: theorems on the large-sample limit behaviour of posterior distributions,

6: for posterior consistency; for posterior rates of convergence; for

7: consistency of the Bayes factor in hypothesis testing or model

8: selection; and a new theorem that explains how credible sets are

9: converted to asymptotic confidence sets. Proofs require the

10: existence of a Bayesian type of test sequence and priors that give

11: rise to local prior predictive distributions that satisfy a weakened

12: form of Le~Cam's contiguity with respect to the data distribution.

13: Results are applied in a wide range of examples and counterexamples.

14: \end{abstract}

15: