Emrik comments on Deference for Bayesians