Matthew Tromp comments on Deference for Bayesians