torekp comments on Knightian uncertainty in a Bayesian framework