Or maybe it could be that, given their prior beliefs, the reactions to your arguments are, in fact, _rational_ in a Bayesian sense? I think, with Bayesian prob

Or maybe it could be that, given their prior beliefs, the reactions to your arguments are, in fact, _rational_ in a Bayesian sense? I think, with Bayesian probabilities, you can have a setup where, given the same data and two different prior distributions, you end up with two wildly different posteriors after updating the initial beliefs. Unfortunately, i don't have an example at hand but i remember there is a very interesting passage on this phenomenon in E. T. Jaynes "Probability Theory: The Logic of Science". Can anyone else expand on this?