Can Bayesian Models of Cognition Show That We Are (Epistemically) Rational?

Abstract “According to [Bayesian] models” in cognitive neuroscience, says a recent textbook, “the human mind behaves like capable data scientist”. Do they? That is: do such models show we are rational? I argue that Bayesian of cognition, perhaps surprisingly, don’t and indeed can’t, Bayesian-rational. The key reason is they appeal approximations, fact carries significant implications. After outlining the argument, critique two responses, seen neuroscience. One can be as approximately Bayes-rational, while other reconceives norms rationality.