Event Detail
Fri Jun 5, 2015 1111 Tolman Hall 12–2 PM |
Richard Samuels (Ohio State) Does Bayesian Psychology Vindicate Human Rationality? |
This talk focuses on the following question: Does recent Bayesian psychological research vindicate the contention that human cognition is rational? Addressing this question in the affirmative turns on resolving two challenges:
-
The In Principle Challenge concerns how best – in light of well-known computational complexity results – to characterize Bayesian norms so that they are, at least in principle, satisfiable by bounded human agents.
-
The Empirical Challenge concerns the extant to which Bayesian psychological research supports the contention that human cognition in fact satisfies relevant Bayesian norms of rationality.
In this talk, I discuss both these Challenges. The first part of the talk provides a partial response to the In Principle Challenge by sketching a broadly Bayesian conception of rational norms for bounded agents. The remainder of the talk focuses on the Empirical Challenge. This is an expansive issue, and I do not pretend to resolve the matter here. Rather, I first set out the challenge in a bit more detail, paying special attention to one major problem for Bayesian claims about the extent of human rationality –the apparently pervasive evidence for probability matching. Finally, I critically assess some of the main ways in which Bayesians might seek to address the problem posed by probability matching. (This is joint work with Shaun Nichols.)