Years ago, I spent a big chunk of my intellectual career studying the rationality of disagreement, mostly via math modeling, but also some lab experiments. My main conclusion was that, for the purpose of accurate beliefs, it seems both desirable and feasible for people to not know...