claims, one of which is clearly wrong … The clearly wrong claim is that one would expect persons who are boundedly rational to behave like cultural evaluators just because they are boundedly rational. It is indeed well established that people conform their factual beliefs both to the apparent view of others (through mechanisms such as “group polarization,” “reactive devaluation,” and “naïve realism”) and to their own values (through mechanisms such as “biased assimilation” and “defensive motivation”). But these dynamics don’t tell us which group commitments (professional or geographic, political or socio- economic) or which values (ideological, religious, aesthetic, etc.) will exert this impact on belief formation. They thus furnish no explanation for any particular distribution of beliefs across persons or issues, and no explanation, in particular, for why beliefs are in fact distributed in ways that express persons’ commitments to hierarchic and egalitarian, individualistic and communitarian worldviews. The most plausible way to make sense of these patterns of belief is to view culture as prior to the cognitive processes through which people perceive facts. … Bounded rationality, then, does not explain why people behave like cultural evaluators; on the contrary, the disposition of people to behave like cultural evaluators explains why established mechanisms of belief formation – social influences, biased assimilation, the availability heuristic, probability neglect, affect, etc. – generate the distinctive array of beliefs that boundedly rational people actually hold.

I find that explanation persuasive. My concern with cultural cognition theory is not with its explanatory model but with its normative implications, at least as implied by Kahan and Slovic, who claim that in a democracy people are entitled to their values and that certain factual beliefs are in a very direct sense expressions of such values. As such, both the values and the factual beliefs are entitled to some normative weight. Nonsense, says Sunstein. Incorrect factual beliefs have no “normative weight,” even where they are expressions of values. That, in my view, is obviously correct. Truth is not about counting votes or respecting people’s values and prejudices. Truth is about underwriting factual claims with the prevailing opinions of a specialized scientific community that follows certain public procedures.