A Cognitivist View on Genetic Factors

Bill Cosby is terrifying.


The Coz is gonna cut you down (?)

Schrieber’s finding that people seem to be terrified of beloved comedian Bill Cosby is the central question from the “Race and Fear” presentations. It implies that something is essentially wrong with how we are describing our findings. To me it argues for the existence of “schema,” a concept from cognitive psychology that we have not addressed in our more evolutionary approaches. This oversight is unfortunate because there is a tremendous overlap already between cognitive and evolutionary psychology, and would be relatively simple to incorporate.

A “schema” is a constructed mental model or semantic meaning. For instance, “genetics,” “genomics,” “evolutionary psychology,” and “structural equations” are all part of the schema for this class. Thinking about one makes it more likely to think about others. There are linear relationships in schemata. For instance, vaguely smelling pie may vaguely recall something, but being suddenly hit by a distinct smell may bring on a distinct memory…


Schemata offer an explanation for the horror that is Bill Cosby’s visage. “Fear” and “Black” appear to be schematically related. Bill Cosby’s face, being highly valence, loads its schema-members with a similar high valence. Likewise, a black face with low valence (say, any random black UNL student) should bring about its schema-members with low valence.

Our biological factors training, however, lets us improve on the cognitivists’ schema theory. We recognize that conscious and rational thought are only a portion of behavior, and hear evidence that rational thought is much closer to after-the-fact rationalization of our beliefs than the before-the-fact rationale for our beliefs. In particular, we are skeptical of self reports, “good reasons,” and the rest of the SSSM carcass that still clings to some of that domain.

Such as multischematic view can look to Craemer for support. His theory of shared and idiosyncratic racial attitudes amounts to a theory of two racial schemata: one built on unique experience and one built on social interaction. His statement that there is “no rhetorical model” (2) for multiple schemata makes no sense from a cognitive perspective, because cognitivists have long focused on individual domains of learning. Society is a separate domain from interpersonal actions, and thus there is no reason to believe that learning in one will substantively effect the other.

An immediate consequence of multischema view of the mind that IATs are challenged because the very concept of a domain-general “attitude” is challenged. Not only may someone have “racist” social attitudes but “non-racist’ interpersonal attitudes, these concepts may change from domain to domain. Perhaps IATs, which occur in domain-general settings, only measure the greatest valence of an “attitude” across any domains. This would seem to be supported by the Kurzban research (“Can Race be Erased”) we went over earlier. Humans are small-group-centric animals, and by changing the membership of small-group coalitions Kurzban was able to completely throw off the valences.

A more specific test of this theory can be devised. First, give a battery of IATs designed to test “racist” “attitudes” across several domains. For instance, one focusing on writing, one focusing Cosbyesque amygdala response, one focusing on economic gameplay. Then manipulate racial interaction in just one domain while leaving the others constant. So give the subjects “neutral” material that would reinforce existing beliefs in every domain but one: for example, everything remains the same except a black behaves somewhat unfairly against the subject in the economic game. Then re-run the IATs. The results should be constant except in the manipulated domain, in this case the economic game. There could also be another condition where the black behaves very unfairly. This high valence behavior should leak out to other domains.

Quite possibly, one may be able to develop a mathematical model of IAT mismeasurement: The Measured Implicit Attitude of a Domain is the Actual Implicit Attitude in that Domain added to the sum of the product of the implicit attitude of each domain by the schematic overlap of the domains. Or, MIA(Domain) * AIA(Domain) + Sum(For every other domain ‘OD”,IA(OD) * Schematic Overlap(Domain,OD)). Is this formula true? Probably not – I don’t think that many breakthroughts are made in weekly reaction papers. But it invites definition of concrete terms and, once operationalized, makes predictions open to falsifiability.