On the existence of archaic racial archetypes

There’s an ongoing debate in the ‘sphere about whether the extremes of human phenotypal expression can be explained better as simple variation in homo sapiens sapiens or as clusters of traits in the hominids which mixed to produce modern man. Proponents of the former say there’s a lot of room in the right tail of the bell curve to accommodate overdeveloped parietal bones, whereas proponents of the latter point to the Paracas skulls and hybrid theory. I’ll make an argument by analogy that says we ought to expect archaic archetypes to exist in trait clusters, then back off the extreme position a bit.

I posted this last week as a simplified version of trait clusters in hybrid theory:

Let’s look at that spectrum from white to black, with African Americans in the middle. Among the hybrid race of African Americans we observe a mean IQ (85) halfway between Europeans (100) and Nigerians (70), using simplified numbers. Among the African Americans with IQs in the >130 range, we observe a higher incidence of European skeletal features because g and skeletal features are both polymorphic and mostly heritable. Sure, there will be some Nigerian-looking people with IQs above 130, because the African American genepool is (roughly) a superset of the Nigerian genepool, and the Nigerian genepool occasionally produces >130 IQs (+4 SD). But the extremely European-looking African Americans will outnumber the extremely Nigerian-looking African Americans in that range by 717:1 (+2 SD vs. +4 SD). So eventually African Americans will learn heuristics for who among them “looks smarter”, and this impression will converge to an image of the archaic European admixture, i.e. an archaic racial archetype.

That’s a pretty simplified explanation, but it should suffice to illustrate why, on the extreme ends of a single trait’s spectrum of expression, we’d expect to find a cluster of traits corresponding to the parent race. Next I’ll back off that a bit by illustrating a dynamic picture of racial mixing.