We can draw two larger points from all of this. First, if among group
trait variation is substantially the result of genetic variation
in groups, then human populations are not ‘natural experiments’
(Auerbach, 2007; King, 2010), as their entangled prehistories
render them non-independent from one another.