PRETEND GENES 67question, to be causal. However, even if we accept those at face value,
there are more serious prob lems. As with the heritabilities arising from
twin studies, the results are virtually inevitable because they are hope-
lessly confounded with environmental infl uences.
Th e biggest prob lem is what is called “population structure.” Pairs of
seemingly unrelated people may share more genes than average, because
they are descended from the same remote ancestors, who have remained
(more or less) in the same social strata, even though not from the same
immediate family. So ge ne tic resemblances will again be confounded
with resemblances in rearing environments. Th e method makes some ef-
fort to correct for such a prob lem, but only at a superfi cial level. As Evan
Charney noted in an article in In de pen dent Science News, there is further
cryptic ge ne tic relatedness “omnipresent in all populations and it wreaks
havoc with assumptions about ‘relatedness’ and ‘unrelatedness’ that cannot
be ‘corrected for’ by the statistical methods [devised].”^37
Th is prob lem arises because of another fatal assumption— that human
socie ties can be treated as random breeding populations with equally
random distributions of ge ne tic material in randomly distributed envi-
ronments. All human populations, however, refl ect continuous emigra-
tion and immigration. Th ese are usually associated with persecution or
the search for jobs. Immigrants with related ge ne tic backgrounds tend
not to disperse randomly in the target society but become associated with
diff er ent classes or strata that are also, for entirely noncausal reasons,
associated with diff er ent IQs. Examples include the infl ux of Huguenots
to southeastern Eng land to become weavers, the infl ux of Irish to work in
Northumbrian coal mines and Yorkshire and Manchester mills, the in-
fl ux of Scots to work in Midlands steelworks and car factories, and large
numbers of Rus sian and Eu ro pean Jews fl eeing po liti cal and racist perse-
cutions to become tailors and retailers.
Again, these genes will have nothing to do with cognitive ability (or
mostly anything else of importance). But their coincidence with social
class has created an entirely noncausal correlation between social class
and ge ne tic background. Moreover, because social class mobility is lim-
ited and controlled, correlations among genes, social class, region, and
cognitive characteristics will persist across many generations. Fi nally,
because IQ tests have been constructed to correlate with social class (see
This content downloaded from 139.184.14.159 on Tue, 17 Oct 2017 13:51:26 UTC