Re: Rushton, why?

Mon, 31 Oct 1994 01:51:00 PST

Rushton may be wrong, but challenging him via arguments that are on the face
of them incorrect hardly shows that he is wrong; e.g., Russell writes:

"... I am quite troubled by the fact that
you plan to proceed examining race as a channel through which genetic
information is spread ..."

Since (emic) races are also the basis for much mate selection (i.e., most
marriages are not random with respect to (emic) races), it follows that race
is a channel through which genetic information is spread.

and further:

" .. how can you possibly try to link an inherited trait (as you're
convinced IQ is) ..."

I take it that the parenthetical comment is inserted to express the author's
disagreement with the assertion that some portion of intelligence, however
imperfectly measured by IQ, has a genetic component and hence is inherited
via genetic transmission. Clearly, individuals differ in intelligence
(the phenotype) not solely by virtue of environmental conditions, but also by
virtue of variation at the level of the genotype (regardless of environmental
conditions, I could never be an Einstein, a Mozart, a Shakespeare, a Fermat,
etc.). The genotype is "inherited"--what is less clear is the extent to
which the genotype constrains the phenotype.

It should be noted that whereas Jensen (and possibly others) have argued that
a 15 point difference in mean score for IQ could not be obtained via the
environment if the heritablity of intelligence is around 0.7, a problem
arises when this argument is applied to the purported 15 point spread
between blacks and whites. The problem is that the heritability argument
requires assuming (at least) (1) no systematic bias in the measure of
intelligence and (2) no environmental/IQ score interaction.

Heritability estimates are based on variances (h-squared = genetic variance
/total variance) and if there is systematic, additive bias (e.g., blacks
systematically score 15 points lower by virtue of past and present
conditions), this additive bias would have no effect on variance estimtes.
That is, one can have simultaneously a high additive bias and a high
heritability estimate. Thus a heritabilty of around 0.7 says nothing about
potential change in black mean IQ scores UNLESS one has FIRST established
that the IQ test is unbiased and not affected by environmental conditions.

Interaction could enter via high intelligence persons being less affected by
an academically poor environment (e.g., bright kids tend to do well despite
poor teaching) while individuals with average to below average intelligence
could be (negatively) affected more by an academically poor environment that
brighter kids. If such is the case it follows that two cohorts, otherwise
equal in average intelligence, could end up with a much greater difference
than would be expected from merely considering heritability of intelligence
if one cohort experiences a largely "good" academic environment and the
second cohort experiences a largely "poor" academic environment.

D. Read