race and I.Q -- demonstrating a causal relation
Mike Lieber (U28550@UICVM.BITNET)
Tue, 18 Oct 1994 15:05:42 CDT
Any first year undergrad learns, or is supposed to learn that a correlation is
not the same thing as a cause. Any correlation can be multi-causal, e.g.,
the birds flying south correlating highly with the children going back to
school. Clearly the birds don't cause the kids to start school, yet the
correlation holds throughout much of North America. Most of what we have as
alleged evidence of a link between race and I.Q. is correlations of test scores
to racial categories. That people called black in the U. S. form a subspecies,
as biologists understand that term, is in question, as Mr. Pflaum has already
shown. That I.Q. measures some "thing" called intelligence has been in
question for decades. But let us say for the sake of argument that there is
consensus on what intelligence refers to and that there is a black race. Then
how could one demonstrate that there is a causal relation between race and
intelligence? What are the minimal requirements of such demonstration?
First, you'd have to identify every gene that is supposed to code for every
character that is supposed to result in a phenotypic trait distinguishing the
population as a subspecies. Then you'd have to identify for each gene the
protein for which it codes. Then you'd have to identify every chemical
reaction that each protein catalyzes. Then you would have to identify each
chemical pathway that each of those reactions take part in, tracing each one
to its final phenotypic expression, which must include the range of possible
expressions for each pathway. Having completed this task, you would next have
to identify every gene that codes for every sector of the nervous system that
is supposed to give the phenotypic expression that we agree is intelligence.
You would then have to follow the same procedure of identification, from gene
to protein to chemical raction to chemical pathway to phenotype that you
already did for the "racial" genes. Then you would have to show that either
the same genes code for both or that the chemical pathways show a union of sets
such that pathways from one intersect with and become parts of the other.
Has Rushton or anyone else made or even attempted such a demonstration?
I've just described the simple part. This is for just one individual. You'd
have to do this on a population scale to get reliable results, and that's
where you'll have to deal with precisely the kinds of genetic variability that
Pflaum is talking about. If a "black" person has mixed ancestry, as millions
of people classified as black do, then what is the range of permissable
variability in alleles at each locus that allow the person to be included in
the sample? What is the range of variability allowable for inclusion as
white and that for black? If a black testee has white alleles (those
determined as allowable for the sample), then does this exclude that person
from the sample? If this is the procedure, then how large a sample population
is possible? How large would the testable black population be? And what can
be determined, if anything, about all those people classified as black who get
excluded from the sample? You have the same problem with whites, too.
Do we have any agreement as to what intelligence is? Do we have any agreement
as to what constitutes environmental variables on I.Q. performance? Do we
have the genetic and chemical data necessary for the minimal demands of
demonstration? Do we know where intelligence is located in the brain? Or at
least, where the intelligence necessary to take the I.Q. test is located? I
suggest that none of the requirements of causal relation are in place, so why
are people talking about this relation as if they were?
This all reminds me of a poster I saw of Hitler and Goehring visiting the
Volkswagon factory in 1943 and looking at a model of the Beetle. The caption
read ADVERTIZING WILL MAKE IT HAPPEN.