the doable and the impossible

Mike Lieber (U28550@UICVM.BITNET)
Thu, 27 Oct 1994 09:33:44 CDT

What seems to bother the critics of my Trib article most is the demand for
proof of genetic causation of the correlation between race and I.Q.. The
demand is alleged to be impossible. Do scientists not know how to identify
genes and their proteins? No, the impressive results of gene mapping belie
that one. Do scientists not know how to trace a chemical pathway from gene
to phenotypic expression? No, there is plenty of work that does just that.
It is not a matter of the impossibility of doing the work. It is doable, if
scientists can agree on what intelligence means and what nerve structures
control its expression. The question really is whether it is worth doing, and
that is a very complex question. Surely it is a lot easier and a lot cheaper
to keep running correlations. What sorts of research get funded for what
amounts of money is a question of funding agencies' and disciplinary priorities
that may have little to do with the intellectual exigencies of race and
intelligence. I will be very curious to see how much money will be spent on
the research effort that goes into prosecuting and defending O. J. That is
an example of research priorities. I'll bet it will be enough to get a good
start on the kind of research I'm demanding. Now for a contrasting case.

In the 1970s, my older brother, Arnold, was working with Miami police and
learned that they regularly gear up for heavy work at the full moon. Curious
about their certainty, he did some library research and found that the
correlation between lunar phases and homicide rates had been researched before
with negative results. He noted that homicides were plotted on an ordinary
calendar. What he did (with a psychologist) was to take homicide statistics
over a 15 year period for Dade, Cuyahoga counties, and New York City and to
plot them against a lunar calendar. Sure enough, what he found was perfect
preiodicity in 72 hour peaks--the highest peak starting at 36 hours before
a full moon, peaking at the full moon, and then declining for 36 hours after a
full moon. The same thing happened at the new moon, but with a lower peak.
The correlations had statistical significance (though replications using
_different_ statistical tests did not get the same results). The problem
then became one of how to account for the correlation. I won't go into all
the details, but the most sensible hypothesis was that the effect of
"biological tides" on those people most sensitive to them was most likely
changes in the rate of endocrine secretions. This turned out to impossible
to test with the technology current in the mid-1970s. Not that changes in
these rates could not be measured, but that the amount and kinds of
equipment that had to be used was so intrusive to the person tested (who had
to remain still during the testing period) that the physiological responses
to the testing instruments altered what needed to be measured--the classic
Heisenberg effect. The hypothesis remains as interesting and plausible, but
not testable, at least on humans. There has also been some interesting work
on root growth of beans and other plants that shows lunar phase to be highly
correlated with accelerated growth of roots when other factors are held
constant in the laboratory. Perhaps these folks can find a way to test
causal hypotheses that might pose alternative sorts of tests on animals, at
least by way of eliminating other possible causal hypotheses. But for now,
the demands for testing an endocrine rate hypothesis are impossible to meet.
This is not the case for genetic causation for the race and intelligence
correlation.

Mike Lieber