dumb but doing

Mike Lieber (U28550@UICVM.BITNET)
Mon, 24 Oct 1994 19:04:36 CDT

There's not much I can do about being taken other than saying how I got took,
but I have listened carefully to others on the Net, particularly John McCreery
and Mr. Roe and a few others whom I should name but cannot. I have also
listened to Dave DeGusta and a couple others who did not like my demands for
scientific proof of a genetic basis for race correlating with intelligence.
>From my critics, I learned that demands without clear context for them will
always seem arbitrary and unreasonable, and I thank all of you. From Roe and
McCreery, I determined that if anthropology has an alternative to the
pseudoscience of the Jensens, Rushtons, Hernnsteins, and Murrays of this
world, then the people who most need to hear it are the general public who
pay the taxes that fund public education (that isn't supposed to help the
genetically challenged). I took all of your suggestions and valuable
critiques and turned them into an article that I submitted to the Chicago
Tribune as an OpEd piece. To my surprise, they published it in the
Sunday Editorial section, first page. I am enclosing the copy, so you all
can see some of the effects of your opinions and critiques (compared with the
earlier version of these ideas posted last week).


I was alarmed by this past Sunday's lead editorial on race and I.Q. in Per-
spectives--alarmed because the most obvious and important scientific issues
got lost, as they always seem to do in a race and intelligence polemic. It
seems that every 20 years or so, eugenicists in modern dress, marketers of
what Ellen Goodman used to call "designer genes", start a new round of pseu-
doscientific "proofs" that racial difference causes differences in I.Q.. The
evidence is always the same: correlations in the average I.Q. test scores of
minorty populations, particularly blacks, and whites. Then there is some
argumentation about genetics and intelligence, usually undemonstrated asser-
tions about what percentage of intelligence is inherited in genes vs. the per-
centage affected by environment. We even get correlations between brain
sizes--whites always seeming to have bigger brains (and therefore greater

Even high school kids in many Chicago public schools learn that all a cor-
relation tells you is that as some A changes, some B will change with it.
What a statistical correlation adds to this is to tell us whether the correla-
tion is a coincidence or not. A statistcal correlation is NOT the same thing
as a cause. For example, there is a pretty high correlation between the birds
flying south and the children going back to school. Would anyone argue that
birds' flight causes the children to go back to school? Or vice versa? Now,
that correlation is not accidental, and a bit of historical research will show
that the timing of both is associated with seasonal changes and agricultural
cycles. What a statistically significant correlation tells you to do is to
look for some process that causes both. While we're on correlations, the
brain sizes of Neanderthals were 25% larger on average than those of modern
human populations. And Neanderthals are extinct!

Once you get a statistically significant correlation, then you have to
guess at a cause, turn the guess into a hypothesis that can be tested with
factual data, and then go ahead and conduct the test. There are two possible
causes of the correlation between I.Q. and race--genetic inheritance and envi-
ronment or some combination of the two. Modern eugenicists, of course, cite
genetic inheritance as the major cause of variation in "intelligence" in eth-
nic populations. Now, if this is the hypothesis, then what are the minimal
sets of facts that adequately test it? Let's assume for the sake of argument
that there is consensus on what intelligence is and where it is located in the
brain, though there is no consensus. Let's also assume that whites and blacks
are both biological races, populations that breed only within their own
subspecies and not with one another, which is also untrue. Here's what you'd
have to do to demonstrate a genetic cause for variations in intelligence
between these two populations.


The hard part is getting scientists to agree what intelligence is and to
come to concensus on which brain structures account for it. If you don't know
what you're looking for, it's pretty hard to devise tests for it. Harder yet
is figuring out how to organize the populations to be sampled. Who counts as
white and who as black? The black population of the U. S. is one of the more
mixed ones genetically, including not only African but also European, Native
American, Asian, and Oceanic ancestry. What shall be the criteria for includ-
ing and excluding individuals to be genetically sampled? Do we establish a
percentage quota for non-African genes, and exclude individuals whose percent-
age of non-African genes is greater than that arbitrary figure? How do we
decide the figure? How do we know which genes are African? Or do we use only
some but not all possible genes, say the more significant ones? If we do
that, can we be sure that the genes not sampled do not affect intelligence?
When we get all those sampling criteria established, who is left to sample?
This situation is typical of human populations, by the way. Genetically,
human populations are all mongrels, because people always breed outside their
local group at every opportunity. This is why anthropologists have largely
abandoned the idea of race as a scientific biological concept. It has cul-
tural, social, and political meaning, but it is scientifically pretty useless.
Now comes the easy part--detailed but theoretically doable.

Now that we have a hypothetical population to test, we have to show that
the same genes responsible for racial traits are also responsible for intelli-
gence. This means identifying every gene responsible for each, but that is
only the beginning, and here's why. All a gene does is to code information
about a protein produced in the cell. That protein triggers a chemical reac-
tion in the cell, which may produce a physical trait by itself. For example,
A, B, and O blood types result from a single chemical reaction controlled by a
single gene. Other traits are the result of a chain of chemical reactions,
like skin color, which takes a chain of eight reactions, each triggered by a
protein coded for by a gene. These chains are called chemical pathways.
After you've identified every gene and every pathway for racial traits, you
have to do the same thing for the formation of the nerve nets that control
intelligence. If the pathways for racial traits connect to those for intelli-
gence in a single individual, you can go on to test the entire population. If
not, then there can be no causal relation between race and intelligence, and
you can stop right there. Even if there is a connection of genes and pathways
linking intelligence and racial traits, you may still have a problem with the
population. That's because in any population every gene usually comes in sev-
eral versions (called alleles), and these may produce slightly different chem-
ical reactions. What that means is that the population will be variable in
most traits. When the populations are all sampled, then you have to show that
form and/or the operation of the nerve nets contolling intelligence in each
population are different and that the range of variation between populations
is greater than the range of variation within each population. Are any scien-
tists anywhere near close to offering this kind of demonstration? No. And
this is the easy part.

All of this boring detail is an illustration of what happens when the rig-
orous standards of scientific investigation are applied to the question of
race and intelligence--and this only to the genetic side of the argument. The
environment side is an even bigger scientific quagmire. But if it is to be
called science, then it is these things that have to be done. People like
Murray, Hernnstein, Rushton, and their predecessors like Arthur Jensen, are


all good at manipulating language to give the image of science. Don't be
fooled. Just as a correlation is not a cause, image is not substance. This
difference is particularly important when pseudoscience gets used as a ration-
ale for public policy, like cutting education funding for inner city children.

Michael D. Lieber is associate professor in anthropology at the University of
Illinois at Chicago.