Re: What Are the Race Deniers Denying?
Sun, 26 Jan 1997 12:02:21 -0500

On Mon, 20 Jan 1997 wrote:

> <> wrote:
> >
> > Since you apparently have the book, would you please explain to us what
> > Cavalli-Sforza *means* by race? Just what is he denying. Also, why he
> > weights all factors equally? Furthermore, can he distinquish man from
> > chimp on the basis of gene counting? Or any two species?
> In _Diasporas_, the definition he's critiquing (p. 228)is "...members of
> an animal or plant species sharing one or more constant features which
> distinguish them from other groups within the same species, and which
> can be transmitted to descendants." He's primarily critiquing 'constancy',
> both in the association of traits -- why should certain traits be privileged
> in determining 'races', when others cross-cut those? -- and through time --
> since we have very little knowledge of the temporal stability of almost
> any of these traits. Those are pretty fundamental critiques.

I think I'm beginning to understand. Now why it is that it has taken
months for me to get something resembling a straight answer to my question
and why so many loudmouths keep repeating themselves that there are no
races in humans without addressing my question is not an unimportant
issue. Methinks they just *feel* that denying the existence of races in
human is (somehow) a good thing to say, so they say it. It's a reflection
of groupthink, really. They also suppose that what they call "racism," an
anti-concept if there ever was one, would vanish if human subpoulations
did not meet the biologists criteria of race.

Anyhow, you say that Luigi Cavalli-Sforza is critiquing a *definition*.
This would mean that race is not a viable *concept*, not that there are
races in certain animals and not in others. Or perhaps he means that there
*happen to be* no species anywhere which can be divided into subgroups
each of which share "one or more constant features which distinguish
them from other groups within the same species and which can be
transmitted to descendants." On this definition, there probably are no
races, since I doubt that any feature would be shared only among members
of a specific race: races are not species, whence there will be overlaps.
It is good idea to critique *this* definition! Races are a relative
matter, not an absolute one.

But what is this business of "critiquing 'constancy,' both in the
association of traits -- why should certain traits be privleged above
others-- ...."? Again, is he saying that "critiquing" constancy is being
inapproriately demanded at the subspecies level? Or is he saying that all
traits are necessarily of equal importance? This is which I call the "gene
counter" point of view of molecular biologists. It's not that genes
shouldn't be counted or the counts ignored, it's that this technique is
only a *means* for constructing taxonomies, which ever since Darwin are
guided by the overall *purpose* of discerning descent, that is,
constructing the true evolutionary tree. There is a whole UseNet group,, devoted to sometimes heated arguments as to the
relevant merits of gene counting vs. other methods for achieving the aim
of tree reconstruction. They do indeed have to be assessed, since we very
rarely, if ever, have the entire evolutionary tree against which the
various methods might be assessed. Unfortunately, I do not know enough of
the jargon to participate in the discussion, unfortunate as I have a
pretty high ability to get into an argument and discern what the
*implicit* underlying assumptions are that the participants have made.
(This ability is reflected, I think, in my very asking what it is that the
race deniers are denying.) At any rate, I have slogged though some of the
relevant articles in Elliott Sober, ed., _Conceptual Issues in
Evolutionary Biology_, and have come away with the though that sometimes
one method works best, sometimes another. And, what is not brought out in
those articles but is very much true, the (egalitarian, as it were) gene
counters themselves argue over which algorithms to use to process the raw
data. A 1971 book, _Mathematical Taxonomy_ by Jardine and Sibson, goes
into the various algorithms. This book requires considerable knowledge of
biology (which I do not have) and also of topology, which happily I do.
Cavalli-Sforza's methods (the guy has been around for quite a while) come
in for some pretty rough criticism. And remember that other schools of
taxonomy (these are the pheneticists vs. the cladists) rely on gross
traits rather than gene counts.

Regards the statement that "we have very little knowledge of the temporal
stability of almost any of these traits," this is very, very true.
Taxonomists work with quite imperfect information. But this is a problem
that is pervasive throughout biology, and I fail to see its special
implication for the issue of races in general and in humans in most

> And how do you decide which traits to weight most highly? The ones that
> would probably be of most significance in terms of selection -- intelligence
> among humans, for example -- are for much the same reason among the least
> likely to vary much among humans.

Such authors a Philippe Rushton have claimed at most a ten percent
variation in brain size and a standard deviation in intelligence among the
three to seven largest subgroups of men. Height varies by more than that,
if pygmies are counted as one of the major subgroups, and other factors,
like melatin in the skin are much more varied than that.

The issue remains one of descent: are the pygmies a group that became
*relatively* isolated from other African populations and, over time,
became distinct, not only with regard to height but with regard to a
number of other factors? Well, Carleton Coon, regards them as a subrace of
the Congoids, since they share in common a great many of the traits that
*he* was observing when he made his studies. But he allowed that it is
reasonable to call them a separate race.

I fail to see how Coon was so misguided. What he did not do was to say
very much about the racial characteristics of the peoples of the Asian
steppes, from the Huns to the Turks, or the peoples, loosely, who spoke
Ural-Altaic languages. But he also said that Caucasoids and Mongoloids do
not vary all that much. It's fascinating the finding of mummified bodies
of people who lived 2400 to 4000 years ago i the Tarim Basin region of
Western China that are strongly European in appearance, some resembling
the Irish or Welsh (_NY Times_ science section, 1996 May 7). The language
known to have been spoken later was Tocharian, a member of the
Indo-European family. Now I looked this place up in the [London] _Times
Atlas of World History_ and these people lay on the western part of the
apparent Turkish homeland. (The Huns' homeland was east of that.) And even
more exciting was the find last year of an *apparent* Caucasoid skull in
Oregon or Washington State (I forget which) that is some 9000 years old.
As you probably know, the Indians now living there, but only for the last
500 years, have gotten the Army Corps of Engineers to keep this skull out
of the hands of scientists.

It certainly seems that both Caucasoids and Mongoloids were a
well-traveled bunch. I don't know what Mongoloid skulls have been found in
Finland, but the language is Ural-Altaic (according to the lumpers, but
Finno-Ugric according to the splitters), but (as you pointed out in
another post), the spread of languages and the spread of genes are by no
means perfectly correlated. What does seem to be the case is that they are
sufficiently correlated to serve quite useful *hypotheses* for the
diffusion of the one and the other. So also with those notorious POTS and
for mythologies. (That the Roman gods came from the same root as the
Celtic gods and not just cheap copies of the Greek gods, as the Hellenists
would have you believe, is coraborated and coraborates linguistic
arguments that the Italic and Celtic branches of Indo-European should be
lumped together, much as Slavic and Baltic commonly, but not universally,

It would be nice if we found pottery suggesting such relationships as
these AND if we had skulls, teeth, and other HARD evidence, and DNA
evidence, such as it be worth, from current inhabitants. Of course, pots
and other hard cultureware will not spread exactly and simultaneously with
languages, religions, and other soft cultureware, and neither of them with

So what is all this race denial about?? I'm glad to find a specialist in
African linguistics. So tell me whether the taxonomic disputes are over
matters more fundamental than splitting vs. lumping. I don't think they
are with respect to Asian and European languages (Yiddish aside, which
I'll be touching on in a future post: a *current* minority holds Yiddish
to be Oghuric Turkic in origin, rather than Germanic Indo-European). I
think American Indian languages also tree up rather neatly.

And cultures also seem to do so, at least according to macrohistorians who
spot some 10-30 civilizations in recorded history. The "civilization
deniers" are called "world-systems" thinkers and think that there has been
enough trade flow (cf. gene flow) between civilizations that the globe
started becoming "one" when Egypt and Mesopotamia started fighting each
other (it took them a whole millenium to get around to it: The Old
Testament is largely the story of the hapless Jews that got caught in the
middle) and from there to incorporate China as soon as the slightest
exchange of luxury goods took place. The whole world was united not long
after Columbus. But no one claims that Japanese Buddhism (Zen is one of
them) thereby makes Japan part of Indian civilization. So the world-system
theorists are basically extreme lumpers. I jumped into an e-mail group
devoted to world history, but by the time I had figured out what the
implicit background issues were, the discussion had totally disintegrated
into name calling.

I hope this doesn't happen with this current thread on what the race
deniers are denying, though for the most part it never took off. Most of
the rest of the UseNet groups on which this thread is running are pretty
degenerate. I hope I have succeeded in elucidating a bit better what the
issues are.

That's why we talk about 'populations' -- we
> can define them for the purpose of a particular study. And you certainly
> can distinguish chimps from humans genetically, although the difference is
> only a few percent. One of the main arguments against one of the most
> common racist propositions -- that human races correspond to sub-species --
> is that human genetic variation is significantly less than equivalent
> variation among recognized sub-species of other primates.

I could say, "so what?", but I'll just ask how it came about that other
primates have so much more variation at the gene count level (I think you
mean) than humans.

> > So you are a race denier also. Explain, please, what *you* mean by race
> > and by "biological reality."
> Yup. I'd critique essentially the same definition that Cavalli-Sforza critiques
> -- a limited number of monolithic, well-defined human populations with a
> constant set of physical/genetic characteristics stable through relatively long
> periods of time. Definitely the sub-species concept as well.
> Scott
> ________________
> Scott MacEachern
> Department of Sociology and Anthropology
> Bowdoin College
> Brunsiwck, ME 04011
> -------------------==== Posted via Deja News ====-----------------------
> Search, Read, Post to Usenet