The brainchild of renowned
population geneticist Luca Cavalli-Sforza, the HGDP is an attempt by human
population geneticists to try and generate interest in "big science" the
way the medical genetics community succeeded with the Human Genome Project.
Cavalli-Sforza reinvented
in the 1960s what used to be called the study of "racial history".
An earlier generation of scientists, such as Harvard's Earnest Hooton,
had been strongly impressed with the reticulate nature of human microevolution,
to such an extent that Hooton drew it literally as a circulatory system.
The computational and statistical breakthroughs associated with "numerical
taxonomy" suggested an application to human genetic markers. Thus,
Cavalli-Sforza and Edwards (1964) presented a dendrogram of the human species.
The diagram became a lot simpler over the two decades since Hooton's figure
from Up from the Ape (1946), which made the figure a lot easier
to interpret.
Only one problem: it was wrong.
Cavalli-Sforza argued
that the major conclusion of this genetic tree was that it linked Europeans
and Africans, and juxtaposed them against Asians. Anatomy, on the
other hand, seemed to link Europeans and Asians, and juxtapose them against
Africans. However, other genetic studies (notably by geneticist Masatoshi
Nei) didn't agree with Cavalli-Sforza's (e.g., Evol. Biol., 14:1-59).
Finally, in 1988, Cavalli-Sforza acknowledged that the genetic result he
had been promoting for twenty years as a result of the tree-building exercise
was wrong
These trees are sensitive
to many things beside divergence: gene flow, the clustering algorithm used,
the markers selected, the demographic history of the populations, etc.
They aren't phylogenetic, although they may look it. But there is
a hope held out that with more populations and more markers we can get
the ultimate microphylogeny of the human species.
That's where the Human
Genome Diversity Project comes in.
The Human Genome Project
was based on the medically-essentialized idea that you could represent
the human species by a thing called "the human genome". Given that
tehre is a "normal" cystic fibrosis gene, a "normal" Tay-Sachs disease
gene, etc., they assumed that one could sequence a "normal" human genome.
This, however, neglects much of the human gene pool. Many genes,
after all, don't come as a "normal" and "disease" type, but as several
normal alternatives -- like blood group genes. I was one of the first
to point this out, in a sarcastic letter to Nature (322:590, 1986).
In 1991, the journal
Genomics
published an article by Luca Cavalli-Sforza, Allan Wilson, Charles Cantor,
Robert Cook-Deegan, and Mary-Claire King, proposing "a worldwide survey
of human genetic diversity." Unfortunately, the proposal contained
a lot of archaic anthropological assertions, such as the idea that studying
the genes of the !Kung San will illuminate "our evolutionary history,"
and that indigenous peoples are genetically isolated. It was accompanied
by great fanfare in the science media, being touted by Jared Diamond in
Nature,
and by Leslie Roberts in Science.
The idea became merged with the
need for immediacy as these indigenous populations are "now" becoming extinct.
After the initial interest,
the HGDP approached some anthropologists to try and sell the idea to the
anthropological community. A number of other anthropologists raised
some constructive criticisms of the HGDP at meetings, but to little or
no avail, for its agenda had been set. It failed pitifully to engage
indigenous communities, it was conceptualized in archaic racial modes,
and when the issue of profiting from the genes of indigenous peoples came
up, the HGDP again failed to take a strong stand. The images it conveyed
were colonialist, exploitative, and racialized -- and it only seemed to
get worse as time passed. It quickly managed to give a black eye
to all serious students of human variation; the support it has engendered
within the community coming tragically from scientists hoping to prosper
from it.
A study of the genetic
diversity of our species would indeed be a valuable undertaking.
But this is poorly problematized, and it seems unlikely to improve.
Below are some more details to explore.