Hello Dr. Wolpoff,
Thank you for the copy of the Templeton article. It was helpful in shaping my opinions, and I want to show you my thinking about the article.
I wasn't satisfied with Templeton's quick dismissal of the ancient candelabra theory of human origins. I'd like to see a more closely reasoned argument against it. Consider that the African/Melanesian case can argue for the ancient candelabra theory: a dispersal of Homo erectus puts some erectus in a tropical environment far from Africa, and adaptive pressures maintain the frizzy hair and heavily pigmented skin, but do not similarly preserve genetic variables not affected by a tropical environment, which consequently diverge across time.
As I read the article, I gathered the impression that Templeton is among the anthropologists who are promoting a kind of racial agnosticism. His basis is twofold. First, human genetic variance (which I assume was estimated globally and without regard for recent mixtures of relatively pure stocks) between groups is smaller than that usually used to justify subspecies categorization. Second, human gene flow is said to have been too great in the past for pure races ever to have existed.
I'll take the latter point first. A reasonably "pure race" can be created from a non-pure ancestral population in a matter of a dozen generations or so—perhaps within 250 years—provided that the rules for producing a thoroughbred stock, long known to work for animals, are rigorously applied to humans. The steps, roughly, are (1) a wise selection of ancestral group for the prevalence of desired traits, (2) a rigid practice of inbreeding apart from calculated exceptions, and (3) culling of defectives. The characters of the subject population become marked and standardized with time, until the emerging "nation" shows a strong "family resemblance" by which they and others can distinguish their type. The end result is a pure race; a race whose characteristic traits have stabilized and which has been purified of bad recessives. Sparta probably came close to creating a pure race. National Socialist Germany might have done so, had it been permitted the necessary time. And, here and there, now and then, pure races have probably come into existence, though none of them managed to spread far before dysgenic forces, of which careless breeding was paramount, began to destroy the purity of the genotype and, with it, the character of the original people, including their culture (the details of their manner of living). That does not mean, however, that such a pure race, once created, cannot ever so spread. Some choices for the ancestral group are much better than others from the standpoint of cultivating intelligence and other desirable qualities in any pure race to come—and that much remains true whether or not we decide that human differences warrant taxonomic distinction.
Beyond that, I have doubts about Templeton's criteria for rejecting race. The globalized (or globally sampled) "every featherless biped" estimate of human FST=0.156 is not by itself a way by which reasonable distinctions of race or species can be made. This methodology amounts to the deliberate construction of a slippery slope argument. Nobody doubts that clines exist, or that political favor everywhere for multiracial societies has made clines increasingly typical. Clines are where the human mongrels live. The existence of mixed breeds proves only that the breeds had to exist separately before they could mix. If you were testing, say, dogs for race, you wouldn't first engage in a program of greatly enlarging the percentage of mutts before working out the FST of the species. Further, the creation of mongrels of greyhounds and collies would not remove the differences between the greyhound and the collie. If you recalculate the FST and use some sort of statistical filter to omit the cross-breeds, you'll probably find that the resulting value is higher.
Templeton says, "As a result, some combination of characters will distinguish virtually every population from all others. There is no clear limit to the number of races that can be recognized under this concept, and indeed this notion of subspecies quickly becomes indistinguishable from that of a local population...." Or, in other words, if you wanted to resolve humans into races, you would have to accept the idea that there are several thousand races. Actually, there is a biological distinction that covers this level of resolution: the sport. In humans, differences of sport are equivalent to those minor ethnic variations that nobody regards as substantial enough to amount to a difference of race. Among apples, the Yellow Transparent is a sport of the Lodi. The Englishman is a sport of the German.
I offer an analogy. The words "orange" and "green" have no precise meaning. They are loosely associated with wavelength bands within the spectrum of visible light, but their meanings have not been made into well-established, universally recognized conventions. People generally know "green" when they see it, and they never confuse it with "orange," although there might be some dispute over where to draw the line between yellowish green and greenish yellow.
Someone might come along with the argument that the usual color categories are too narrow, asserting that orange is nothing more than yellowish red and that green is nothing more than yellowish blue. But nobody would deny that color is a real sensory impression. Nobody would claim that the existence of yellow proves that green and orange are the same color, or that "color does not exist" because "it's all ONE KIND of light." Nobody would abandon the descriptive use of color for reason that there is no clear limit to the number of colors that COULD be recognized, if the notion of color were resolved down to bands of nanometer or Angstrom width. Someone might argue that a mere sensory impression isn't important enough to distinguish light on the basis of color because, as a practical matter, color makes no difference. But anyone making such an argument would be wrong: color, the sensory impression, is well correlated with physical phenomena that don't depend on subjective interpretations of the brain and that do make practical differences. Consider the photoelectric effect.
The analogy with race should be obvious.
The best test of the importance of genetic differences isn't a statistical summary value of genetic markers, especially a summary value of samples corrupted in regard to the quantity being tested for. Such exercises are no doubt interesting, but the competition between genotypes hasn't been carried on at the molecular level since the molecular replicators figured out how to make cellular organisms. Ever since then, the quality of genotypes has been best judged by the qualities and abilities that they give to their "survival machines": the organisms they create.
If I gave you complete DNA maps for two standard (non-hybrid) strains of corn, could you tell by some sort of statistical methodology which will grow more hardily, bear the greater quantity of grain, and taste sweeter to someone sampling the harvest? Not easily, you couldn't. Maybe not at all. But grow them both, and examine the results appropriately, and the verdict of which corn genotype is the better from the point of view of the major agent for natural selection (you) becomes clear. After having both types of corn on your dinner table, you would know which of them you would rather invest your labor in propagating from that day forth.
Going back to the color analogy for a moment, I alluded to the photoelectric effect because it illustrates a correlation between a continuous phenomenon and the outcome (pass or fail) of a related test. A difference of 20 angstroms in the wavelength of an incident photon will SELDOM change the outcome of a potential ionization event (yes or no). But a difference of 2000 angstroms will OFTEN do so. Similarly, a genetic distance of 20 of Cavalli-Sforza's (x10000) FST units will SELDOM affect whether either of the peoples so distanced will be able to meet and overcome a difficult social or physical challenge to their welfare (live or die). But a genetic distance of 2000 such units might OFTEN do so. Or, as someone else once put it...
"However much the soil, for example, can influence men, the result of the influence will always be different depending on the races in question. The low fertility of a living space may spur the one race to the highest achievements; in others it will only be the cause of bitterest poverty and final undernourishment with all its consequences. The inner nature of peoples is always determining for the manner in which outward influences will be effective. What leads the one to starvation trains the other to hard work." (A. Hitler, 1923.)
Social anthropologists seem to be so fond of the "one human race" idea that they will not apply the simple test of examining the fruits of the garden. Testing corn by taste is one thing, to them, whereas testing humans by performance is another. The difference in motivation is, I think, political in nature. All the genetic statistical analysis they can do probably won't inform them that the Negro brain weight averages 11% less than that of the Caucasoid, or that the Negro cortical supragranular layer is 14% thinner and exhibits significantly less sulcification. Such analyses wouldn't, either, predict the 15 point gap in Negro-Caucasoid IQs measured by tests designed to be culture-free. They couldn't have informed anyone as to which race was more likely to discover the principles of celestial navigation, or of rocketry. I doubt that the gene-statisticians could have foretold that Blacks needing organ transplants generally require another Black to be the donor, or that they could explain from their studies why it is that a physical anthropologist can sort human skeletons by race with very good accuracy in a procedurally "blind" test.
Thank you again for sharing with me the Templeton article.
Jerry Abbott