Population Distances: Biological, Linguistic, Geographical, and Environmental [and Comments and Reply

Abstract

IN THE EARLY DAYS of physical anthropology, population ("racial") differences were assigned entirely, if seldom explicitly, to differences in genetic origin, that is, to ancestral CCpure races," modified only through mixture. It was Boas who demonstrated for man the possibility of environmental response; and biologists were simultaneously recognizing that such shifts could involve either phenotypic plasticity, a pure environmental effect on an unchanged gene pool, or actual Darwinian adaptive selection. Partly because of this, and partly because of their own clear genetic basis, blood groups took primacy of interest away from anthropometry in both descriptive and genetic studies. Here, also, the original assumption was one of stability in population blood group differences, differences whose origin was not explained except by mutation. Only a decade ago did the probable action of selective agents on blood factors, and still more recently the importance of genetic drift, become generally acknowledged. In addition, however, to evidences in some regions of long-time stability and in others of drift, there are also strong signs of nonrandom trends in the ABO system in the decline or loss of gene B, and perhaps of A, in small isolated or marginal populations (America, Oceania, Switzerland -see Benoist 1964 for a West Indian example). This suggests that still other factors, more familiar from mathematical population genetics, may be of real significance in man: population size and effectiveness of isolation, i.e., rate of gene flow (as distinguished from the mere fact of hybridizing). We have become theoretically sophisticated as to the factors italicized above, but we are still by no means advanced methodologically in applying them to the materials of physical anthropology, which are normally derived from a complex historical situation. We gather data first and ask questions afterwards, an old anthropological habit learned from justly revered masters, who adopted it because of the rush of ethnic and ethnographic change. Our publications are apt to consist of a report of the data followed by a discussion section in which these evolutionary factors are browsed among and ruminated, but are not or cannot be submitted to test. Livingstone (1963), using material from the Eastern Highlands of New Guinea, determined intergroup distances or differences in miles, blood group frequencies, and language and got low correlations among these three distances, especially the last two. Thus, if language is used as the index of common ancestry, the latter was not reflected in blood group likenesses, which should therefore, Livingstone concluded, not be incautiously used as a measure of community of genetic origin. That genetic drift may have been important in this case appears from Giles, Walsh, and Bradley (1965; see Cavalli-Sforza et al. 1965; Gajdusek 1965), who showed its significance in differentiating blood types of two clearly related and ecologically similar villages, also in the Territory of New Guinea. Laughlin and J0rgensen (1956) demonstrated the probable effect of drift on morphological characters, using Eskimo crania from Greenland. One can think of other studies of the effect of one particular factor. Relatively few studies attempt to assess several factors at once. Among these is that of Hiernaux (1956), who used measures of "distance" in anthropometric traits (D2) and in blood traits (sum of chi-squares) among a number of tribes in Kivu and Ruanda Urundi to gather evidence relating to environmental effects, persisting genetic differences, and gene flow. This paper is also an attempt to view the operation of several factors, in a case in which controls are not good but which deals with a definable population area, important to Oceanic history: Bougainville in the Solomons. The method is to use biological, ecological,