Background
Along with the birth of anthropology, the nineteenth century saw the development of semi-scientific to wholly unscientific disciplines, such as anthropometry, craniometry or phrenology. Unquestioningly accepting the prevalent concept of race, some scientists constructed facial and nasal indexes or claimed to measure the skull's volume for every race, of course with the result that the white race's cranium was the most capacious and its owner, therefore, the most intelligent; others went further, insisting that amidst the white race, only the Germans were the "pure" descendants of the "Aryan race" which was destined the rule the earth.
In India, from 1891 onward, Herbert H. Risley, an official with the colonial government, set about defining in all seriousness 2,378 castes belonging to 43 "races," all of it on the basis of a "nasal index." The main racial groups were Indo-Aryan, Turko- Iranian, Scytho-Dravidian, Aryo-Dravidian, Mongoloid and Mongolo-Dravidian.
Unfortunately, this imaginative but wholly unscientific work weighed heavily on the first developments of Indian anthropology; in the 1930s, for instance, B. S. Guha studied skeletons from Mohenjo-daro and submitted a detailed report on the proto- Australoid, Mediterranean, Mongoloid and Alpine races peopling the city, all of them "non-Aryan" of course. Long lists of such fictitious races filled academic publications, and continue to be found in Indian textbooks today.
In the wake of World War II, the concept of race collapsed in the West. Rather late in the day, anthropologists realized that race cannot be scientifically defined, much less measured, thus setting at naught a whole century of scholarly divagations on "superior" and "inferior" races. Following in the footsteps of pioneers like Franz Boas,1 leading scientists, such as Ashley Montagu,2 now argued strongly against the "fallacy of race." It is only with the emergence of more reliable techniques in biological anthropology that anthropometry got a fresh chance; it concentrated not on trying to categorize noses or spot "races," but on tracing the evolution of a population, on signs of continuity or disruption, and on possible kinships between neighbouring populations.
In the Indian context, we are now familiar with the work of U.S anthropologists Kenneth Kennedy, John Lukacs and Brian Hemphill.3 Their chief conclusion, as far as the Aryan debate is concerned, is that there is no trace of "demographic disruption" in the North-West of the subcontinent between 4500 and 800 BCE; this negates the possibility of any massive intrusion, by so-called Indo-Aryans or other populations, during that period.
Die-hard proponents of such an invasion / migration have therefore been compelled to downscale it to a "trickle-in" infiltration,4 limited enough to have left no physical trace, although they are at pains to explain how a "trickle" was able to radically alter India's linguistic and cultural landscape when much more massive invasions of the historical period failed to do so.5 Other proponents still insist that "the Indo-Aryan immigrants seem to have been numerous and strong enough to continue and disseminate much of their culture,"6 but do not explain how the "immigrants" failed to leave any trace in the anthropological record.
A powerful new tool
In the 1980s, another powerful tool of inquiry came on the scene: genetics, with its growing ability to read the history contained in a human body's three billion bits of information. In particular, techniques used in the identification of genetic markers have been fast improving, leading to a wide array of applications, from therapeutics to crime detection to genealogy. Let us first summarize the basic definitions relevant to our field.
In trying to reconstruct ancestry, biologists use two types of DNA, the complex molecule that carries genetic information. The first, Y-DNA, is contained in the Y- chromosome, one of the two sex chromosomes; it is found in the cell's nucleus and is transmitted from father to son. The second, mtDNA or mitochondrial DNA, is found in mitochondria, kinds of power generators found in a cell, but outside its nucleus; this mtDNA is independent of the Y-DNA, simpler in structure, and transmitted by the mother alone. For various reasons, all this genetic material undergoes slight alterations or "mutations" in the course of time; those mutations then become characteristic of the line of descendants: if, for instance, the mtDNAs of two humans, however distant geographically, exhibit the same mutation, they necessarily share a common ancestor in the maternal line.
Much of the difficulty lies in organizing those mutations, or genetic markers, in consistent categories called "haplotypes" (from a Greek word meaning "single"), which constitute an individual's genetic fingerprint. Similar haplotypes are then brought together in "haplogroups," each of which genetically identifies a particular ethnic group. Such genetic markers can then be used to establish a "genetic distance" between two populations.
Identifying and making sense of the right genetic markers is not the only difficulty; dating their mutations remains a major challenge: on average, a marker of Y- DNA may undergo one mutation every 500 generations, but sudden changes caused by special circumstances can never be ruled out. Genetics, therefore, needs the inputs from palaeontology and archaeology, among other disciplines, to confirm its historical conclusions.
India's case
Since the 1990s, there have been numerous genetic studies of Indian populations, often reaching apparently divergent conclusions. There are three reasons for this: (1) the Indian region happens to be one of the most diverse and complex in the world, which makes it difficult to interpret the data; (2) early studies relied on too limited samples, of the order of a few dozens, when hundreds or ideally thousands of samples are required for some statistical reliability; (3) some of the early studies fell into the old trap of trying to equate linguistic groups with distinct ethnic entities — a relic of the nineteenth-century erroneous identification between language and race; as a result, a genetic connection between North Indians and Central Asians was automatically taken to confirm an Aryan invasion in the second millennium BCE, disregarding a number of alternative explanations.7
More recent studies, using larger samples and much refined methods of analysis, both at the conceptual level and in the laboratory, have reached very different conclusions (interestingly, some of their authors had earlier gone along with the old Aryan paradigm8). We will summarize here the chief results of nine studies from various Western and Indian Universities, most of them conducted by international teams of biologists, and more than half of them in the last three years; since their papers are complex and technical, what follows is, necessarily, highly simplified and represents only a small part of their content.
The first such study dates back to 1999 and was conducted by the Estonian biologist Toomas Kivisild, a pioneer in the field, with fourteen co-authors from various nationalities (including M. J. Bamshad).9 It relied on 550 samples of mtDNA and identified a haplogroup called "U" as indicating a deep connection between Indian and Western-Eurasian populations. However, the authors opted for a very remote separation of the two branches, rather than a recent population movement towards India; in fact, "the subcontinent served as a pathway for eastward migration of modern humans" from Africa, some 40,000 years ago:
"We found an extensive deep late Pleistocene genetic link between contemporary Europeans and Indians, provided by the mtDNA haplogroup U, which encompasses roughly a fifth of mtDNA lineages of both populations. Our estimate for this split [between Europeans and Indians] is close to the suggested time for the peopling of Asia and the first expansion of anatomically modern humans in Eurasia and likely pre-dates their spread to Europe."
In other words, the timescale posited by the Aryan invasion / migration framework is inadequate, and the genetic affinity between the Indian subcontinent and Europe "should not be interpreted in terms of a recent admixture of western Caucasoids10 with Indians caused by a putative Indo-Aryan invasion 3,000–4,000 years BP."
The second study was published just a month later. Authored by U.S. biological anthropologist Todd R. Disotell,11 it dealt with the first migration of modern man from Africa towards Asia, and found that migrations into India "did occur, but rarely from western Eurasian populations." Disotell made observations very similar to those of the preceding paper:
"The supposed Aryan invasion of India 3,000–4,000 years before present therefore did not make a major splash in the Indian gene pool. This is especially counter-indicated by the presence of equal, though very low, frequencies of the western Eurasian mtDNA types in both southern and northern India. Thus, the 'caucasoid' features of south Asians may best be considered 'pre-caucasoid' — that is, part of a diverse north or north-east African gene pool that yielded separate origins for western Eurasian and southern Asian populations over 50,000 years ago."
Here again, the Eurasian connection is therefore traced to the original migration out of Africa. On the genetic level, "the supposed Aryan invasion of India 3000-4000 years ago was much less significant than is generally believed."
A year later, thirteen Indian scientists led by Susanta Roychoudhury studied 644 samples of mtDNA from some ten Indian ethnic groups, especially from the East and South.12 They found "a fundamental unity of mtDNA lineages in India, in spite of the extensive cultural and linguistic diversity," pointing to "a relatively small founding group of females in India." Significantly, "most of the mtDNA diversity observed in Indian populations is between individuals within populations; there is no significant structuring of haplotype diversity by socio-religious affiliation, geographical location of habitat or linguistic affiliation." That is a crucial observation, which later studies will endorse: on the maternal side at least, there is no such thing as a "Hindu" or "Muslim" genetic identity, nor even a high- or low-caste one, a North- or South-Indian one — hence the expressive title of the study: "Fundamental genomic unity of ethnic India is revealed by analysis of mitochondrial DNA."
The authors also noted that haplogroup "U," already noted by Kivisild et al. as being common to North Indian and "Caucasoid" populations, was found in tribes of eastern India such as the Lodhas and Santals, which would not be the case if it had been introduced through Indo-Aryans. Such is also the case of the haplogroup "M," another marker frequently mentioned in the early literature as evidence of the invasion: in reality, "we have now shown that indeed haplogroup M occurs with a high frequency, averaging about 60%, across most Indian population groups, irrespective of geographical location of habitat. We have also shown that the tribal populations have higher frequencies of haplogroup M than caste populations."
Also in 2000, twenty authors headed by Kivisild contributed a chapter to a book on the "archaeogenetics" of Europe.13 They first stressed the importance of the mtDNA haplogroup "M" common to India (with a frequency of 60%), Central and Eastern Asia (40% on average), and even to American Indians; however, this frequency drops to 0.6% in Europe, which is "inconsistent with the 'general Caucasoidness' of Indians."
This shows, once again, that "the Indian maternal gene pool has come largely through an autochthonous history since the Late Pleistocene." The authors then studied the "U" haplogroup, finding its frequency to be 13% in India, almost 14% in North-West Africa, and 24% from Europe to Anatolia; but, in their opinion, "Indian and western Eurasian haplogroup U varieties differ profoundly; the split has occurred about as early as the split between the Indian and eastern Asian haplogroup M varieties. The data show that both M and U exhibited an expansion phase some 50,000 years ago, which should have happened after the corresponding splits." In other words, there is a genetic connection between India and Europe, but a far more ancient one than was thought.
Another important point is that looking at mtDNA as a whole, "even the high castes share more than 80 per cent of their maternal lineages with the lower castes and tribals"; this obviously runs counter to the invasionist thesis. Taking all aspects into consideration, the authors conclude: "We believe that there are now enough reasons not only to question a 'recent Indo-Aryan invasion' into India some 4000 BP, but alternatively to consider India as a part of the common gene pool ancestral to the diversity of human maternal lineages in Europe." Mark the word "ancestral."
After a gap of three years, Kivisild directed two fresh studies. The first, with nine
colleagues, dealt with the origin of languages and agriculture in India.14 Those biologists stressed India's genetic complexity and antiquity, since "present-day Indians [possess] at least 90 per cent of what we think of as autochthonous Upper Palaeolithic maternal lineages." They also observed that "the Indian mtDNA tree in general [is] not subdivided according to linguistic (Indo-European, Dravidian) or caste affiliations," which again demonstrates the old error of conflating language and race or ethnic group.
Then, in a new development, they punched holes in the methodology followed by studies basing themselves on the Y-DNA (the paternal line) to establish the Aryan invasion, and point out that if one were to extend their logic to populations of Eastern and Southern India, one would be led to an exactly opposite result: "the straightforward suggestion would be that both Neolithic (agriculture) and Indo-European languages arose in India and from there, spread to Europe." The authors do not defend this thesis, but simply guard against "misleading interpretations" based on limited samples and faulty methodology.
The second study of 2003, a particularly detailed one dealing with the genetic heritage of India's earliest settlers, had seventeen co-authors with Kivisild (including L. Cavalli-Sforza and P. A. Underhill), and relied on nearly a thousand samples from the subcontinent, including two Dravidian-speaking tribes from Andhra Pradesh.15 Among other important findings, it stressed that the Y-DNA haplogroup "M17," regarded till recently as a marker of the Aryan invasion, and indeed frequent in Central Asia, is equally found in the two tribes under consideration, which is inconsistent with the invasionist framework. Moreover, one of the two tribes, the Chenchus, is genetically close to several castes, so that there is a "lack of clear distinction between Indian castes and tribes," a fact that can hardly be overemphasized.
.. if anyone can explain?
