THE ARYAN WARS
A HISTORICAL DEBATE GETS POLITICAL AGAIN AS NEW GENETICS RESEARCH SUGGESTS OUR ARYAN ANCESTORS CAME IN FROM THE WEST OVER 4,000 YEARS AGO
Where did our ancestors come from? This can be a highly emotional—and political—issue. It is clearly an important question for many people and one that modern genetics has answered to a great extent. In recent months, a scientific paper published in the journal BMC Evolutionary Biology has sparked a heated controversy in the Indian media by outlining an ‘Indo-European expansion, with an ultimate source in the Pontic-Caspian region’ into the Indian subcontinent.
Behind the gently arcane scholarly language, the paper argues that the genetic ancestry of all modern Indians displays evidence of significant mixing with populations that moved to the subcontinent from northern Iran and the Caspian region some 4,000-5,000 years ago. Tempers are fraying because these findings have reignited a long-simmering and highly politicised debate about the ancient origins of the Vedic ‘Aryans’. For decades, historians, linguists and archaeologists have debated the relationship of the Aryans to India, with little resolution—indeed the argument has been subsumed by its implications for the contemporary battles between the hyper-nationalist politics associated with the Sangh parivar and the secular liberal opposition. Many Hindu nationalists are uncomfortable with the idea that the ancient roots of the Indian peoples may lie outside the sacred geography of the subcontinent. Meanwhile, there is often an element of liberal schadenfreude in embracing the narrative that suggests a parallel between the Vedic Aryans as conquering ‘invaders’, not unlike the later visitations of Islamic and European empires.
Over the last few years, genetics has begun to offer its own findings, which much more definitively indicate that a people—who may have been Aryans—moved into the subcontinent approximately 4,000-5,000 years ago. Soon, the field of ancient DNA research may close the case. The BMC paper that has drawn so much interest and ire is just the latest in a line of research that goes back decades, and grows more precise and insightful with every new technological advance. By looking at patterns of genetic markers in modern humans, geneticists have been able to sketch the family tree of our species. Researchers are also testing genetic material from remains tens of thousands of years old. It has already highlighted the Neanderthal heritage extant today in all humans outside of Africa.
Because genetic science has been driven by US-based researchers, biases have crept into the sort of questions asked. But the democratisation of the field, due to a surfeit of data, is now enabling exploration of more diverse topics, including questions related to the Indian subcontinent.
Eurocentric ideologies have spawned their own counter-theses. While, in 1903, the Indian nationalist Bal Gangadhar Tilak could write The Arctic Home in the Vedas, in keeping with the migrationist beliefs of the era, today, Indophilic westerners such as Koenraad Elst are promoting an ‘Out of India’ theory, an inversion of the older route.
These arguments exist outside of politics and nationalism. The author Sanjeev Sanyal contends that “the idea of a unidirectional ‘Aryan’ invasion or migration around 1500 BC is now conclusively proven to be wrong”. As outlined in his book, The Ocean of Churn, Sanyal emphasises the reciprocal movements of peoples. He notes that archaeology does not tie any Indus Valley civilisation site to Central Asia, and the Vedas themselves seem ignorant of geography outside of South Asia.
Ten years ago, one could reasonably support Sanyal’s suppositions from a genetic perspective. The exploration of mitochondrial lineages, the direct maternal ancestry out of Africa, remained the dominant method of inference. That line of evidence strongly suggests that South Asian populations are deeply rooted in the subcontinent.
Other researchers were looking at the descent of males, the direct paternal lineage as recorded by mutations of the Y chromosome. The evidence from these results was more equivocal. One of the
more common South Asian Y lineages, R1a1a, is also very common in Eastern Europe and Central Asia. The discoverer of this lineage, the geneticist Spencer Wells, says that his work “in the late 1990s strongly supported a significant migration from the steppes of Eastern Europe and Central Asia into India in the past 5,000 years”. Wells connected this to Indo-European speaking nomads, and believes the latest results have borne that out. Obviously, this is in conflict with the mitochondrial results. Because R1a1a is not very diverse, it was difficult to get a sense of where or when it may have originated. Many researchers, contra Wells, contended that R1a1a may have been indigenous to South Asia.
Today, we know more about R1a1a than we did in the 2000s. Whereas then researchers looked at a few hundred markers on the Y chromosome, or perhaps some regions with very high diversity, today they can sequence most of the Y chromosome.
In line with Wells’ original suspicion, after looking at whole genomes, many scholars now surmise that R1a1a entered South Asia within the last 4,000-5,000 years from the Eurasian steppe. The reason R1a1a is not diverse is that it underwent a massive, recent expansion; not much time has elapsed for mutations to accumulate. With whole genome analysis, one can see that East European R1a1a is one lineage, while Central Asian and South Asian R1a1a strains form another. Martin Richards, co-author of the paper in BMC, explains, “This high resolution allows for both very detailed genealogical information and quite precise genetic dating—so we can see where and when lineages branch off into a new territory.”
Ancient DNA has also shed light on the relationship of the various branches of R1a1a. Extinct Central Asian steppe pastoralists, the Scythians, and their geographic kin, the Srubna people, who dwelt north of the Caspian Sea 3,750 years ago,
“THERE IS A VERY MARKED SEX BIAS IN THE ARRIVAL [IN INDIA] OF NEW PEOPLES FROM THE STEPPE ZONE DURING THE BRONZE AGE” MARTIN RICHARDS Archaeogenetics professor
also carry this Y lineage. It is notable that the R1a1a lineages of Scythians and Srubna are the same as Central Asians and South Asians, not Europeans.
A resolution to the paradox of mtDNA and Y chromosomal lineages pointing in different directions has now presented itself. The answer: migration into India was not sex balanced. This is a major point in the BMC paper. Richards states, “There’s a very marked sex bias in the arrival of new peoples from, ultimately, the steppe zone of eastern Europe, in the Bronze Age.” (This is the case in Europe too, with steppe migrants overwhelmingly male.)
Our understanding today does not rest on Y chromosome and mtDNA alone. With whole genomes available for analysis, scientists have reshaped our understanding of the past of South Asia. In 2013, geneticist Priya Moorjani and colleagues published research concluding that at least two very distinct populations were mixing in the Indian subcontinent 2,000-4,000 years ago. Moorjani says that “4,000 years ago, there were unmixed ANI and ASI groups in the Indian subcontinent”.
ANI and ASI are acronyms for Ancestral North Indians and Ancestral South Indians. The former population was genetically very similar to Near Easterners and Europeans. One of the original researchers who developed this model of Indian origins, Nick Patterson, characterises the genetic distance between ANI and European populations as so small that if you did not know of the provenance you might say it was a European population. The ASI, in contrast, do not have any close relatives. Rather, they were distant relatives of indigenous Andaman islanders. Moorjani reiterates that though there were unmixed populations representative of these groups in the relatively recent past, today all native South Asian groups display ancestry from both.
In Moorjani’s 2013 paper, she estimated that Dalits from Tamil Nadu are 40 per cent ANI (the remainder presumably being ASI). Pathans are 70 per cent ANI. Kashmiri Pandits are 65 per cent ANI, while Brahmins and Kshatriyas from Uttar Pradesh are 55-60 per cent ANI. When it comes to the mix of ANI and ASI, there are two rules of thumb one needs to consider. The further northwest you go, the more ANI you will get. Upper castes have more ANI ancestry as well (Bengalis and Munda tribes have East Asian heritage that is neither ANI nor ASI). This is exactly the pattern you would expect from Y chromosomal lineages such as R1a1a, which many geneticists posit have arrived from Central Asia in the last 4,000-5,000 years.
Indian observers of historical population genetics have noted these findings and integrated them into their own understanding. Sanyal says he believes “Indians are the mix of several genetic streams, particularly the ANI and ASI who have been living in the Indian subcontinent from the Stone Age”. Moorjani states that her “study did not specifically look into the direction of migration”. But, she also admits that “the direction of migration leading to ANI is probably into India”.
Why would Moorjani state this? First, let us take a step back and address two points of the Out of India framework. As Sanyal observed, the archaeological connections between South Asia and Central Asia are tenuous. The Aryans do not seem to recollect a time before India. But the past 10 years of discoveries of ancient DNA have shown us there were mass migrations where archaeolo-
“SOME 4,000 YEARS AGO, THERE WERE UNMIXED ‘ANI’ AND ‘ASI’ GROUPS IN THE SUBCONTINENT... THE DIRECTION OF MIGRATION LEADING TO ANI WAS PROBABLY INTO INDIA” PRIYA MOORJANI Geneticist
gists suspected none. The physical record is incomplete, and it may be difficult to connect with the broader patterns of history. But part of it is that some populations, such as nomads, likely do not leave much of an archaeological footprint. As for the argument based on Indian religious and oral history, it must be observed that the Greeks also do not have any memory before Greece, and yet they are just as antique an Indo-European people. The argument about cultural memory alone cannot be trusted to adjudicate on this matter.
Which brings us to what could ultimately resolve uncertainties: ancient DNA. The testing of samples from the Near East and Europe over the period between 5,000 and 10,000 years ago has shown that a few pulses of migration mixed together to create the predominant genetic patterns we see around us today. We do have a reasonable sampling of ancient individuals from the Near East, Central Asia and Europe, and what we see are massive population changes over the past 10,000 years. Should we expect India to be any different?
For the purposes of an understanding of the South Asian genetic landscape, two ancient populations from Western Eurasia share strong affinities with people from the subcontinent. First, the earliest farmers of Western Iran, in the Zagros, whose heritage is now found all across Eurasia, evince high affinities with many Indian populations. Second, Copper Age pastoralists of the Yamna culture of the Pontic steppe, who flourished 4,000 to 5,500 years ago, also exhibit a strong affinity to South Asians, in particular populations from the northwest and upper caste groups such as Brahmins.
Researchers in David Reich’s lab at Harvard have tested what possible groups could be combined to create the ANI element in South Asians. After exhaustive comparisons, they find ANI is best modelled as a combination of the Pontic pastoralists and early Neolithic Iranian farmers!
In The Ocean of Churn, the thesis is presented that ideas and people move in a bidirectional fashion. Indian religious and philosophical ideas did impact the West through Pythagoras and Plato. Conversely, many Indian alphabets quite likely have their origins in the Near East, while Christianity and Islam have both taken root in the subcontinent.
So too with genes. South Asian genetic markers are found in Southeast Asia, from Thailand to Bali. Conversely, Bengalis, Assamese and Munda peoples show their Southeast
Asian heritage on their faces, their genes, and in the case of the Munda, their languages. But this idea of ubiquitous gene flow has limitations.
The distinctive genetic heritage of India, the ASI component, deeply rooted in the subcontinent and not closely related to populations elsewhere, exists in low proportions in Iran and Afghanistan. But with the exception of the Roma people, ASI ancestry is notably absent throughout Western Eurasia aside from India’s near neighbours. This suggests there has been very little westward movement out of India over the past few thousand years because, as Moorjani observed, all Indian populations have ASI ancestry within the past 4,000 years.
Many geneticists now believe a major migration of people from Central Eurasia and West Asia into South Asia during the Neolithic and Copper age is the simplest and most parsimonious model to explain the data we have. The Out of India model is not theoretically impossible, but it strikes many as far-fetched and a stretch to explain the pattern of the accumulated data.
To move beyond probabilities, we need to make recourse to what has rescued us in the past: ancient DNA. Unfortunately, there is currently no ancient DNA data from South Asia proper. Even now, researchers are trying to get genetic results from samples at Rakhigarhi in Haryana dating to the Harappan period. Supposition can quickly be replaced with certitude when we sample these individuals. india today learns that results of the Rakhigarhi samples will be announced in early September this year. Dr Vasant Shinde, an archaeologist at Deccan College, Pune, which conducted this project in collaboration with geneticists from Seoul National University, is understandably reluctant to offer any pointers as to what the Rakhigarhi samples suggest. “It’s very politically sensitive,” he says. But given the fact that the graves from which DNA was extracted were dated to somewhere between “2300 and 2500 BC”—the same period in which Martin Richards and his colleagues suggest a pulse of migration from the Pontic-Caspian region into India—it must remain a possibility that Rakhigarhi will yield R1a1a DNA and not settle the debate.
And yet this is just one site. There are hundreds of samples for Europe and the Near East, and from those hundreds we have gleaned startling results. At some point, there will be hundreds of samples from South Asia, and there is no doubt we will glean some fascinating results.
The tide in historical population genetics has turned towards migration, but some still hold to the model of continuity dominant in the 2000s. Gyaneshwer Chaubey has been publishing and researching human genetics for over 10 years, with a substantial contribution in the area of Indian population history. He is not persuaded by the hypothesis of a mass Aryan migration. Rather, he observes that published research has shown that “the genetic imprint of this migration (if we want to maintain any) is minimal”.
Without ancient DNA, we can only perceive broad coarse outlines from the variation of living human beings; fragments of the past, rearranged and reconstructed using statistical frameworks and data from people alive today. Our conjectures have assumptions. Chaubey does not deny the data showing Neolithic Iranian farmers and Copper Age Pontic pastoralists having genetic similarities to modern Indians. He argues for an equal probability that groups that “lived in India in the Bronze Age or in Neolithic time having quite similar ancestry as the Steppe belt populations or Neolithic Iranians…” By reframing assumptions, he also disagrees with the revision in regard to the history of R1a1a. He admits that R1a1a was the primary reason he took himself off the BMC paper. Though unconvincing to many, Chaubey’s rationale does have a basis in theory and data. The disagreement is on the matter of probabilities. This is not uncommon in statistical genetics. Ancient DNA should resolve many of these disputes rather soon, but this issue is overly politicised.
We do know some things. Geneticists have confirmed divergences of caste and region in the genomes of Indians. A great deal of mixing seems to have occurred over the past 4,000 years. Before that period, much of the subcontinent was inhabited by people genetically very different from those alive today. By coincidence—or perhaps not—the extremely common R1a1a paternal lineage, which binds many Indian men to Europeans and Central Asians, begins to expand rapidly just as the last major mixing event in South Asia between ANI
“THE GENETIC IMPRINT OF THIS MIGRATION (INTO INDIA) IS MINIMAL... THERE WERE GROUPS LIVING IN INDIA IN THE BRONZE AGE WITH SIMILAR ANCESTRY” GYANESHWER CHAUBEY Geneticist
and ASI lineages occurred.
Prominent population genetics laboratories that have reshaped our understanding of the history of Europe and the Near East through ancient DNA studies are now looking to India. It won’t be long before new tools are brought to bear on old contentious questions. One can no longer say that the thesis that Indo-Aryans arrived in the subcontinent in large numbers is refuted. Connections to ancient groups outside of the subcontinent seem highly likely, and many prominent geneticists are now promoting a viewpoint predicated on migration and population turnover, which seems to have been the norm in Europe and the Near East.
Yet there remain credible scientists at prominent institutions who are sceptical of the new models. While a new consensus may be emerging, it has not crystallised. Genetics can speak in broad strokes with errors of the order of a thousand years here or there, and each new discovery refines details of the broader picture. We are in a time of significant transition and great intellectual ferment. That ferment extends to politics. Just as the original Aryan invasion model was promoted on a political basis, so are Out of India theories given political validity. But the reality is that these models of human history are either true—or they are not. Whether or not they were discernible, the facts have always been with us. It is our political interpretation of them that seems to change. Humans are protean, but nature is timeless.
We see through a glass darkly. In a few years, it may be crystal clear that a new people arrived in the Indian subcontinent 4,000 years ago. That now seems to be the belief among the majority of prominent researchers. A century of theorising and ideologising has armed us with answers and objections, but history as unveiled by genetics may hold some bracing surprises for our rigid grandiose pretensions. That may be the most exciting aspect of these lines of research, not how they align with century-old arguments.
Razib Khan is a geneticist with an interest in population histories and personal genomics. He works at Insitome and is a Ph.D candidate at UC Davis. He writes the blog ‘Gene Expression’