Over the past week, there have been lots of reactions to the two papers which came out last week, The formation of human populations in South and Central Asia and An Ancient Harappan Genome Lacks Ancestry from Steppe Pastoralists or Iranian Farmers. The Insight is still on hiatus, but I managed to interview Vagheesh Narasimhan for my other podcast, so check that out. Like many people, Narasimhan is not keen on the “Aryan invasion theory.” Myself, I don’t have a problem with the term, but it turns out that many Indians dislike the connotations of “AIT” quite a bit.
Since I’m not very invested in semantics, I’m going to just move on and propose another term that identifies a real dynamic. I present then the new AIT, the “The Aryan Integration Theory.”
For various reasons, Narasimhan et al. propose that steppe pastoralists who flourished between 2000 and 1500 BCE are the most likely candidates for the “steppe” contribution to modern Indian genomes. In the Swat valley samples, which date initially to ~1000 BCE, the authors noticed over time the proportion of Iranian-farmer-related ancestry decreased over time to give way to steppe and Andamese-related ancestry.
This pattern over time is related to something you see in the geographical and communal distribution of ancestry in the “three-way admixture” you see:
|Pandit||4||Jammu and Kashmir||0.159||0.616||0.225|
|Dogra||5||Jammu and Kashmir||0.178||0.601||0.222|
|Muslim_Kashmiri||9||Jammu and Kashmir||0.197||0.599||0.204|
For those of you vague on the geography, “Bihar” is the state on the eastern edge, near Bangladesh, south of Nepal. Gujarat is in the west, near to southern Pakistan.
What I want to observe is that there are groups in Bihar, such as the Bhumihar, who are higher in steppe ancestry, and, AHG ancestry, than many populations to their west. I believe this is related to the simultaneous increase of AHG and steppe in Swat.
In the revised interpretation of the above papers the Kalash of Chitral are reasonable proxies for “Ancestral North Indians.” They are a mix of Indus Valley Civilization or related peoples (~70% of their ancestry), and steppe peoples (~30% of their ancestry). The ~30% is a rough floor on their “Indo-Aryan” ancestry, because by the time the Indo-Aryans arrived in South Asia they may have been less than 100% “steppe”, accreting Iranian-like ancestry which has affinities to the IVC peoples.
An initial stylized model of the ethnogenesis of South Asian populations along the “ANI-ASI cline” (ASI being “Ancestral South Indians”), as these two populations mixed in various fractions. But it seems quite likely, and the authors of the Science paper admit as such, that period of the intrusion of the Indo-Aryans after 2000 BCE was marked by several distinctive populations interacting, mixing, and synthesizing.
It is a possibility (though not definitive) that while the Indo-Aryans were penetrating from the northwest, Austro-Asiatic farmers were pushing from the northeast. In northeast India, these people may have encountered “pure” AHG populations. Why pure? Because the cultural toolkit of the IVC civilization seemed to be optimized for the northwestern 25% of the subcontinent. In my reading, I have seen it suggested that though Gujarat and Maharashtra have toponyms of Dravidian linguistic origin, this is not the case in the Gangetic plain.
The simplest reason for the patterns of AHG, IVC-descended, and steppe, ancestry across the northern half of India, and the peculiar west to east pattern, is that relatively unmixed steppe tribes pushed eastward and mixed with local groups who lacked IVC-related ancestry. My intuition tells me (and some prior theory-reading) that a diffuse expansion along the frontier of Aryavarta would not exhibit this pattern. Rather, the Indo-Aryan tribes were highly mobile, and likely expanded into a patchy ecological landscape where they moved as socio-political units en masse.
South along the fringe of the Arabian Sea the Indo-Aryan expansion would have met denser agglomerations of IVC-descended populations. These regions were after all part of the broader IVC civilization. This explains part of the enrichment for IVC ancestry. In the Gangetic plain at a certain point, the Indo-Aryans clearly pushed beyond the limits of the IVC frontier and began mixing with non-IVC tribal people.
In the northwest of the subcontinent, the Indo-Aryans assimilated and were assimilated into, the local post-IVC populations. Over time the fraction of steppe ancestry declined in the Indo-Aryan speech community because that speech community eventually encompassed the whole population. But in the eastern frontier, the Indo-Aryans mixed with local groups. Their steppe fraction likely declined fast and stabilized quickly because it was probably a male migration, with few women.
But cultural assimilation was not uni-directional. Almost all Dravaidian-speaking South Indian groups have some steppe ancestry, and even some adivasi groups have high fractions of R1a1a associated with Indo-Europeans. This means that Indo-Aryan groups were assimilated very early into non-Indo-European speaking groups. Indo-Aryans that moved eastward along the Gangetic plain did not encounter a particularly sophisticated group of peoples (perhaps with the exception of Mundas). Cultural assimilation was toward the Aryan identity. In contrast, in the west and south, there were large numbers of non-Indo-European speaking groups with more sophisticated cultures. There were clearly cases where Indo-Aryan assimilated into the non-Aryan society.
The arrival of Indo-Aryans to South Asia seems to have coincided with a phase of admixture and integration across the subcontinent. The presence of Indo-Aryan Sinhalese in the far south is suggestive of the possibility that the non-Indo-Aryan cultures which came to light during the historical period did not have roots much deeper in the south than the Indo-Aryans in the north. An “Indo-Aryan” international probably developed in South Asia due to common speech religious rituals. But genetically there was a great deal of variance due to differential mixing with diverse local populations. The increase of AHG and steppe in Swat is probably due to the Indo-Aryanization of the region after 1000 BCE (remember than Burusho is found nearby, and it is an isolate). That process occurred partly through migration, and these cosmopolitan migrants naturally had more steppe and AHG.
Traditionally the Aryavarta has been restricted to a broad zone in northern India, the very conceptualization of territories ruled and dominated by people of common and comprehensible speech implies the existence of its converse. Though South India and Mesopotamia both were outside of the Aryavarta, the region south of the Vindhya mountains clearly exist in active and dynamic tension with the Aryan territories.
The Aryan invasion theory conjures up death, destruction, and physical domination. Some forms of the theory posit that barbarian invasions destroyed the Indus Valley Civilization. The fall of civilizations, especially Bronze Age ones, are overdetermined. It seems likely that the Indo-Aryans were able to intrude precisely because of the IVC was in decline, or decrepit. The Aryan integration theory is different because it emphasizes the creative energy and synthetic consequence of the arrival of the steppe pastoralists. Though the Indus Valley Civilization was massive compared to its Near Eastern analogs in geographical expanse, it was still sharply delimited compared to modern India. For whatever reason, it was the arrival of the Aryans which set the preconditions for the integration of diverse polities into a coherent civilization.