13 April 2010

Genetics and the Afro-Dravidian Hypothesis

The global distribution of Y-DNA haplotype T

French anthropologist Bernard Sergent made a strong linguistic case in La Genèse de l'Inde (1997) (linked excerpt translated from French into English by Sunthar Visuvalingam), that the non-Indo-European Dravidian languages of India show a very strong linguistic affinity, too specific to be attributed to chance, with the Niger-Congo languages. He also notes that the material culture of South India, which is home to the Dravidian culture, underwent a dramatic change strongly parallel to that of West Africa around 2500 BCE, when the area adopted agriculture. He notes too connections such as a shared game, shared religious beliefs, shared attitudes towards cousin marriage, similar home construction methods, and similar musical instruments. The linguistic suppositions he makes are shared by some linguists. See, e.g., Upadhyaya,P & Upadhyaya,S.P.(1979).Les liens entre Kerala et l"Afrique tels qu'ils resosortent des survivances culturelles et linguistiques, Bulletin de L'IFAN, no.1, 1979, pp.100–132.
and Upadhyaya,P & Upadhyaya,S.P.(1976). Affinites ethno-linguistiques entre Dravidiens et les Negro-Africain, Bull.de L’IFAN,No.1, 1976,pp.127–157. Aravanan,K.P. (1976). Physical and cultural similarities between Dravidians and Africans", Journal of Tamil Studies 10, 23-27. Aravanan, K. P. (1979). Dravidians and Africans. Madras. Clyde Winters is a less distinguished proponent of the theory but offers notable citations in support of it and has inferred some proto-Dravidian-African agricultural vocabulary.

The package of crops that arrived in India at that time, including pearl millet, are crops that were domesticated in West Africa around 7000 BCE, were in cultivation by Niger-Congo language speakers in West Africa around 2500 BCE and are not native to the Near East, South Asia, Southeast Asia or East Asia. The fact that the Indian crops are Sahel African domesticates and the timing of their arrival around this time has recently been confirmed with agricultural archeological work, by individuals including Dorian Fuller. The timing of the appearance of these crops coincides with the linguistically inferred date for the proto-Dravidian language.

Sahel crops don't grow well in the conditions that Fertile Cresent crops do, and visa versa. The crops grown in Egypt, Europe and the Fertile Cresent, while well suited for the Indus River Valley, would not be suitable for South India. But, Sahel crops do grow well in South India.

Contemporary scholarship is also increasingly clear that the Dravidian culture was not an outgrowth of the neighboring Harappan culture of the Indus River Valley which was contemporaneous with it. Witzel at Harvard shows an absence of a Dravidian substrate in the early Rig Vedic Sanskrit (the Indo-Aryan source language in a manner similar to the way that Latin is the source for the Romance languages) that surely had some sort of Harappan substrate given the place where it arose and the matters discussed in the Vedic literature that relate to that geographic location. The interaction of the two cultures may have been limited to one or two trading port outposts on the West Coast of India on the Harappan frontier. Harappan warehouses at those trading posts are the earliest signs of Sahel crops in South Asia. Genetic evidence from Y-DNA also supports a lack of a link between the Harappans and the Dravidians:

Associated microsatellite analyses of the high-frequency R1a1 haplogroup chromosomes indicate independent recent histories of the Indus Valley and the peninsular Indian region. Our data are also more consistent with a peninsular origin of Dravidian speakers than a source with proximity to the Indus and with significant genetic input resulting from demic diffusion associated with agriculture.

Annotations to sources for these points can be found here.

Looking For A Population Genetics Link

The most important missing link is in the population genetics.

West Africans overwhelmingly belong to paternal line inherited Y-DNA haplogroup E1b1a, with haplogroups E1b1b found in the herders to their North and basal haplogroup E1a* found in Western and Northern Africa, particular in Mali, and haplogroup E2 found in many places in Africa, particularly those with Bantu origins. The Y-DNA haplogroup commonly found outside Africa is the E1b1b group associated with herders including the Berbers, which spread to the North Mediterranean during periods of cross-Mediterranean commerce and through admixture with Arab herders who are prediminantly Y-DNA haplotype J1.

The dominant maternal line inherited mtDNA haplogroups in West Africa are those of the L haplogroups (particularly L2 and L3 in West Africa). There is a small frequency of mtDNA haplotypes M1 and U6 in Africa, which appear in locations similar to the North African Y-DNA Y haplotypes. The M1 mtDNA hapolotype is found only in North Africa, the Mediterranean and the Middle East; it is not found in South Asia. All other M type or descended mtDNA hapolotypes are found in South Asia or points East of there (except M48, a rare haplotype found only in Saudi Arabia), except in case attributable to recent back migration.

There is a community of people with African genetic origins in South Asia, but they are found in the South Indus Valley and trace their origins to a historically documented Islamic military unit from Africa about two thousand years after the South Indian Neolithic, and about a thousand years after the Indo-Aryan invasion of India which had culturally transformed that part of India before this community arose.

Yet, it is hard to imagine that a culture that had become the dominant culture of South India by the time of the Indo-Aryan invasion, over the course of a thousand years, transferring the leadership group's language, agricultural practices, religion and material culture to India, as Sergent hypothesizes, would have left no trace on India's population genetics. The genetic lines of failed cultures disappear into isolated relict populations or become vanishingly rare, buried in the leading culture. The genetic lines of a successful, expanding, transformative food producing culture, in contrast, should leave unmistakable population genetic marks.

One can imagine cultural borrowing limited to crops from a distant land. But, a cultural borrowing of a pre-literate dominant language and total culture from afar seems very implausible. Such a total cultural transfer needs to be carried by people, or at least it did in 2500 BC.

Now, these marks need not be equal in the paternal and the maternal line. There are many examples of major cultural expansions in which men belonging to the expansion group left a strong paternal Y-DNA haplogroup mark, but had children with local women, leaving the indigenous mtDNA population mix largely unchanged. If an original proto-Dravidian group of colonists from Africa were entirely composed of men, it would leave no mtDNA mark at all. Even if there were some women in the leadership group, if they were greatly outnumbered by the locals and had purely cultural rather than genetic seletive advantages, they might vanish without a trace in a few generations from the gene pool.

The Case For Y-DNA haplogroup T as a Dravidian leadership group marker

So, is there any way that the cultural evidence and the genetic evidence can be reconciled? It turns out that there is a way.

Y-DNA haplotype T is a particularly plausible candidate for this transmission.

Y-DNA haplotype T is most common in the pre-Aryan Dravidian area of India. For example, more than 22% of Telagu speaking men in India have this Y-DNA haplotype, and in some South Asian populations it exceeds 50%. This percentage is on the same order of magnitude as, but somewhat lower than, the genetic contribution of the formative Indo-Aryan population in parts of South Asia that were never Dravidian, as expected from a population that is believed to have arrived in India somewhat earlier.

Y-DNA haplotype T is also virtually absent from the areas associated with strong Indo-Aryan influences, so its presence in South Asia isn't likely to be a result of admixture of Sumerian and Harappan populations in connection with their long trade associations with each other by sea and possibly also by land. Matrilineally inherited mtDNA evidence also supports this conclusion:

West Eurasian mtDNA haplogroups H, JT and W represent 6–7% of north and central tribes, which are located in the area where Indo-European languages are spoken. In contrast, these west Eurasian mtDNA types are virtually absent in south tribes, which are located where Dravidian languages are spoken. This might reflect different responses of local people to the Indo-European settlement of India. In the north and center, Indo-Europeans may have admixed with local people, concomitant with the spread of Indo-European languages. In contrast, in the southern part of India, local populations may have challenged the arrival of Indo-European newcomers, resulting in limited admixture, reduction of tribal population sizes and retention of their original languages, thus explaining why Dravidian languages survived the spread of Indo-European languages in south India.

Tibeto-Burman language speakers in South Asian tribes have strong East Asian genetic as well as linguistic affinities in both mtDNA and Y-DNA which are found nowhere else in South Asia, suggesting that "these populations remained relatively isolated."

The populations in South India where Y-DNA haplotype T is most prevalent coincide strongly with the linguistically inferred proto-Dravidian homeland within India. See also here.

Type T is also one of the few Y-DNA haplotypes not of the E type that are found in a Fulani population of West Africa, where about one in six Fulani men in Cameroon are currently of that type, and the Fulani speak a Niger-Congo language family language quite close linguistically to the proto-Dravidian language.

The Y-DNA T haplotype is also found in the Horn of Africa, in the Nile River basin, in the Fertile Crescent and in low frequencies in places where the non-Atlantic part of the Neolithic expansion took place. In the Nile River basin proper, Y-DNA haplotype T is most common in Upper Egypt and Sudan.

Sergent makes the observation, and I offer it for what it's worth, which may not be very much, that "the resemblances among Nubians, proto-dynastic Egyptians, Dravidians and what was formerly called "Hamites" appear henceforth through multivaried cranial measurements "as being very evident," and that among the so-called "Hamites" (in reality of the kushitic languages, that is to say the linguistic family called semito-hamite), the Somalis and Galla are black-skinned, it is probable that the Dravidians have conserved their color on leaving Africa; their installation in the Indian tropical zone could have subsequently only confirmed and augmented this pigmentation."

Making The Connection

How did type T get to Africa and to the other places where it is found? And, where did it come from?

Y-DNA haplotype T is an offshoot of Y-DNA haplotype F, which is found, with its descendants, across the non-African world from Europe to Australia, but is not found in Africa except among the Fulani of Cameroon, on the Horn of Africa, and in the Nile River basin.

One plausible scenario is that mutations that distinguish Y-DNA haplotype T from its direct ancestor Y-DNA haplotype K, are part of the migration out of a "Eurasian Eden" (probably South Asia or Persian) into the Fertile Cresent that also brings Y-DNA haplotype R1 with it, probably prior to the Neolithic Revolution.

When agriculture arrives from the Fertile Cresent in Egypt, where it arrived before it did anywhere else but the Fertile Cresent and the Indus River Valley, men with Y-DNA haplotype T settle in the Nile River Valley and become part of the core population of the agricultural era of the Nile River Valley. Other men from the Fertile Cresent with Y-DNA haplotype T are part of the population that brings agriculture to Europe in the LBK population that brings agriculture to Eastern and Central Europe.

Sometime later, a group of Y-DNA haplotype T Egyptians continue towards the source of the Nile, which splits in two directions, the Blue Nile makes its way East to Ethiopia. The other branch extends to the West until it reaches the divide between the Lake Chad inland basin, and the Nile River basin. An exploring group on foot looking for the source of the Nile and forking to the West would easily end up in the Chad Basin and from there to the home of the Fulani in the Sahel, where the crops that make their way to India have their origins.

Another group of men, with Y-DNA haplotype R1b1a (also known as R1b-V88) is also found primarily in North Cameroon, where it is spoken mostly by Chadic language speakers. The relatively undiluted mix of Y-DNA haplotypes of these men, compared to those with Y-DNA haplotype T suggests that the haplotype T group arrived long before the R1b-V88 group did.

In the Sahel these Y-DNA haplotype T men learn Sahel agriculture (the only kind of agriculture in sub-Saharan Africa until the Bantu people develop tropical agriculture many years later), and adopt a Niger-Congo family language. The Sahel crops have origins on the opposite sides of the Sahel and Y-DNA haplotype T is not widespread in sub-Saharan Africa, so they are unlikely to have domesticated these crops themselves. Trade conducted by the Mesopotamian or Egyptians with the Harappans brings them the discovery that there is some place suitable for Sahel crops in South Asia which can be reached by ship. These men set out up the Nile Valley to Egypt with their crops to Mesopotamia to South India, or alternately to the Horn of Africa to South India. One of these possibilities seems most likely.

It isn't clear if the Y-DNA haplotype T in Ethiopia and Somolia and Tanzania has its origins from Egyptian sea traders migrating inland, or from people moving up the Nile and taking the Blue Nile fork rather than the White Nile fork and head East as well as West. Either way, Sahel farmers in Ethiopia could bring crops and culture, as well as their Y-DNA haplotype T from Ethiopia to South Asia by boat.

But, a Fulani source seems more likely than an Ethopian one, because otherwise it is hard to explain why the men bringing Sahel crops to South Asia would speak a Niger-Congo language, although this could have been the language of the region prior to an influx of Afro-Asiatic languages from Egypt and the Middle East. The Kordofanians (who have a relatively high frequency of Y-DNA haplotype T) may be a relict population that is the most direct linguistic and cultural descendant of the East African Niger-Congo language speaking population, something supported by the fact that they are believed to be the oldest layer of people in the Nuba Mountain area of Sudan which has been a refugium for many layers of peoples over past millenia.

Also, Sahel crops may have come relatively late to East Africa: "Wiegboldus (1996) found no evidence of millet and bicolour sorghum being cultivated in East African countries until late antiquity, millennia after African millets were being cultivated in the Sahara, West Africa" and in South Asia . . . Wigboldus,J.S. (1996). Early presence of African millets near the Indian Ocean. In J. Reade, The Indian Ocean (pp.75-86), London: The British Museum."

A third possibility is that the Fulani or a kindred people with high levels of Y-DNA T traded back and forth all the way to the West African coast and took a boat from there, but there is considerably less evidence of sea trade reaching that far around 2500 BCE.

There are lots of populations with Y-DNA haplotype T in Europe, in Egypt, and in the Fertile Cresent, but those populations would have no way to learn how to farm Sahel crops, or enough exposure to the Niger-Congo language and culture to account for the affinities unlikely to be a result of chance that Sergent attributes to an Afro-Dravidian link. The documented dates of domestication of the Sahel crops also leave no doubt that the domesticated were brought from the African Sahel to South India ca. 2500 BCE, and not the other way around.

Standing alone, these might seem implausible stories, but these extraordinary stories would provide a population genetic link that could explain the many coroborating details found in the cultural connection made by Sergent, and explain the unusual distribution of this Y-DNA haplotype.

The fact that Y-DNA haplotype T is quite concentrated in South Asia, with high frequencies in some groups, while being virtually absent many other parts of South Asia, also point to its relatively recent arrival in South Asia relative to other Y-DNA haplogroups found there.

Updated and Expanded on November 15, 2010; minor corrections made December 13, 2010


Anonymous said...

The presence of HG T in South Asia is among the tribal groups. These are the original, indigenous people of India and have been there for more than 50,000 years so a recent African migration is unlikely. Second, the origin of T is West Asia so it's presence in Africa is from a back migration.

Andrew Oh-Willeke said...

Not only tribal groups, and the fact that a group is currently "tribal" doesn't insure its antiquity. It isn't unheard of for farmers to "revert" to hunting and gathering or marginal agriculture when circumstances promote that, or for an ethnic outgroup to have that status despite not originating in an area.

The scenario of T being in West Asia or SW Asia ca. 10,000 years ago, migrating to Africa and Europe in the Epipaleolithic or Neolithic, and then moving from Africa to India ca. 2500 BCE, is not particularly far fetched.

It is very unlikely that hg T in South Asia is 50,000 years old, as its frequency would have reached fixation by then, rather than being highly distinct in distribution geographically.

Anonymous said...

I think history would be very clear if the Dravidians ,who originated from Mesopotamia , were not confused with the indigenous peope of india ..
The Dravidians travelled far and wide spreading their culture even to south America etc etc

Anonymous said...

Confusing the Dravidians ( who came from Mesopotamia ) with the original indigenous peole of india who were non Caucasian
Will give rise to many discrepancies in history

Andrew Oh-Willeke said...

I am not deleting the last two "anonymous" comments, although it appears to be a double post by the same person, perhaps accidentally.

We certainly don't know that "Dravidians originated in Mesopotamia," or that any late migrants to India who may have been formative in Dravidian ethnogenesis were or were not Caucasian or originated in a particular place. We have inconclusive clues. The Elamo-Dravidian linguistic hypothesis also doesn't have very solid support.

We certainly know, however, that nobody from the Old World spread their culture to South America between the advent of Dravidian culture ca. 2500 BCE and 1492 CE - and the arrival of people other than Europeans to the New World came many decades after that.