Caste is not ancestrally arbitrary

First, thanks to Zack for the opportunity to blog here. More importantly, thanks to Zack for the Harappa Ancestry Project! I've learned a lot from him in terms of the optimal way to go about "genome blogging," and have been able to benefit from his experiences in my own African Ancestry Project. It's really great that in 2011 we don't have to wait for academic researchers to explore the topics which interest us at the intersection of genetics and history.

Prior to being interested in South Asian genetics on such a fine-grained level I had read works such as Nicholas B. Dirks' Castes of Mind. To give you a sense of Dirks' argument, here's the summary from Library Journal:

Is India's caste system the remnant of ancient India's social practices or the result of the historical relationship between India and British colonial rule? Dirks (history and anthropology, Columbia Univ.) elects to support the latter view. Adhering to the school of Orientalist thought promulgated by Edward Said and Bernard Cohn, Dirks argues that British colonial control of India for 200 years pivoted on its manipulation of the caste system. He hypothesizes that caste was used to organize India's diverse social groups for the benefit of British control. His thesis embraces substantial and powerfully argued evidence. It suffers, however, from its restricted focus to mainly southern India and its near polemic and obsessive assertions. Authors with differing views on India's ethnology suffer near-peremptory dismissal....

One of the inferences which people draw from this model, perhaps unfairly, is that the endogamy and biological separation of caste groups is relatively new, and that genetic variation is likely to be arbitrarily distributed across caste groups. The most extreme interpretations almost seem to turn the British into the culture-creators of all that is Indian. In any case, genetics can obviously test the power of this thesis in relation to ancestry.

First up, below I have taken all the HAP samples where N >= 2. I've done some semantic shifting, so that "Tamil Iyer" becomes "Tamil Brahmin." I know that some of you have more information about the samples than is listed in Zack's spreadsheet, but I've been conservative. I will also use the word "community" sometimes instead of "caste" in future posts, because I don't know what the proper word for Syrian Christians or Bihari Muslims would be. But really same difference to me. I want to focus on groups with caste/religious labels intersected with a specific region here. The bar plot below is not going to be a surprise, and you see the clusters in Zack's dendograms, but I thought it would still be useful.

Caste is not genetically arbitrary. To me this strongly falsifies any contention that the endogamous units which we know as castes (or jatis) derive predominantly from the past 200 to 300 years. Tamil Brahmins number in the millions, so it does not seem that they plausible that could have expanded so rapidly from a very small homogeneous founder group two to three centuries ago. Rather, their origins are almost certainly more ancient. Some of the results are also not that surprising. Northwest Indians have the genetic profile you'd expect in comparison to other groups. The Bengali Brahmins consistently have more of an "East Asian" trace than other Brahmin groups, while Tamil Brahmins seem elevated in the "SW Asian" fraction in relation to other Brahmins. Both of these trends I think illustrate the likelihood of some admixture with location populations.

Now let's look within regions a bit. I'll divide South Asia into four quadrants. The classification will be self evident from the bar plots.

I'm the third to last Bengali, while the last two are are my parents. My parents are not related, and from opposite ends of Comilla east of the Padma. My mother is the last bar plot, and from a family with attested Middle Eastern ancestry (non-South Asian focused ADMIXTURE runs tend to bring the small, but non-trivial, element out more clearly). I believe that that is what is elevating her "SW Asian" fraction. It is notable that the two other individuals from eastern India who show this balance between "SW Asian" and "European" are also of Muslim background. I doubt that that is coincidental. Though South Asian Muslims are overwhelmingly indigenous, they do seem to have some outside admixture since the arrival of Arabs, Persians, and Turks, to the subcontinent. The most obvious marker of this to me isn't the elevation of "SW Asian," but the common presence of African ancestry among Pakistani Muslims. This certainly is due to the arrival of Africans and people of part African origin in the retinues of Indo-Islamic rulers.

Aside from this it seems more clear to me now that like in South India the Brahmins of the east are also relatively new and intrusive. All show an elevation of "European," though the trace of "East Asian" suggests admixture. That probably indicates their arrival after the absorption of the Mundari populations, and perhaps Tibeto-Burmans, into the substrate of eastern India.

I find nothing important to say here aside from the fact that we need a lot more samples for UP! The UP Kayastha indicates that there's a fair amount of variation here which is not being sampled.

We are obviously rich in samples from South India. One interesting aspect is the bias toward "SW Asian" as opposed to "European" among non-Brahmins, especially what I think are termed "Forward Castes" (e.g., Reddy). The proportions are low, but consistent. This is the inverse of what we see among non-Brahmins in East India. I am liable to dismiss the the East Asian admixture among many South Indians, especially non-Brahmins, as noise, but it may be signatures of absorbed Mundari substrate. Who knows? The Kerala Christian samples have the most "SW Asian." We need better references from other non-Brahmin non-tribal/Dalit castes in Kerala (a Nair is coming up), but I wonder if this validates the idea of some Semitic admixture of yore among Nasranis (or, perhaps just as likely long term trade and marriage connections over the centuries).

Now let's just look at South Indian Brahmins.

Very similar, huh?

Finally, the last cluster in western India:

Not much to go on, though I've been told that several of the Gujaratis are Patels.

Overall I think we can reject a strong recent post-colonial social construction of caste as a plausible model going by genetics. What replaces it? There probably won't be a neat model. But hopefully as HAP expands it can fill in some of the gaps. The 1000 Genomes Project will be releasing Assamese Ahom, Bengali Kayasthas, Marathas from Maharashtra, and Punjabis from Lahore, this year.


  1. { Brown Pundits } » Against social constructionism - pingback on June 7, 2011 at 9:12 pm
  2. This rejects a maximal claim, especially for Brahmins, but for that group that was never a good assumption to begin with.

    Trickier are claims in between. Susan Bayly for instance has argued (Caste, Society, and Politics in India) that large sections of the peasantry, particularly in recently cultivated areas, acquired a jati identity only relatively recently. This is trickier to tell. Brahmin/non-Brahmin identity is very different in the South; but that's in large part a legacy of the fact that Southern Brahmins owe some northern ancestry. The situation in Punjab for instance is more mixed; maybe caste identity has been more fluid there?

    It does seem that caste/non-caste populations have very different population histories, and various groups have maintained endogamy for a long time. One way to clear this up would be to digitalize the records at Haridwar and track surnames.

  3. I think you make a good observation regarding the elevated SW Asian among forward caste, socially dominant South Indians like the Reddys. In the future we might have some Kamma participants, so we'll see if this trend continues to persist among them. Reddys and Naidus (unfortunately, we have no Naidu samples as of now) have been described as chieftains who prevailed among the Vijaynagar Kshatriyas who's modern descendants are likely represented by the Nayak and Raju castes of Andhra Pradesh and Karanataka. Likewise, the Kallars and Vellalars of Tamil Nadu used to serve as chieftains under the Cholas. Strangely enough though, HRP31, who's a Sri Lankan Vellalar is only 2% SW Asian and 1% European, much too little to be considered anything significant, and was only 2% more Baloch/Caucasus than the average Sakilli/Chakilliyan Tamil in the standard K=12 runs. Of course, n=1 is hardly a sample to make any concrete conclusions.

    Regarding the SIBs, while they seem to be rather similar in terms of general admixture proportions, their inferred ASI scores seem to vary a bit. Tamil Brahmins seem to be around 38-40% ASI whereas the non-Tamil, Pancha Dravida Brahmins range anywhere from the 35% to 40%.

  4. BTW, from what I've read the Halai Bhatia (HRP108 is one) are also Sindhi Rajputs, like the Thathai Bhatias. Hence I think Sindhis; n=3, not 2.

  5. Razib, thank you very much for the post. We, Assyrians, as you know, are also, much to our surprise, a 19th century invention of the West. I honestly believe this might have contributed to the apparent lack of interest in our genome, by the academic community. Jewish populations, Druze, Samaritans, Arabs, Iranians, Kurds, Turks, Maronites, and others, have all been subjects, some multiple times, of academic studies. Without Harappa, Eurogenes, Dodedad, and the rest, we would still be an unknown, or, our identity significantly misunderstood (ie believing we are simply Christianized Kurds or Arabs).

    The Kerala Christian "SW Asian" values are interesting. I recently received my father's Dr. McDonald analysis.
    The following are possible population sets and their fractions, most likely at the top:

    Bedouin = 0.098 Armenian= 0.860 Finland= 0.000 S_India= 0.042
    Bedouin = 0.101 Armenian= 0.845 Sardinia= 0.009 S_India= 0.045
    Bedouin = 0.101 Armenian= 0.851 Finland= 0.000 N_India= 0.047
    Bedouin = 0.102 Armenian= 0.848 Basque= 0.001 N_India= 0.048
    Bedouin = 0.108 Armenian= 0.826 Sardinia= 0.014 N_India= 0.052
    Bedouin = 0.113 Armenian= 0.790 Sardinia= 0.025 Sindhi= 0.071
    Bedouin = 0.099 Armenian= 0.845 Lithuani= 0.000 Sindhi= 0.055
    Bedouin = 0.107 Armenian= 0.824 Basque= 0.009 Sindhi= 0.060
    Druze= 0.351 Armenian= 0.592 Chuvash= 0.000 Sindhi= 0.057
    Druze= 0.293 Armenian= 0.673 Tuscan= 0.000 S_India= 0.034

    To add more intrigue, I recently came across the name Beth Hindawaya in a list of surnames of residents of a particular Assyrian (now predominantly Chaldean Catholic) village in northern Iraq. Beth Hindawaya means "House of the Indians" in our Aramaic dialect. I have no known link to the town, but it certainly does suggest Nasranis were living among us in northern Mesopotamia.

    • paul, i'm sure you know that the syrian christian community was originally in communion with the church of the east, right? the jacobite connection only was formed after the rupture due to forced catholicization during the portuguese period.

    • also, just to note, ancient sumerian documents do list tax receipts from villages where the "merchants from meluhha" settled. so a connection between south asia and mesopotamia may be ancient....

      • Thanks, Razib. That is indeed interesting. On the topic of the Sumerians, the one Mandaean with Y-DNA results reported at FTDNA has a distinct hg H haplotype. Although I believe Mandaeans are a principally central to northern Mesopotamian people, the fact they lived in southern Iraq for possibly the ~1800 prior to the most recent Iraq war, creates some intrigue. But, then again, where did the Sumerians originate? And, were they a homogeneous group, or a heterogeneous mix of different types at the end? Did any of their remnants remain at the end of the Babylonian civilization?

        There is also some haplogroup H among Iranians, of course.

        And, I expect the Y-DNA results of two new Mandaeans within the next few weeks.

    • Paul, here is a link to Al-Hindya my tribe traces an oral history to this place. a few of them moved as recent as 1932 (to afghanistan/nwfp/kashmir/punjab), after iraq got independence from Britain.
      these people also went with the honorific title of 'hussaini-brahmins'
      if you do come across more research on this i would be interested.

  6. Paul, as an Assyrian, what are your thoughts about the various Christian and Jewish groups whose mother tongue is Arabic, that live as minorities in various Arab countries? Do you regard them as Arabs? If not, what are their ethnicities according to you?

    • that live as minorities in various Arab countries

      Actually many of them live as diaspora populations in the West today.

      There are also significant numbers of them, especially Jewish ones, that live in Israel.

    • Hi Onur. This question is a virtual minefield among many of the Aramaic-speaking communities of today. I would absolutely never tell anyone how he or she should or should not identify. How one wishes to identify, as you know, is a complicated matter. If you know any of my various fora monikers, feel free to pm me, and I will be happy to continue the conversation as it relates to my own people.

      • I will be happy to continue the conversation as it relates to my own people.

        Thanks, Paul. But I think today there is no ethnic identity problem in Assyrians, as they still speak Neo-Aramaic instead of Arabic as mother language. Among non-Muslims and non-Druze of the Greater Middle East, I think the only ethnic identity problem is with the Christians whose mother language is Arabic - especially those of the Levant. Jews whose mother language is Arabic have largely overcome that problem by migrating to Israel and switching to Modern Hebrew as mother language. Copts largely see themselves as a direct continuation of the Ancient Egyptians, no matter what their current mother tongue is. But Levantine Christians have ethnic identity problems (probably partly because that the Levant has been an ethnically heterogeneous region throughout most of its history; moreover, there were Christian Arabs in the Levant, especially in its southern part, before the Muslim Arab conquests).

  7. Flavors of Afro-Asiatic | Gene Expression | Discover Magazine - pingback on June 10, 2011 at 6:48 pm
  8. Flavors of Afro-Asiatic | Biology News by Biologged - pingback on June 10, 2011 at 7:31 pm
  9. African Ancestry Project: Flavors of Afro-Asiatic - pingback on June 11, 2011 at 5:33 am
  10. Nirjhar mukhopadhyay

    Dear Rajib i am from ghatal west bengal and ofcourse a brahmin. But i must tell you that from rg vedic times castes were not made by colour, race or noses shape! But by your abilities! even by having ability low born guys like vyaasa, vashistha became the supreme sages of all time and they also had their gotra. Unfortunately in later times the flexibility of the caste system got corrupted.

      • What I mean to say is that that's an apologist's view for the idealized concept of varna, and there are hints of that view in Sanskritic literature-- but you're wrong if you think that this idealized system ever existed, and that it subsequently got corrupted-- that's hogwash from the colonial age, fanned by the Arya Samaj and Brahmo Samaj.

        Varna was rarely ever a social reality, although its ideology may have been very influential to some subsections of Brahmins.

        • John, I am curious to know the basic findings based on which you are saying that varna system was never practised in India. Could you please explain? Thanks in advance.

          • Please also inform me the period of the data based on which you are conclusive about non-actualization of the varna system.

  11. Razib, in your North India ADMIXTURE bar graph compilation, you have miscounted the Jatt participants. At the time, there were only three Jatts who had been run for K=11. The fourth Jatt you have included in the image is HRP006's results using his FTDNA Famiy Finder data.

  12. Hey Razib,

    You're misunderstanding Nicholas Dirks' thesis-- he's not saying that individual caste groups were "invented" by colonial enterprise, but rather, he's saying that caste as an identity was pushed forward and made more important by the British during the colonial period-- if you read Cynthia Talbot's "Precolonial Andhra in Practice," you'll see that during the Kakatiya Age, caste was not the most important aspect of social identity.

    Also, what he is at issue with is the idea of an all-Indian caste SYSTEM-- that is, a way all of the castes relate to each other hierarchically, like a jigsaw puzzle. If you read Quigley, you'll see that what we call "jati" is not a subset of varna at all, so Sanskritic texts talking about varna have very little to do with the actual reality (you can also read Patrick Olivelle's papers on dharmasastra for this.)

    Also, Axel Michaels points out that there is no Indian word for "caste--" that the social groupings we call castes today were put into that category because of colonial enterprise-- you may be thinking that "jati" corresponds to "caste," but it doesn't-- jati pretty much just means any group.

  13. Interestingly enough, a study by Bamshead et. al. found that the "middle" castes in Andhra were more related to Europeans than the Brahmins-- I'm not a geneticist, but this doesn't seem surprising given the constant migration into India; you can't just assume that the Aryan Migration was the only one into into.

  14. Believe me i didn't say that with any emotion or trick but by vast experience and searching.

  15. Interesting and making me more curious despite the few grammatical errors and typos.

    The article starts with the subject of Nicholas Dirks book "castes of mind", which I have been reading.

    As per my understanding, Nicholas Dirk says that the caste structure that exists in India as we know now has a lot of influence from British. The summary of Library Journal is correct about Dirk's view that the British politically 'manipulated' the caste system to have a good control on India. Dirk is not denying about anything genetic there. Actually, Dirk is not at all talking anything even remotely about genetics in this book ! All his discussions are limited to ethnographic findings.

    However, the article is indirectly trying to misinterpret that Dirk is saying that the caste system has no biological relation and British has 'invented' the biological structure onto castes, and then trying to disprove that theory.

    Obviously, Dirk is not saying it, and the article seems to me as another proof of straw man logic. (Sorry !)

    I also notice that the author is using the terms that are based on the Aryan Invasion Theory. Those terms are 'brahmins' and 'non-brahmins', which is the notorious political aliases of Aryans (Brahmins) vs Dravidians (Non-brahmins).

    In addition, author's interpretations find that the so-called brahmins' genotype indicates they are relatively new and intrusive, especially in Southern India and Eastern India. To quote him:

    //Aside from this it seems more clear to me now that like in South India the Brahmins of the east are also relatively new and intrusive.//

    It is not clear from where he derives such a conclusion, especially when other community groups also share almost akin genotype pool, which the author is not finding 'new and intrusive' !

    The difference that his chart is providing between Viswakarma/Reddy and the so-called brahmin groups makes me wonder about the simulated data. However, I do not want to take that path.

    He makes it clear that he is a muslim whose mother is from Bengal but of Middle Eastern origin. Then he goes ahead and 'finds' that the origin of Pakistani muslim has a lot gene pool from Africa !

    Are not we all ?

    I also wonder how unassailabl these genotype findings are and how conclusive they could be without discussing anything about phenotype. Darwin, I miss you there !


  16. Pattom Radhakrishnan

    St. Thomas visited Kerala and introduced Christianity .After St. Thomas other missionaries from Persia came to Kerala and converted many people.. The early converts in Kerala belonging to different grades in society on the basis of occupation were commonly called Christians But the name Christian was changed to Nasrani when the Arabs (Moplahs) came to Kerala. Muslims used the word Nasrani in a contemptuous and derogatory manner. Christians and Jews were hated by Muslims in the Middle East and so they used the word 'Nasrani' in a derogatory and spiteful manner, as the Greeks called others 'barbarians.' Muslims quote the Koran to call Christians and Jews as 'Nasranis.'
    This is the basis of the verse. " Jews say: 'Uzair is the son of God' and the Christians say: 'The Messiah, son of God'. Such is their saying with their mouths; they imitate the infidels saying earlier. Allah has cursed them, how are they to turn? "(Surat al-Tawbah: 30) By this interpretation, Muslims regard Christians and Jews infidels and collectively call them ‘Nasrani.’
    Even today all Christians are contemptuously called Nasranis in the Middle East by the Arabs. In Kerala also, after the Arabs gave the derogatory name 'Nasranis' to Christians, Brahmins and Nairs also used that word in a derogatory and insulting manner till the arrival of colonial powers. Vrahnins and Nairs made Christians work as coolies, agricultural labourers in their farmlands and compulsory service called ‘oozhiyam.’ The true economic and social condition of the Christians who were ridiculously called Nasranis was recorded by a missionary when European powers were slowly establishing their authority; Abbe J.A. Dubois, a missionary in Mysore, in his letter dated August 7, 1815, writes: “The Jesuits, on their first arrival in India, hearing of them, in one way or other converted the greatest part to the Catholic faith. Their liturgy is to this day in the Syrian language, and in the performance of their religious ceremonies they use this ancient dead tongue. There remains still among them large congregations, consisting of 70 or 80,000 Christians, of whom two-thirds are Catholics, and a third Nestorians. They are all designated under the contemptuous name of Nazarany, and held by the pagans in still greater contempt than the Christians of this part of the country. The Nairs chiefly keep them at a greatest distance, and they form a separate caste.”
    When the Portuguese and the British educated these hapless Christians and appointed them as soldiers, businessmen and planters, Christians ascended to top position in the fields of education, business, plantations, and commerce making Brahmins and Nairs inferiors. The Portuguese were in Kerala for 150 years and during this period there was widespread inter racial marriage between Portuguese and local Hindus. Albuquerque officially encouraged such marriages to increase Christian population. Later the offspring of such marriages married local Christians which eventually led to change in skin color and personality. Gradually the contemptuous name Nasrani was given up and the words, ‘Christians’ and ‘Syrian Christians’ became common usage. But in Arab countries Christians are even today called in a vulgar form, ‘Nasranis.” In a recent riot in Cairo when Muslim fanatics attacked Christians, the Huffington Post dated August 17, 2013 reported the following incident
    “Naguib’s home supplies store on a main commercial street in the provincial capital, also called Minya, was torched this week and the flames consumed everything inside.
    "A neighbor called me and said the store was on fire. When I arrived, three extremists with knifes approached me menacingly when they realized I was the owner," recounted Naguib. His father and brother pleaded with the men to spare him. Luckily, he said, someone shouted that a Christian boy was filming the proceedings using his cell phone, so the crowd rushed toward the boy shouting "Nusrani, Nusrani," the Quranic word for Christians which has become a derogatory way of referring to them in today's Egypt.”

  17. Sampan Chakraborty

    @Razib Khan:recently I was discussing about it in Indian/Chinese IQ puzzle.

    Among famous Bengali brahmin such as Rammohan,Tagore.physician subhas Mukherjee,statistician Mahalanabis most of them looked like caucasian.But you will find many Brahmins with dark brown skin who look like south Indian Brahmins.

    So do you think there are both pure Aryan brahmins and mixed Mongolo-dravidian Brahmin living in Bengal?

    Also I did not find a single research abut Bengali Kayastha?May be you have not done it cause none of those legendary Boses won Nobel prize but you should know they were far more brilliant than any tam brahmin Nobel laureates.

    Actually in Bengal there was no specific job for Kayastha which is why many Brahmin used to do the job of Kayastha and became Kulin Kayastha.that probably explains why Bengali Kayastha are so damn smart.Also Satyajit Ray and Vivekananda came from Kayastha community.

    But in recent times Bengali Brahmins are doing better than any other community with DR.Anandamohan Chakraborty,the pioneer in microbology and genetic engineering,Anirban Banerjee(who won Nobel laureatesignature award in Chemistry),and several inventors...

    BTW there were few Brahmin and Kayastha were converted to Islam which is why Bangladesh has also produced couple of good names.

    I have read Fazlur khan's ancestors were Bengali elite kayastha.Same may be true with Sal khan.

Trackbacks and Pingbacks: