Bhadralok are made not born

Tanushree Dutta is a Bengali Kayastha

I have two samples of full ancestry from West Bengal. A Kayastha and a Brahmin. You can see where they plot.

Bengali Brahmins are very similar to North Indian Brahmins (often they have some “eastward” shift). In contrast, the Kayastha individual looks like the Bangladeshi samples, except with far less East Asian ancestry.

I do want more samples. Though I’ve gotten a few Bengali Brahmins and they exhibit the sample pattern as above. I am curious about non-Brahmin West Bengalis. But from the above, I think I will conclude that the hypothesis that Kayasthas are a cultivator caste which uplifted themselves occupationally is probably the right one.

36 thoughts on “Bhadralok are made not born”

  1. *Ignorance alert*
    I guess there is no need to add Brahmins from UP to the pot since Bengali Brahmins are of the same stock as UP Brahmins that were seeded under Niyoga practice. Does this hold true?

  2. So does it look most probable that our ancestors came from West Bengal Kshatriya class and moved further north east to cultivate more land and mix with tibeto burmese tribal groups that were already there?

    My ydna is R2, common in West Bengal and my mtdna is M49, present in Austro asiatic tribal groups in NE India like the Khasi who like the Garo are heavily tibeto burmese shifted.

  3. The overlap between Bengali and Telugu samples is interesting. A cline is to be expected given the geography. Many South Indian middle castes also present with epicanthic folds. E.g. old time actresses Savitri Ganesan, Krishna Kumari.

  4. Excluding the East Asian ancestry, are bengalis closer to Bihari (as my harappaworld results suggest) or Telugu people?

      1. Does that mean almost all of them converted from a single caste group. Or those areas didn’t have caste ?

        1. when i looked at the runs of homozygosity my inference is that the bangladesh samples by and large exhibit ‘random mating’ compared to other 1000 genomes. this includes samples of pakistani punjab, where there’s lots of inbreeding.

          basically, the bangladesh samples look like european ones in their inbreeding/assortative mating patterns. not too much structure subnationally.

          1. Does it mean after conversion to Islam there is enough intra caste mating that this might have destroyed any structure which was there before and there is no way to find out originally which caste they are from ?

          2. Saurav,

            Does it mean after conversion to Islam there is enough intra caste mating

            I doubt stratification could have been removed in 500 years.

            Very likely Sri Lanka is also going to have very little stratification. There is/was caste but more based on trade/obligation to king.

            Feature of non Aryanized societies?

  5. I doubt stratification could have been removed in 500 years.

    yes, it can be. 20 generations is a long time. the uygurs don’t show much stratification either, and they mixed ~1500 years ago.

    also, south asian stratification is also less regional so random mating could happen quickly since it doesn’t require gene flow over distance.

    Very likely Sri Lanka is also going to have very little stratification. There is/was caste but more based on trade/obligation to king.

    the tamils in the 1000 genomes were sampled in the UK but are sri lanka tamils. they look like an indian population (there are clear dalit subsets and a few brahmins in the data set).

    Does it mean after conversion to Islam there is enough intra caste mating that this might have destroyed any structure which was there before and there is no way to find out originally which caste they are from ?

    yes. or it might have happened earlier. the admixture with east asian-like group seems to have occurred 1,500 years ago or so. that component is pretty evenly mixed too, though there is a geographical cline. i don’t think it’s a coincidence that my parents are among the most east asian, and both are from comilla (the west bengal samples always have less east asian).

    basically the bangladesh samples probably show geographical variation, just like european ones do. but they don’t show too much evidence of jati endogamy except for the 5 individuals who look like dalits.

    1. is that the bangladesh samples by and large exhibit ‘random mating’ compared to other 1000 genomes. this includes samples of pakistani punjab, where there’s lots of inbreeding.

      if it was Islam that brought about “random mating” in Bangaladesh, should not that be true of other Indian regional Muslims. Any indication of non stratification, say among Punjabi or Kashmiri Muslims.

        1. it happens. but it is rare. the genetic analysis i’ve done confirm this; cousin marriage leaves very distinct signatures.

          the punjabi samples (collected in lahore) show

          1) lots of cousin marriage
          2) also, indian-like stratification. some of the punjabis are more like south indians. some of them look like pathans. and there is the full range.

      1. the pakistan punjab samples are highly stratified. just like the indian samples. and it’s not just cousin marriage. the only hypothesis i have is that tribal identity or something must be enforcing endogamy.

        1. People tend to marry within the same “biradri” if not to their actual relatives. This is not an issue so much in cities but it definitely matters in villages.

        2. This would mean that the intra caste marriages in bengal would have happened before they converted to Islam. Do you see similarities between the N-Eastern people and east bengalis then? They have tribes but not caste.

          Also as an aside i used to think that genetic data carries identity marker of each generation, and nothing is “lost”, shouldn’t it mean that if there was stratification at one point of time we could/should have evidence on that. Probably that or i am ignorant on this issue 😛

          1. search my other blog. blogged on this. short of it is that some of the east asian in bengalis is probably from tibeto-burmans or perhaps even tai. not just austro-asiatic munda. i don’t know the details about when/if intra-caste marriage happened. i assume so.

            Also as an aside i used to think that genetic data carries identity marker of each generation, and nothing is “lost”, shouldn’t it mean that if there was stratification at one point of time we could/should have evidence on that. Probably that or i am ignorant on this issue 😛

            you’d have to use local ancestry deconvolution and that is hard when the genetic distances aren’t great.

          2. i don’t know the details about when/if intra-caste marriage happened. i assume so.
            the pakistan punjab samples are highly stratified. just like the indian samples. and it’s not just cousin marriage.

            So its a possibility that “random mating” was pre Islamic culture in eastern India.
            Maybe the Sinhalese in SL will show such “random mating”.

        3. Biraderi.

          Which is pretty much Jati. Often with the same names — such as Muslim jats, Muslim khatris, Muslim gurjar.

          Pre partition, there was some attempt at jati/biraderi unity across religions, but it didn’t come to much.

  6. the tamils in the 1000 genomes were sampled in the UK but are sri lanka tamils. they look like an indian population (there are clear dalit subsets and a few brahmins in the data set).

    Lankan Tamils are stratified to this date and very endogamous.

    Sinhalese ideal is endogamy, but historically they had loose civil unions. Well one day we will see from data.

  7. “probably from tibeto-burmans or perhaps even tai. not just austro-asiatic munda”

    But shouldn’t the facial structures than match more towards them rather than matching mainland Indians? Even with different type of mix i feel east and west bengalis look similar.

  8. 15% admixture in the east. closer to 5% in the west. so why?

    anyway, i have plenty of relatives who look east asian. i have very little body hair, which is supposedly rare in south asians according to russell peters 😉

    1. Im also from Comilla(Homna) and have around 13% east Asian. Most of my relatives are pretty hairy and show little to none east Asian traits, some of them looks quite west Asian influenced, infact even Pakistanis mistake them as Afghan or Iranian. Also some of the common folks from Homna have west Asian influenced look. It’s hard to believe that Bengali muslims have lower western Eurasian than Bengali Brahmins. I dont think there’s any significant difference in appearance of Bengali muslims from Middle-upper class and Bengali Brahmins. Honestly there are some middle-class Bengali muslim that have strong west Asian influenced appearance compared to Brahmins.
      I’ve also noticed that in the city where I live, majority of Bengalis have “educated” middle-class background, so their appearances are similar and sometimes west Asian influenced, but the bengalis that live in our neighbouring town are of poor background and they come from remote villages of Kishorgonj, their facial features are also a bit different from ours,their features often seems more indigenous tribal influenced.
      Isn’t that result of some sort of Endogamy?

        1. Ill try to convince some people to get a dna test, but the problem is nobody is showing any interest unfortunately.

  9. “15% admixture in the east. closer to 5% in the west. so why?

    anyway, i have plenty of relatives who look east asian. i have very little body hair, which is supposedly rare in south asians according to russell peters 😉”

    Im also from Comilla(Homna) and have around 13% east Asian. Most of my relatives are pretty hairy and show little to none east Asian traits, some of them looks quite west Asian influenced, infact even Pakistanis mistake them as Afghan or Iranian. Also some of the common folks from Homna have west Asian influenced look. It’s hard to believe that Bengali muslims have lower western Eurasian than Bengali Brahmins. I dont think there’s any significant difference in appearance of Bengali muslims from Middle-upper class and Bengali Brahmins. Honestly there are some middle-class Bengali muslim that have strong west Asian influenced appearance compared to Brahmins.
    I’ve also noticed that in the city where I live, majority of Bengalis have “educated” middle-class background, so their appearances are similar and sometimes west Asian influenced, but the bengalis that live in our neighbouring town are of poor background and they come from remote villages of Kishorgonj, their facial features are also a bit different from ours,their features often seems more indigenous tribal influenced.
    Isn’t that result of some sort of Endogamy?

      1. I agree compared to other south Asians Bengalis on average show some east Asian influence, a few Bengali can look even fully east Asian. But I guess Bengalis in general show good diversity otherwise as a Comillan I wouldve had atleast a few relatives with east Asian feature.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.