Brahmins were made in India, not the steppe


The above are Y chromosomes from ancient samples in the steppe, Iran/Turan, and South Asia. The time periods are obvious. EMBA = “Early Middle Bronze Age”, MLBA = “Middle Late Bronze Age” and LBA = “Late Bronze Age.” IA = “Iron Age.” H = “Historical.” And the other periods are Neolithic or Copper Age. This is from Narasimhan et al. (click the image above for the supplements).

The Forest/Steppe samples are most from what Sintashta archaeological sites. One thing that is evident in early Indo-European pastoral people is that they seem to be highly patrilocal and patrilineal. One particular genetic lineage group of males seems to dominate different early groups. The data from Narasimhan et al.  show us that:

R1a is overwhelming in the Sintashta.

R2 & L is found in pre-Indo-European Iran.

Q & N is in Sintashta too.

H1 is mostly found in South Asian populations.

You can see the distribution in modern populations on Wikipedia, but data from a paper is illustrative:

Kashmiri Pandits519.89.85.8811.763.9219.6113.73
Kashmir Gujars4910.26.1216.332.0440.868.16
UP Brahmin3116.133.233.2167.743.23
Punjab Brahmin4921.437.1435.7125
Himachal Brahmin3010.535.265.265.2647.375.26
UP Kols3011.1133.3440.7414.81
UP Gonds3859.4618.928.11
MP Brahmins427.1423.817.1438.111.9
MP Gonds3156.256.2518.75
MP Saharia5723.43.2422.828.075.37
Bihar Brahmins382.6313.165.2660.535.26
Bihar Paswan273.73.77.413.711.1140.74
WB Brahmins305.5672.2222.22
Maharashtra Brahmins326.6716.67103.33043.333.33
Gujarat Bhils229.0918.1818.189.0918.18
Gujarat Brahmins641.5615.637.819.3832.819.38

​The numbers don’t add up to 100% because I deleted some columns. Brahmins have a lot of R1a. We knew this. But please note that they have lots of other Y chromosomal lineages as well.

We need to be cautious here about not overreading these data. But please notice that Punjabi Brahmins have more IVC-derived Y chromosomes than R1a. R2, J2, and L are all likely related to eastern Iranian farmers. In contrast, UP and Bihar Brahmins have more R1a, and less the of the Iranian farmer Y chromosomal groups. Finally, observe that UP Brahmins have a substantial minority of H1, which is quite plausibly from the AASI.

Here is a hypothesis that I will present which explains the origin of Brahmins, and to some extent Indic civilization as we understand it. The high fraction of IVC Y chromosomes in Punjabi is a function that this region was the center of a more complex society when the Indo-Aryans began filtering into the region. The “Brahmins” was created from an amalgamation of the Aryan priestly class with IVC priests. Many of the things unique to Brahmin culture are from the IVC.

Second, Narasimhan et al. report that IVC ancestry declines as one goes east in the Gangetic plain, AASI goes up, but “West Eurasian” ancestry stays more constant because Indo-Aryan ancestry is high in groups like Brahmins in Bihar. I believe that the Gangetic plain was liminal and more open to Indo-Aryan expansion than the core Punjab, where Aryans assimilated a much larger group of IVC people. Rather, in the Gangetic plain Indo-Aryans assimilated more “jungle people.” Perhaps some hunter-gatherers, but also slash and burn agriculturalists of predominantly AASI origin. The fraction of H1 in UP Brahmins indicates that this was the case even in core Aryan tribes.

The emergence of Brahmanism, and later Hinduism, was a complex multivalent synthesis. There wasn’t one single mixture, one single invasion, but rather multiple peoples interacting and integrating. Though Brahmins are enriched for Indo-Aryan, Narasimhan et al. could not define them as a lineage with common descent. That means parallel cultural processes created multiple founding Brahmin lineages, rather than a single process creating a single lineage.


141 Replies to “Brahmins were made in India, not the steppe”

  1. Mahabharata is instructive here – read Arjun clearing the khandava forest – basically a crime committed by one of the ‘good guys’ – love the Mahabharata for its moral complexity.

    In either case.. my point is that this process of entering new lands with more tribal folk may have occurred well into the puranic period.

    Great blog post btw – you won’t find this content anywhere else.

  2. Wait, are R2, J2, L related to the Neolithic Iranian Farmers around 8,000 bce or the newly discovered lineage of herders/farmers, distantly divereged from those Neolithic people. I recall a recent paper; farming could have been indigibious to the IVC and not mediated by the Iran_N farmers in the Neolithic. Instead , a highly divergent group in Iran (seperated to the predecessors of Iran_N), immigrated to IVC region perhaps at the Palaeolithic age and mixed with AASI.

    1) Are R2, J2, and L lineages from this lineage that you suggest, it’s not from Iran_ N, right? I mean I know some Of these lineages do appear in M/N and EN Iran. So are these lineages linked to Iranian Farmers, or from those Palaeolithic people who diverged from those farmers however stayed around? I hope my question seems coherent.

    2) why is R2a so high in Brahmin clases, Razib? I thought we were Dasa toilet cleaners lol. No, seriously. R2a seems to have a huge presence in even punjabi Brahmins. What’s the link?

  3. Wait, are R2, J2, L related to the Neolithic Iranian Farmers around 8,000 bce or the newly discovered lineage of herders/farmers, distantly divereged from those Neolithic people

    the two groups of eastern farmers split after LGM but before holocene. i think they share common haplogroups.

    2) why is R2a so high in Brahmin clases, Razib? I thought we were Dasa toilet cleaners lol. No, seriously. R2a seems to have a huge presence in even punjabi Brahmins. What’s the link?

    toilets are sacred! but yes, i think R2 is from ivc ppl

  4. 3) Finally, there seems to be a R2 presence in Iron Age Iran but more of its successor lineage R2 in Neolithic and Bronze Age Iran. This must mean that R2a was introduced from the IVC and does not come from Iran. I posit that R2 were forager Male groups in Central Asia who made their way into NW India. Their successors, R2a among the new IVC community mixed with local populations (AASI, J2, L, females etc vice versus), and formed the Brahmin caste of the new fledgling status pole. They may have moved to Iran for whatever reason explaining their small presence there. This, until R1a makes from the Steppe integrated themselves and became part of the Brahmin hierarchy. This makes sense, right Razib?

  5. Its a good hypothesis.

    All this makes me wonder if in a way the previous class of Rig Vedic Gods ie Indra, Varun, Surya are in a way pre-Hindu gods (since there mention in Mittani), and the real native god is actually Vishnu. There is some evidence that Shiva is more IVC than Aryan, debatable.

    That would make the oldest Hindu text ironically, non-Hindu 😛

    1. That would make the oldest Hindu text ironically, non-Hindu 😛

      Indra is considered of the Aryan affiliation, possibly shared with Greeks, Romans in common lineage.
      But in India Indra was relegated to a lower Tier(King of the Gods in title but still very clearly NOT the Primary God head or on that Tier or clearly subservient to a few other Gods, namely the Trinity and even certain Female God-head entities) and most importantly Indra usually had an excessive amount of mocking stories about him, making him to look like a fool/incompetent.

      A senior primary God of the conquering people simply CAN NOT met such a mocking fate. That defines basic logic of human interaction. We know from Sumer’s multiple City-States era and diverse Gods how obsessed people become with their primary deities.

      Maybe Vishu-Shiva were local but what about Brahma, when did Trinity arise and how did Indra end up getting so side-tracked. This is a question which geneticists can’t answer, we need other domains to give a plausible enough answer to this.

      1. Vishnu could be either. His original version was a contemporary and like comrade of Indra however the later Vishnu was influenced, maybe even defined by Krishna traditions, whose origins I do not know of. I am leaning on Shiva being more local however.

        1. the later Vishnu was influenced, maybe even defined by Krishna traditions, whose origins I do not know of.

          A possibility could be assimilation of local/regional Origin-Stories.

          Vishu is supposed to have multiple Avtars, they could very well be from different communities. Even in recent memory we have people who try to include Buddha as Vishnu’s avtar.

          Hindu-Civilization (whatever it was at that time) seems to have a very high capacity to absorb other traditions. It could absorb the entirety of Abrahamic doctrine inside it without compromising itself.

          So from hypothetical, Vishnu could have been a normal God who got turned into this or maybe didn’t exist before anywhere but during the process of Assimilation of Avtar Cult/Myth became powerful enough to be part of Trinity.

          1. Brahma is the creator god, mostly a side show. His job could have done by essentially any other god. He was laid off

            Shiva is the mountain god. He is either IVC or pure Aryan god, considering that mountains are not a big deal in Hindu-ism’s mecca (UP-Bihar)

            Vishnu/Krishna/Ram is the real “son of the soil” god. Son of cowherd, lives in forests for 14 years. Sets up cities in foreign lands. Lies ,cheats ,steals. Runs away from battleground , even while preaching others not to. So a true Indian, and “native” god. 🙂

  6. Finally, there seems to be a R2 presence in Iron Age Iran but more of its successor lineage R2 in Neolithic and Bronze Age Iran. This must mean that R2a was introduced from the IVC and does not come from Iran.

    you should look it up, but i bet “R2” here is just because the older samples don’t have enough markers to make the a vs b call.

  7. Is H1 really AASI?

    Seems like maybe it;s possibly some Paleolithic Central Asian hunter-gatherer marker, with a connection to West Eurasia. There’s European H too.

    If H isn’t AASI though (it might be, it might not), what y-dna in India can be tied to AASI?

    I can’t think of any. India’s AASI ancestry seems to be maternal heritage.

    1. That would be so strange!!! Surely there couldn’t have been a complete replacement of the native AASI ydna. No way could a male group be so effete that any outgroup males simply takes their womenfolk.

      I think In this case, the population of AASI must’ve been so small that the males were naturally replaced by more ‘fit’ outgroup West Eurasian males. Consider the fact that AASI males were probably Pygmy looking males.

    2. H definitely wasn’t introduced by the Iran HG considering how old it’s south Asian branches are, has to be pre-LGM (well over 20,000 years old). However I do think that it wasn’t a part of the original east Eurasian settlement of south Asia either. For now I think that ydna H1 was brought to India by a Levant late Aurignacian population some 30,000 years ago which was entirely west Eurasian and 0% basal Eurasian. Subsequently it got mixed with other populations in its original homeland and with proto aasi in India. The basal Eurasians probably started out in Arabia before their later expansion into all of west Asia between Anatolia and Iran sometime around or just before the lgm (not entering south Asia until late ice age/ early holocene in the form of Iran hgs who were ane + basal + para-whg). I think that dzudzuana might be one of the first cases of this basal population mixing with west Eurasians. The reason why I place the basal Eurasian pre-lgm homeland in arabia is because of its vicinity to west asia and the concentration of mtdna r0a in the region.

      Anyway that was a justification for Levant aurignacian moving into south Asia without leaving a trail of H1 in its path. I also think that these Levantine cro-magnon-like people brought mtdna u2a, u2b, u2c, u8c with them. I saw an unusually large number of the rare mtdna u8c in a report regarding the mtdna of yadav people who are not very different from chamars autosomally. It was only a handful but that is still unusually high considering that was are talking about u8c here.

      1. ” For now I think that ydna H1 was brought to India by a Levant late Aurignacian population some 30,000 years ago which was entirely west Eurasian and 0% basal Eurasian. ”

        That would make folks from the eastern part of South Asia more Western Eurasian and less basal Eurasian than their western cousins. Basal Eurasian probably produced stereotypical middle eastern look, while the West Eurasian proto-European look. I can notice resemblances between these sportsmen:

        What percentage of Levant late Aurignacian autosomal DNA do modern Levantines/Arabs have? I know there is very little H left there.

        1. I think that the stereotypical middle eastern look is recent and probably tied to the regional agricultural selection. The early epigravettian Natufians were said to be very much like European hunter gatherers of the mesolithic in appearance, while the more modern Mediterraneanid appearance came later on in late Natufian groups due to selection. On the other end of west Asia, John Lawrence Angel emphasized the similarities in appearance between Hotu cave people (which turned out to be very basal rich, the most basal rich Eurasians known so far) and European hunter gatherers of the ice age. A part of the stereotypical modern MENA, okay lets be more direct, Arabid appearance is chamaecrany, while the Hotu cave female skulls were high ortho- to nearly hypsicranic in spite of long skull values for females.

          The question regarding how much of the Levant Aurignacian ancestry persists in modern Levant is something that I do not have an answer to. I also don’t know how much of it is present in AASI since it is a guess for now based on haplogroups alone.

          1. @Son Goku
            It is hard to say what is and isn’t a phenotypical relic because the mixture, if it happened, would have happened so long ago and in an unknown proportion such that we don’t know that it would be hard to make a strict connection. One example of getting bamboozled by phenotypes and ancestry would be proto-Iranid vs Eurafricanid. Proto-Iranid is a type found in Iranians today and is hypothesized to be a relic of the old stone age Iranian population. On the other hand, Eurafricanid is considered to be one of the relics of north African late stone age people. However, when you actually look at the stone age skulls of these regions, the stone age north Africans look more like proto-Iranids (skulls that are broader than their height from Iberomaurusian sites); meanwhile stone age Iranians look more like Eurafricanids (Hotu cave). Heck even in the IVC zone, the skulls that were initially classified as proto-Australoids were then later on, at least briefly, re-classified as Eurafricanid. Probably from the stone age Iran influence. If physical taxonomy when used alone and with broad brush can confuse ancient north Africans with ancient Iranians with proto-Australoids then it is an indication that you cannot rely on it alone especially not in far fetched comparisons between population data points that are separated by thousands of km in space and thousands of years in time.

            If you want to use it then use it sparingly and when comparing groups over a relatively small frame of time (a few thousands of years, complete with the sequence of changes instead of a broad sweeping stereotype) and across specific regions which have populations whose existence you know about beforehand (as opposed to picking large region with who knows how many unknown populations as variables, the more the unknowns, the more liable you are to make mistakes). And of course, you only use it as your only approach if you don’t have better methods like genetics at your disposal, or to make comparisons about environmental impact on otherwise genetically similar groups.

            As far as Alpinid is concerned, the emergence of most Alpinids tend to be associated with neolithic farming, so it would be kind of weird if it came from Levant Aurignacian peoples. The Alpinid group in India is of a much more recent origin, pretty sure it has nothing to do with changes from 30,000 years ago. I don’t think we have enough Levant Aurignacian skulls for a good comparison anyway.

    3. “If H isn’t AASI though (it might be, it might not), what y-dna in India can be tied to AASI?”

      I support DaThang’s opinion ATM thus H might be pre-LGM west Asian. If that’s the case, Y haplogroups C/D would be valid candidates for proto-AASI.

      1. Add K2 to the list. There is divergent K2a/K2e in south India and a very divergent K2b2 subclade in the Andamans (distant proxy). C1b + D + K2. I wonder if the Fs in Paniya etc are really F (XGHIJK) or just unidentified subclades within GHIJK, kind of like how early 2000s study gave plenty of P to central and north Asians with very little Q but the more recent studies given much higher Q along with pretty much no P (xR, xQ)-> basically showing that earlier unidentified subclades of Q were being counted as P outside of Q because Q wasn’t very well known..

  8. @Grewal, Razib
    According to YFull, South Asian H and West Asian/Euro H are separated by a TMRCA of 45,600 ybp.

    Isn’t that around the time West and East Eurasians diverged? It’s possible that TMRCA indicates this early split among Eurasians, with H1 representing an East Eurasian lineage (AASI likely).

    But then there’s the absence of any early H1 branch in East Asia, SE Asia and further beyond the Wallace line. Maybe they had H but it went extinct.

    1. “It’s possible that TMRCA indicates this early split among Eurasians, with H1 representing an East Eurasian lineage (AASI likely).”

      The question is if AASI is a homogeneous group or a heterogeneous group? We need to dig out AASI autosomal DNA to find out the answer.

  9. I don’t know why but brahmins are not considered good in my community and considered opressors. Some people in my extended family have this perception that could’ve happened due to media perception or could’ve some historical truth to that.
    But the people that i have met are very nice,mostly educated, very germaphobic and overall good people.

    But given enough food and nutrition the pygmy looking males can turn into beasts. But some people(most of them) like Jatts,gujjars are naturally built i guess but can’t say the same about the brahmins though. The population replacement is really difficult for people whose tribe has small numbers(SINTHASTHA) mixing into larger tribes.(AASI)

    ONE question is : Did the geography, the fertile land of the Gangetic plains helped the Aryans to settle in India?
    Did the brahmins have monolpoly over knowledge and philosophy and other people were just labourers, agriculturalists , sewage workers etc?

  10. Excellent post!

    I remember my first post (here or maybe GNXP) where I asked whether it is possible that only the first veda (rigveda) was composed entirely by Indo aryans and rest of the literature (begining from Yajurveda) were created by IVC priest after appropriating the ideology + some DNA exchange.

    Well this theory puts that same thought in much better /plausible way. It will however by very difficult, if proven true, for lot of people who have black & white approach to the whole AIT. Where Aryans riding horses impose caste system, vedas, Hinduism into the egalitarian wonderland that exists before through the agency of Brahmins

    I think this theory should satisfy the RW too. Majority of what is quintessential indian culture is defined not by Aryans coming from Europe but the IVC. It is their ideas that can tell apart Hinduism from other pagan religions. Indo aryans arrival however does play a big part in uniting the whole region with creation of hinduism + Brahminism.
    Staying true to the lore, steppe people always end up uniting through disruption.

  11. inb4 anthrogenicaoid GED match retrofitting in R to suit narrative and cherry picking data gang comment and rage on this all over the internet. weird racialist types who take pride in their complexes lmfao

    S Asia is complicated. The Steppe:AASI ratio is one of the ultimate guides but definitely not the end all be all.

    Is there evidence of caste in IVC. Now that would be interesting. My impression is that pastoral societies tended to be more egalitarian, due to the harshness of the steppe. Hence how turkic people are defined more by culture than be race. Is this a misconception?

    1. Is there evidence of caste in IVC. Now that would be interesting. My impression is that pastoral societies tended to be more egalitarian, due to the harshness of the steppe. Hence how turkic people are defined more by culture than be race. Is this a misconception?

      pastoralists have more ‘churn.’ in any case, India is unique in its stratification. china and Europe don’t have this.

      turks are not defined by race (though they were in the 9th century, there is a long Islamic treatise on why they are superior to Arabs and persians), but Turks all have some east Asian ancestry. so that’s the ‘original’ signal.

      there is some evidence of caste in areas in the south.

      1. //but Turks all have some east Asian ancestry. so that’s the ‘original’ signal.//

        Razib, Turks like Krygz and Kazhaks are predominantly East Asian(60%>) while Uzbeks and Uygurs are around 40% and 49% respectively. So, it not “some” east asian ancestry but more.

  12. I think this backs what I have largely thought, which is that the Indo-Aryan supremacy was not a result of an invasion, but rather a result of religion.

    Judging by the timings, admixture occurred towards the end of the Cemetery H period, which likely means this was not an actual IVC synthesis.

    It is apparent that areas outside of the main areas of settlement were the centres of cultic worship. When we look at the Agnicayana and Rudram, for example, we see a ritual and liturgical structure that firmly rejects the main centres of physical settlement. When combined with the broader of content of the Vedas, Aranyakas, and Brahmanas, we see the same example. The Vedas are learned by young Brahmins in areas far from the villages and towns, and this continues into the early classical era.

    My idea is that the post-IVC society had some type of religious complex centred on open spaces and the wilderness. The Indo-Aryans, with their steppe background and religious practices, clearly fit very well into this structure, which is why the priestly class is the most Steppe enriched, not the royal or warrior level.

    It also explains why their language, which was that of prayer, was so easily imposed. That said, though, we don’t really know that the linguistic map of India was, and the strange language borders we see today may have been the result of relatively modern migrations. I suspect, for example, that large areas of Madhya Pradesh, Chhatisgarh, Maharasthra, Jharkand and Orissa were non-Indo European speaking until the post-classical era.

    1. “I suspect, for example, that large areas of Madhya Pradesh, Chhatisgarh, Maharasthra, Jharkand and Orissa were non-Indo European speaking until the post-classical era.”

      Or as some would say the “less Hindu” regions of that time. 🙂

  13. The oldest parts of the Rgveda (Mandalas 2,3,4,6,7) mention Aryas as belonging to the Puru tribe, specifically the Bharatas. Other tribes are never described as “Arya” in a “us vs them” criterion. Some individuals from those tribes are however specifically described as Arya. The functional expression is entirely in a communal sense and perhaps even cultural sense, never a racial sense. Sri Aurobindo and Ambedkar came to the same conclusion independently in the 1940s. More recently, Talageri also states the same based upon his “old and new books” analysis.

    I do not think either Arya or Brahmana represented any sort of racialized anthropological unit 3500 yBP ago, irrespective of current day stratification.

    This hypothesis falls into the demarcation gap of Popper, the zone where neither confirmation or refutation is possible. Unless an ancient grave is found with the headstone “Here lies a Brahmana”, this one will remain safely outside the realm of falsification to sceptics.

    1. I don’t think that establishing a connection is as hopeless as it may seem. If we can get some translations from IVC giving us an idea about their society and religion, we might be able to make connections to the later mixed Vedic traditions.

      1. If IVC script is deciphered it will be like a season finale.

        Also it would be max fun if everyone is proved wrong at same time.

        1. On the other hand I hope it’s never deciphered.

          Or else how will we play all this cute games?

  14. This hypothesis falls into the demarcation gap of Popper, the zone where neither confirmation or refutation is possible

    i think popper is fine. but if you work in science you know no one cares about philosophy of science, and if you work in philosophy of science you know that popper is ‘old hat’ and not held in such high esteem anymore. so incantations to popper have zero impact.

  15. – people should not think AASI are pygmies. no evidence for this. the pygmy negritos of modern day s and se Asia are not representative.

    – we know very little about the genetics of AASI aside from a 50,000 foot level

    1. Sorry about that. I assumed AASI might be Pygmy looking because their closest (although highly diverged group) of East Eurasian HG tend to be short. That, and most if not all populations with a significant AASI admixture (East and South Desis) tend to be much shorter than their highly enriched West Eurasian ANE cousins in NW South Asia.

      I apologise bhai, I will never undermine you again. Anecdotal perceptions is baseless against genetic evidence.

      1. Linking height with the ancient origin is not congruous. Height in a community can increase within a few generations on condition that there’s enough sustenance.

  16. Razib

    I have a few questions on the history of South Asia.

    What do you suppose the Paleolithic and Mesolithic population looked like in terms of autosomal and ydna? Did rice farming really date to 7000 BC in the Ganges Valley and why were the Mesolithic Ganges Valley people so tall (183 cm for men and 181 cm for women)?

    Is the native population of Assam really Austro-Asiatic? I thought this was a recent migration to South Asia (no older than 2000 BCE)? Also is there a reason South Asian like people (either Iran_N or ASI or AASI) did not populate the high mountainous regions south of the Himalayas and left that open to Tibeto-Burman colonization late in history (4000 BC?)? This applies more to the Eastern Himalayas than the Western Himalayas.

    Do you know if the Tarim R1a is really Z93-?

    What do you think of the significance of R1b-L51+ in Volosovo and R1a-Z93+ in CT and Usatovo? This is the opposite of what everybody predicted.

    1. “Is the native population of Assam really Austro-Asiatic?”
      Are there any Austroasiatic tribes(except Garo tribe) in Assam?

  17. I hope that comment number 60,128 goes through, don’t feel like writing it all over again only to have the new version fall through the abyss as well.

  18. Very illuminating ideas in this post and the earlier ‘Sintashta’ one. Some great comments as well.

    I think its clear that despite the general egalitarianism among pastoralists, the Indo Aryans (and maybe their ancestors, the Sintashtas) already had a three fold division of society comprising of priests, warriors and laymen before they came to India.

    Clans may have jumped from one division to another but it was not something easily achieved. eg. Vishvamitra’s clan became Brahmins from Kshatriyas. Similarly we have many examples of Brahmins ruling kingdoms. A lot of laymen clans may have gained power (thus becoming Kshatriyas) during times of political fluidity.

    Also there was at times significant hostility between these divisions of society as mentions of a war between Bhrigus (a clan of Brahmins) and Haihayas (a clan of Kshatriyas) shows.

    One thing I am not sure about is claim that the formation of the Brahmin caste had some contribution from an already existing Indus-Valley priesthood in the Punjab region. As in, Brahmins are a result of a merging between Aryan priests and the Indus Valley priests. There seems little evidence of this. More probably, the Aryans priests married any suitable native and over a few generations this led to the Brahmin caste.

    Ofcourse, the Aryan religion absorbed many tenets of the Indus-Valley faith to finally become Hinduism as is recognizable today. But I feel that may have come about organically because Aryans found themselves in an overwhelming Indus Valley milieu atleast in the Punjab region.

    1. That makes absolute no sense if you consider the high rate of R2a Brahmins, almost parallel to R1a in the various populations. If R1a were exclusively Brahmins by marrying in, then why the large percentage of R2a Brahmins almost on par (give or take 5-10% in some cases) with R1a. Obviously, they had huge significance. The aryans most likely incorporated themselves into an already organised IVC hierarchy.

      The question is, why were they accepted. They didn’t come with womenfolk so by default any integration will be patriarchal. Was their integration forceful? They were proud warrior types. Or after a series of migrations, accepted due to their own traditions/culture? Perhaps, they too had a sophisticated religion. Or perhaps, their appearance. We all know how much desis love light skinned brown people. If, I imagine, these swarthy toned beautiful entered the subcontinent, they may have been admired for the appearance alone. Dark (ish) skinned IVC people would have lapped up to marry their daughters off to these chads. Perhaps, perhaps.

      1. // I imagine, these swarthy toned beautiful entered the subcontinent, they may have been admired for the appearance alone. Dark (ish) skinned IVC people would have lapped up to marry their daughters off to these chads. //

        This is the quality content i come to Brownpundits for, lol.

    2. Bhrigus i.e. ‘Brigi’ (something as – highlanders) were a Serbian tribe which originated in today’s Montenegro and migrated to Asia Minor. Later they got the better-known name – Phrygi(ans). Well known is a Mita Gordyev (Gordian) dynasty (similar to Mino in Crete) which was a competitor with Karanovic’s dynasty and occasionally they had local wars. Mita is also known by his knot (Gordian knot) which Aleksandar (the Great) Karanovic cut with his sword.

    3. The problem with that narrative is you give examples of rigid groupings, that are difficult to jump, from much later time. If you read the discussions here and sintashta shamans most of the material (mythology) you mentioned has formed after Indo Aryans were deep in India. That tells nothing about their behaviour before.
      Also it is interesting to notice that earlier there is 3 fold system in parallel to other aryans elsewhere. Later it is not restricted to varna division alone but results in much larger number of castes (in thousands). There is not one group called Brahmins since most likely they didn’t even come in a single wave. Even in one culture/language there exists different castes within same varna which didn’t intermarry. What can explain that?

    4. “I think its clear that despite the general egalitarianism among pastoralists, the Indo Aryans (and maybe their ancestors, the Sintashtas) already had a three fold division of society comprising of priests, warriors and laymen before they came to India.”


      Correct and good thinking. Serbs had so-called ‘three-functional’ division for thousands of years and they brought this to South Asia. I wrote about this before. SA caste system is probably a consequence of the system brought by Aryans. I have no knowledge that caste system existed before, but I do not exclude this. That was a merit-ory system. For example, in so-called ‘ancient Greece’ (oxymoron expression), Sparta had meritory system while Athens had a tribal system which we now call – democracy. Democratic system was highly corrupted and could exist only based on enormous number of bureaucrats and slaves which number exceeded the number of free people (only in Athens more than 100 000).

      It is worth mentioning that Serbs never had slaves in their many thousands of years of history. Even prisoners of wars were free to move in general population and very often they got assimilated. Enemies knew this and because of this, they tried not to be killed in a battle, knowing that as future prisoners, they will be free.

      That is also an important consequence for SA history, they had caste merit-ory system but not slavery.

  19. “Many of the things unique to Brahmin culture are from the IVC.”

    Could you please elaborate on that or point me to more info?

    Great post.

        1. very few societies practice obligate vegetarianism. BUT, vegetarianism does seem to crop up in ascetic subgroups in complex societies with variegated ideological systems (essenes and pythagoneans in the west). steppe pastoralists are never vegetarians (obv). i think IVC societies could had groups which engaged in this practice for various reasons

        2. IVC is most likely not a monolith but cosmopolitan in nature. So it doesn’t have to be all in or all out.

          If you chart the percentage of vegetarian people across different states of India you can see the peak in people along IVC. This is where the IVC haplogroups also peak, if I am not wrong.

          The article is an example of the people I was refering to before, who see everything in binary (black or white, no grey areas)

        1. yeah. i think the diff btwn brahmins and other north indian indo-aryans is endogamy was stronger and so they mixed less with local people.

      1. ” vegetarianism and excessive focus on ritual pollution”

        They are related to my pristine IVC-Dravidian culture, and not to Aryans!

        Oh my Periyar!

        1. lmao, AIDIAMKA or whatever they are called are probably linguistic and genetic descendants (much more so mid-castes like Kallars than Tamil tribals/chamars), as are North, particularly NW+West, Indians genetically, but pure IVC culture emanated from Balochistan/Mehrgarh. Thats where youll find evidence of the proto culture when you peel away later NW Iranian (Balochi language) and Indo-Aryan contributions

          South-Dravidian, one branch of the lost Dravidian proto-family from which all modern South Indian Dravidian languages are derived from is likely only one offshoot of the proto-IVC culture given the linguistic time depth is so low (paper can be provided if interested) and South Indians further heavily mixed with local peninsular AASI culture; IVC migrants brought NW AASI with them South but further admixed culturally and genetically with local ashmound South AASI cultures to top up their total AASI %.

          So IVC was overall much lower in AASI than modern South India IMO and modern South India obviously has seen a lot of local cultural AASI assimilation as well as cultural innovation/evolution derived from local geography and technology flows such as local flora/fauna and retrofitting possibly far away NW cultural motifs into local scenarios

  20. If R1a were exclusively Brahmins by marrying in, then why the large percentage of R2a Brahmins almost on par (give or take 5-10% in some cases) with R1a. Obviously, they had huge significance. The aryans most likely incorporated themselves into an already organised IVC hierarchy.

    i don’t think r1a is exclusively ‘brahmin’ cuz brahmins are more recent than indo-aryan intrusions. r1a is found all over the subcontinent. even in Dravidian speaking tribal people.

    also, you shouldn’t presume that the ivc people were all so dark. ppl in sindh are mostly ivc and can be on the lighter side, right?

    1. You are right, Razib. I just searched Sindhi people on google images and they are really light.

      Thank you for not admonishing me harshly. We sit at your feet, O Guru!

      1. Maybe Sindhis are considered light by Indian standards bit in Sindhis are considered the darkest Pakistanis.

        The order from lightest to darkest is usually


    2. Sindhis are often remarkably light-skinned, in some ways looking less characteristic of the indic average than haryanvis, punjabis, and kashmiri pandits. The indus people may have yet had a discernible color stratification, and the encounter between the aryans and those light-skinned Indus people wouldn’t have been one of perceived racial strangeness to the extent we imagine. A putative proto-caste stratification in indus peoples might have been the distant legacy of the Iran-related post-holocene people and the AASI, which may have been a practically black-white racial encounter of two vastly divergent branches of humanity. The Aryans, already socially expressing a “tri-functional” division as you noted, may have subsumed the structure and given substance to it ideologically over time.

      1. @girmit
        Agree with most of the outline

        Quote /Aryans, already socially expressing a “tri-functional” division as you noted, may have subsumed the structure and given substance to it ideologically over time./
        I am not sure they had the required power in the new society based on the fact that their chief god indra had to be pushed down considerably and with mocking stories to add to injury.

        Whether their religious ideas + language was used as useful tools to tell someone else’s story or whether they were sly enough to change their own story entirely to fit in the new society?

        Interestingly the other 2 groups in original tri- functional society are nowhere as distinct today! Or am I missing something?

        1. If the Indus civilization was in late collapse, its conceivable for Aryans to have commandeered one of the more important fragments, perhaps in present Haryana, say through demographic force concentration, and made it the nucleus of a new culture that harnessed whatever remaining vigor there was in the productive forces of the old society. Also taking into consideration that an InPe folk exodus occurred to distant lands to the south and east (with or without Aryan infusion) somewhere in this timeline without carrying much if any literary culture or meme complexes with them. The latter makes one imagine an expended root civilization. Not convinced yet if these Indus wanderers were dravidian in whole or part.

          1. I get your point of IndoAryans getting the expended civilizational capital easily in post IVC world and wanting to fashion it in light of their thoughts. Maybe like Tesla getting traditional manufacturing site + equipment for peanuts as compared to the price it would have commanded in past.
            But then why would they compromise with primacy of their head god for the gods of old society? Well I am assuming that someone like Shiva is more of IVC god from information available.

          2. IamVY,
            quote //But then why would they compromise with primacy of their head god for the gods of old society? //
            Put into the perspective of long enough timelines, from late harappan —> proto-vedic aryan ingression—->late bronze age rg vedic—> mahajanapadas, more than a 1000 years of cultural and demographic churn would have passed ( late-bronze age collapse as an inflection point?). The aryans themselves, intermixture notwithstanding, would have become unrecognisable plains-dwellers who’s history would have become mythology. The primacy of a pastoral sky-god may have attenuated due to a novel material cultural ,ecology, and social economy. The expanse of india would get settled more thickly and agricultural, herding, and extraction economies established with networks of trade and all the attendant technologies. Imagine all the varieties of AASI (i think of the variation we could see between american tribes from iroquois to seminole) and InPe diaspora being assimilated into this expanding civilisational complex, and like mithraism gaining influence in rome, the various cults that fused into the shiva, vishnu, and durga were canonised in aryavarta.

          3. youareVY – Maybe the comparison is not coincidental. Tesla was a tall Aryan (R1A, 187cm) –😊. Shiva (i.e. Ziva) also existed in the oldest Serbian mythology. She had couple names – Ziva (=alive) or Vesna. Vesna is a modern and frequent (Jatt_S would say – “cool”) Serbian name. In Russian is still Spring=Vesna.

      2. @Girmit

        Solid 100 to your post. Caste system likely predated Indo-Aryans in some form given its literally only a feature in South Asia compared to all other IE societies

        As are things such as Agamic traits such as asceticism, yoga vegetarianism etc not seen anywhere else in the IE world so likely either purely local or local culture influenced

    3. They can be. But what I also find interesting is that the more AASI shifted end of the IVC range was a 50% or so AASI.

      From the diagram in Rakhigarhi paper post

      Modern day Gujarati Patels, Punjabi Dalits, and S Indian mid castes like Reddys have little steppe and are about half IVC and half AASI on many popular calculators. Interestingly, the geographic cline makes sense here. The IVC people I would imagine ranged for these looks to a Baloch type look, essentially the looks of the non majorly steppe impacted people of the modern day area that makes up the IVC region.

      I also think that perhaps the threshold for declaring an outsider “alien” looking was less than it might be today. I would imagine societies were more homogenous, so even relatively small deviations in phenotype would look quite different, let alone the difference between a typical Patel look like Paresh Rawal and say the look of Sintasha (take perhaps modern day Georgians as the proxy for argument’s sake).

      Even the North African Moor in Othello, likely someone of beberid ancestry with minimal SSA, was noted to look quite different from the Italian locals of the play (of course British actors played them all), with particular references to his thicker lips.

      1. “I would imagine societies were more homogenous”

        Agree with your point by and large but people might not be as homogenous as one might imagine if the IVC society itself was stratified into endogamous groups, as some have hypothesized here.

    4. Fantastic write up

      @Razib to your point funnily enough the most genetic connection to IVC peoples IMO outside Balochi/Brahui people are Gujarati/Sindhi Lohanas and Gujarati Muslims in general (at least the Sindh shifted ones) from the data ive seen. They also have a very low Steppe to Iran N ratio which is almost the reverse of UP Brahmans (high Steppe to Iran N ratio)

      The closer to Balochistan you are (approaching from the east from the Gangetic plains) the less Indo-Aryan contribution (steppe adna and cultural) so its easier to sift through and make deductions. Can you do a piece on these samples?

      Ive noticed Sindhi people also dont have Brahmins, and dont really have a formal caste/jati system (staple middle Indo-Aryan high culture features) that to me shows that Sindh wasnt a core Indo-Aryan region and most of the Indo-Aryan genetic/cultural component there was introduced and existed largely in deference to the local elements aside from the language (Indo-Aryan)

      1. That is so egregiously false that it needs to be called out. Please read

        Go to the supplementary tables as a shortcut and look at Lohanas. They literally are among the MOST steppe shifted groups out there. They score even more steppe than some North Indian Brahmin groups. Their steppe is similar to Khatris and just under the Kalash.

        Gujarati Muslims also have a fair bit of steppe and a decent recent West Asian input, as seen by caucus admix. Neither of these groups comes even close to a good proxy for any part of the cline of the Indus Valley.

        Again, the Indus Valley has two components

        “the-origins-of-the-neolithic-in-the-indian-subcontinent” on this website

        Please look at this post ^

        It is made of a cline of Iranic Mesolithic HG related (not exactly but shares common ancestor with) ancestry (common ancestor with Iranic Mesolithic HG- the predecessor of Iranic Neolithic Farmer [what was the old model for IVC but this turned out false] and Iranic Neolithic HG)- and AASI. We can make it short hand Iranic and AASI cline. The Iranic ranges from 50-100% and the AASI from 0-50%. There is NO steppe or Caucus type late input, as seen in Gujarati Muslims like Bohras, Memons, and Ismalis. Lohanas again are quite Steppe shifted. Even look at the GEDmatch oracles and search Lohanas if you trust the anthrogenica unverified GED kit data more.

        We DO NOT have a good proxy for that “pure” iranic related end of the IVC cline. We do have a good proxy for the half AASI, half iranic end. Those proxies are groups that are about 40-50% AASI, with little steppe input. Those groups include, once agian, S Indian mid castes like Reddys, Gujarati Shudras like Patels, and Punjabi low castes like Ramaghiris or even dalits like Chamars.

        A Mahzabi Sikh like Daler Menhdi or a Gujarati Shudra like Narendra Modi are frankly the best living relic type proxies of the AASI end of the IVC spectrum.The other end really doesn’t exist because of mixing with other W Eurasian groups like the steppe peoples.

        And there is caste in Sindh dude. Look at how low castes are treated in Karachi. And many of the low castes are Pak Christian converts. Many of the Hindus of Sindh remaining btw are those with little means to have left and are low caste. Sindhis also refer to their merchant as Vaniyas like Gujaratis.

        And the Sindhi Brahmins likely converted to the Syeds and Shahs. Sindh was just under Muslim control longer. So clarity of caste hierarchy in original form was replaced. Syeds btw generally claim to be upper caste converts as a rule.

  21. What’s your opinion about the Brachycephalic/Alpinid South Asian phenotype, is that a reduced version of proper Levantine cro-magnon that came with H-men?

    idk dude. the only phenotypes i keep a close eye on are hot women 😉 i’m a geneticist

    1. Asking a pseudoscience-related question in a genetic thread wasn’t a good idea, I guess.

  22. “Though Brahmins are enriched for Indo-Aryan, Narasimhan et al. could not define them as a lineage with common descent. That means parallel cultural processes created multiple founding Brahmin lineages, rather than a single process creating a single lineage.”

    Does this apply to South Indian Brahmins, as well?

    Or is it that multiple Brahmin lineages formed in the north (in the most general sense) first and then some of them migrated to the south (where they then further mixed with locals)?

    If latter, are they more similar to certain northern Brahmin lineages over others?

  23. Or is it that multiple Brahmin lineages formed in the north (in the most general sense) first and then some of them migrated to the south (where they then further mixed with locals)?

    i think something like this. i don’t have enough sample for non-UP northern brahmins, but basically UP & bihar brahmins are the same.

    bengal brahmins are 75% UP and 25% bengali
    tamil brahmins are 75% UP and 25% tamil
    and so forth

  24. I am British-Bangladeshi from Sylhet.

    Y hg L-M357

    mtDNA M35b

    Phenotypically we vary a great deal with mish mash across north Indian subcontinent, East/Southeast Asian, central Asian and middle eastern.

    My paternal line were most likely converts as my great-grandfatherX4 had the name Mohan prefix.

    Some examples my mother’s siblings have blue and green eyes. My nan had greyish hazel eyes and these people all have fair skin, narrow nose, thin lips, etc.. I am waiting do some tests on my mum’s paternal line. It will massively help assume a possible migration pattern to Bengal. I also believe they’ve managed endogamy very well due to their physical appearance across a few generations. My father’s side appears to be more diverse albeit still some very NW Indian and Central Asian.

    I am quite tall 6ft 1 and have a soft long hooked nose. Most Pakistani Punjabis I meet think I am l one of them. Most Bengalis I meet assume I am Pakistani or Indian as do Europeans. There assumption is Bengalis are short, dark, and Indo-Chinese in appearance.

    Going by my research on L-M357 it appears that it’s pretty high with Jatts and many Punjabi Sikhs.

    1. “I am quite tall 6ft 1 and have a soft long hooked nose. Most Pakistani Punjabis I meet think I am l one of them. Most Bengalis I meet assume I am Pakistani or Indian as do Europeans. There assumption is Bengalis are short, dark, and Indo-Chinese in appearance.”
      I’m Bengali too and have many hook-nosed, fair-skinned, 6ft+ tall cousins/relatives, and I can confirm you they look unmistakably Bengali. 
      From when Europeans can tell apart South Asians?

      1. Well my friend go and speak with these people who have a stereotypical view of the Bengali phenotype. I hear it time and time again from people including South Asians.

        I do know what Bengalis look like and they can vary vastly. At the same time I am not ignorant to the fact we are product of people mixing not so long ago let alone hundreds of years ago.

        My point I made is in relation to the data presented and Y hg L appears to be mostly non-Bengali groups. Identifying the varying phenotypes in my extended family may also help with establishing admixture. You may be interested in this paper

        My purpose here is to work out where the different parts of my ancestry came from through dialogue and research.

        1. 1) that’s not a good paper to cite. Y strs can’t tell you that

          2) bengalis do vary in phenotype. but a huge number ARE short and vaguely east Asian looking. not all. similarly, most Punjabis are lighter. but there are dark Punjabis. e.g. the actress parminder nagra.

          3) the random sampling indicates that the vast majority of bengalis are not descendants of recent migrants. the vast majority who are NW Indian looking are not descendents of recent migrants. those who look like tamils (dark+curlier hair?) are not tamils either.

          1. Thanks for your input. It’s quite a struggle to find much work on Y hg L and this paper was the only thing that I could find.

            I am ~12-15% E/SE Asian based on GEDmatch calculators. I believe that’s quite normal for Sylhetis.

            It would be great to see future work on Y hg L including my subclade L1c.

  25. Also do we know howrice got to the Ganges Valley and ultimately the Indus Valley?
    Tea is also native to the eastern end of the Indo-Gagentic Plains (Assam). How did that get there and why wasn’t it more widely cultivated?
    Also is there evidence of WSHG ancestry in South Asia? What about Central Asia and the Iranian Neolithic? Is that ultimately associated with y R2?

  26. Also what’s the order of migration of ydnas into South Asia?

    I would assume the first lineages were a combination of F, C and H. R1a last and everything else in between during the Neolithic. Also are South Asia F,C and H considered West Eurasian?


    I’ve met very few Bengalis who actually look Indo-Chinese. Most are shorter than people from the West Side of the Subcontinent but I always doubted it is due to genetics.

    1. >During the neolithic
      We don’t have Indian mesolithic neolithic DNA yet but I don’t think that there was a major migration to India during the neolithic itself. The mtDNA dates for south Asian U7 put it at around 12,000 years before present.

      1. *more specifically talking about northern India as far as neolithic movements are concerned, IDK about the northeast.

      2. *more specifically talking about northern India as far as neolithic movements are concerned, IDK about the northeast.

      3. Then where did ydnas E,G, J1, J2a, J2b,L1a,L1c, R2a and T in South Asia come from? Maybe they arrived to the outskirts during the Mesolithic?

        1. I don’t have ancient DNA to make solid claims, but indirect inference would say at least so. The mesolithic of India dates back to 12,000 BP at the earliest. On the other hand, the age of the south Asian mtDNA U7 is at a conservative ~12,000 BP and apparently the L1c/L-M357 in Afghanistan has an age of ~15,000 BP. All of this points to a very long presence of these populations at the edge of India since long prior to the Indian neolithic. There is also the Q1b/2 split that goes back to around the same era. So while I don’t have the late UP to mesolithic south Asian DNA, the indirect stuff points to a presence since prior to the neolithic. This, combined with the Rakhigarhi paper is as close I can get to saying that the Iranian component was not brought by neolithic farmers. The only way to make it more definitive would be DNA from India and Pakistan from 15,000 BP to 10,000 BP. My current view is that they were in India at least since the mesolithic if not the late UP.

    2. Yeah height is an interesting topic. Emperical data shows the tallest groups in India currently are Punjabis and Keralites (170cm). But the tallest religious groups are Jains (largely vegetarian and genetically diverse in proportion of major ancestral components) and Sikhs (170cm average- also genetically diverse but with rather divergent clusters [eg. chamars vs. jatts]).

      The trend seems to be related more to who is has food and who doesn’t. Kerela and Punjab are relatively rich states. Jains and Sikhs are relatively rich people. Homelessness, a decent proxy for starvation, tends to be uncommon among these groups.
      I think nutrition will allow many groups to attain better heights and for the some of these stereotypes to die. There may be differences, but I think they are way exaggerated by environmental factors.

      One thing to note that bone structure does some seem to vary a bit. S Asians have smaller bone structures on average no doubt relative to Europeans. A good proxy for this and also used in the bodybuilding and strength training world is wrist and ankle size. There are pretty good predictive tools on things like max bench, given a certain wrist size. Of course, there are exceptions.

      The following is all anecdotal:

      In India, I noticed some variation with this. Almost all groups were small boned. But some were closer to being medium boned, though none seemed to have the larger structure of Europeans and West Asians on average. The larger boned groups tended to be those of the steppe heavy agricultural tribes of the NW.
      Legendary wrestlers like Dilip Singh and Great Gamma could in part tend to come from the NW due to cultural factors. However, I would hazard a genetic component as well. While, the average person in these groups, when controlling for environment, given these groups tend to heavily prize good nourishment and physicality relative to other Indian groups, isn’t honestly that much bigger than the average person in the other groups, the right side of the distribution in these groups I think may be larger, even given the likely slight right shift of the relatively normal body frame size curve.

      There is some emperical data on this, as in this study of Jatt vs. Bania girls in Punjab. However, confounders are very hard to control for in this situation. It was published in the early 1980s as well, when caste differences in lifestyle were probably even more pronounced than they are today. Environment is very hard to control for, in such studies.

      Regardless, S Asia has a lot of room to grow, in terms of both average nourishment and physical activity. The people who do a lot of work tend to not have enough food. Those who do have none and Victorian standards on the value of exercise, leading to a population full of skinny, obese, and skinny fat people, besides some army personnel and athletes. However, in this is radically shifting.

      I myself and I am pretty into strength training, so I pay attention a lot to the degree of participation in places. Every time I go to India and even on social media now, I am always pleasantly surprised to see more physical engagement among the youth in these activities than I would have expected.

      Oh and just to note, the Indian weightlifting team has pretty good representation from both the N and S. India’s most popular bodybuilder right now and most conventionally successful is Sangram Chougle, a non-brahmin Tamil.

      Jai Shree Ameen

      1. It must be the AASI ancestry giving us smaller bone structure as well as rounder , softer facial features given the rest of our ancestry is Iran_N and Steppe_MLBA.

        Also I think NW Indians are taller given that Kerela has a much higher HDI than the rest of India and probably a lower rate of vegetarianism.

        But like you I am happy to see young South Asians be more into physical fitness. Also much better hygiene, grooming and sense of fashion too.

        1. I will push back a bit here. Generally, farmers tend to be less robust than hunter gatherers facially. There are a lot of smaller boned but facially robust hunter gatherer groups in the Amazon for example. Additionally, many AASI heavy tribes like panniya do have some quite robust looking people. Australian aboriginals are another example.
          I think the softer and rounder features are from the Iranic farming population. Btw, new evidence in Raghikiri papers hows it was a mesolithic Iran HG related population with a common ancestor. Break from ancient iranic populations was further back than previously thought, not to be Iran N. Look at the Raghikiri paper blog post above and the paper link should be in there to go more in depth into that.

          Also, I am vegetarian. I have been my whole life. So are big chunks of my family. A lot of us are above average height. I notice that even in mandirs, And Jains and Sikhs had similar heights in the Secular Height study linked. The other thing, if you want a more specific example. My father worked for Isher Tractors way back when a few decades ago. A lot of manual laborers from Jatt and Gujjar areas would come to work. Many were vegetarian and many were not. Just by looking at them no one could tell who was and who wasn’t. What was consistent among them was high calorie diet with sufficient protein. Those who didn’t eat meat ate sufficient lentil and dairy products. Protein and calories, when held equal on a lacto-vegetarian diet, shouldn’t make a difference in muscle mass or growth. But yeah it’s just easier when you are eating low calorie food to get sufficient protein, if you eat some meat. Animal proteins do offer a slight edge over plant proteins in terms of biological availability. However, milk is an animal protein. So even that part is pretty null.

          The other thing is that Punjab is the bread basket, especially relative to its population. Sure Kerela can have a higher HDI but Punjab’s nutrition will be disproportionately good relative to it’s HDI given it’s agricultural status. Also Punjab ranges in an HDI of .70-.75 and Kerela just .75-.8. It isn’t a tremendous difference.

          There does I think appear to be a tendency actually, if anything, to overplay physical dominance for certain S Asian groups, particularly those of the NW, thanks to heavily pseudoscientific “Martial Race Theory” that came about post 1857.

          Culture is a huge thing. People in my family told me not to strength train when I started. I was 5’9.5 115 lbs, when I started lifting weights. Most men in my family are somewhere in that height range and weigh like 145lbs. After training heavy for a few years, I range from 170-180 with a reasonably good body fat and pretty good squat (~350 IPF depth), deadlift (430), and pause bench (250) numbers. Everyone in my family would tell me it’s not in my blood etc.

          Most Indian groups are way under their potential. But Punjabis are the least under, given that their culture tends to prioritize physicality more, particularly groups like Jatts.

          Jai Shree Ameen.

          1. Thanks for the response. I am curious how you can simultaneously say Indians have smaller bone structure compared to West Asians (so an association of larger bone structure with West Asian ancestry) and say that the smaller/rounder facial features are from Iranic farmers (so an association of long, sharp and angular features with West Asian ancestry) when most West Asians especially Iranians have bigger bone structure and are much longer headed and sharper featured than South Asians.

            I think tribal populations will be more robust than farming populations all things equal which is why some South Indian tribals are more robust compared to Northern farming populations. But South (I would actually say Eastern is more significant than Southern) Indian farming populations are less robust than North Indian farming populations. But in general I think Eastern Indians are less robust and South Indians and West Indians are closer to each other than East Indians.

          2. Robust Paniya? Robustness is width in comparison to height, they seem thin boned in comparison to their height.

        2. Because I think the agricultural Iranic related population was smalled boned. The big bone structure in iranic people comes from groups like CHGs, something S Asians tend to very much lack. Groups that are massive boned like N Iranians or Ossentians that dominate high weight class weightlifting and wrestling, tend to have elevated CHG levels.

          The West Asians of city states like Sumer, Mesopotamia or the majority West Eurasian derived ancestry of the IVC was likely small boned people. I think CHG influx is the key to big bones in many of these groups.

          Also look at Gilgit Baltistanis, they are about 70% IVC and 30% steppe. They tend to have very robust faces but quite small bones. And their faces are quite europid relative to most S Asians. Likely, the steppe influenced the facial robustness. But the small bones tend to come from the IVC. Many Indian mountain populations are actually like this. They have little AASI but tiny bones. Actually, the asymmetry in bone size between Gilgit Baltistanis and facial robustness is interesting to me. They are quite small boned compared to some agricultural tribes of Punjabis and Haryanvis who tend to be closer to medium boned on average. Granted, no s asian groups are big boned on average, from what I’ve observed.

          Additionally, the separation from Iranic people was more in the distant past than previously thought. The ancestor in common is one even before the Iran Mesolithic HG. That gives sufficient time for divergence, particularly if there is a lifestyle change that results in shift of selection pressures to smaller boned peoples, as what long term farming tends to do.

          I think tribals are robust from HG influence. I think groups that are high in steppe tend to be more robust from Steppe HG influence. Those that are more similar to IVC look I believe are more soft featured.

          1. Are there any facial reconstructions from Raghikiri or anywhere else in the Indus Valley?
            As far as I know there isn’t much CHG or Steppe MLBA in modern Iran. So unless Anatolian farmers were much more robust than East Iranian farmers I fail to see why modern Iranians are much more robust.
            On top of that Pashtuns and Tajiks don’t have much CHG or Anatolian Farmer but are bigger boned than Indus Valley populations with longer heads and sharper facial features.

            Also people of the Caucasus are shorter, rounder faced and softer featured than Iranians imo.

          2. You are making the mistake of defining ancient phenotypes not from ancient bones but rather from modern descendants thousands of years later who have had very different kinds of selection pressures. Do we have CHG skeletons for some comparison with known Iranian hunter gatherer skeletons? We don’t, so we can’t compare them for now.

            The Iran HGs that moved to India would be more like the Hotu cave people than the Ganj Dareh farmers. And the Hotu cave females were stated to have features similar to the ice age European hunter gatherers.

            The gracialization in India is a more recent phenomenon, likely tied to the onset of agriculture in the neolithic.

        3. I think Anatolian farmers were robust than Iranic Farmers secondary to CHG common origins. I am talking about body frame and not facial robustness. The two don’t have to be congruent. Those groups with elevated CHG or CHG related ancestry tend to be bigger boned.

          And Pashtuns and Tajiks have both elevated steppe and caucus relative to populations of S Asia.

          And facial morphs need to be correlated with AASI vs. Iranic related, given IVC had a cline in and of itself. To my knowledge, a robust set like that don’t exist to draw phenotypic trend conclusions from.

          Figureines are rather limited as well

          1. So you think Iranians are close to South Asians in terms of robustness given the paucity of Steppe_MLBA and predominance of Iranian farming ancestry?

        4. Btw, the reason I think CHG and related ancestry is bigger boned is because of cold adaption. Body frame comes more from that in Eurasian populations. Warmer weather promotes, thinner bones and a lankier frame for better surface area to volume ratio for heat dissipation. Those with ancestry from more cold adapted groups tend to be shorter and bigger boned. Of course, local environmental factors over a long period of time can also contribute to divergence from this.

          I think the Anatolian farmers had more cold weather adopted ancestors than Iranic ones. The end result is bigger boned people to start with in terms of just base size. And Iranians still do have more CHG than Indians.

          Facial robustness is more a matter of lack of neonatany. Neonatany tends to be correlated with agriculture. So ideally, a cold adapted hunter gatherer would have both facial robustness and body robustness. Hence, what is seen in certain European populations today.

          So the key to all of this is not only farmer vs. HG ancestry but the type of farmer ancestry and proportion of ancestry that derives from warm vs. cold adapted groups. Facial robustness comes from being HG in ancestry. Body robustness comes from more cold adapted ancestors.

          Cold adaptions promote, longer torso, shorter limbs, thicker skeleton, etc. Just a lower SA:volume ratio to maximize heat conservation.

          1. That makes sense. Thank god for some minor Steppe_MLBA ancestry preventing us from being super gracile.

          2. I’m guessing Balochis are the best approximation for what the farming ancestry of South Asians look like. I still feel like they are more robust and sharper featured than most South Asians.

          3. Baluchistan is a word which consists of two Serbian words – Balochi (i.e. Belici) and ‘stan’. Belici (from Bel – the first Aryan leader aka Dionysus, Bacchus, Nimrod) were the Serbian tribe which came to SA with the first Aryan leader. Belici is now a modern surname, meaning – Whites. For e.g. Bel-grade=White City.

  27. the replacement of gods is not a big deal. most of the greek gods have non-indo-european origins (indigenous, or foreign like aphrodite/astarte and dionysious) . zeus is the only who is indisputably indo-european

    1. Do you have any opinion on whether the Mittani were Indo-Aryans or not? They definitely had cooler names than modern Indians.

      1. Balochis are an iffy estimate too. Baloch have a lot of iranic IVC but other components as well. The only group that resembles the original indus people conponents wise are those on the aasi end of the spectrum aka Punjabi Dalits, Patels, and mide caste S Indians.

        Also, look at a group like Kashmiris. They have less steppe and more AASI than Haryanvi Jatt but are streotyped to look the most europid. Local influences matter too. We are talking thousands of years. And artificial selection standards based on culture can really select for certain phenotypes with relatively less overall autosomal genotypic alteration.

        Anyway facial “robustness” and “sharp features” are pretty pseudo scientific statements. You nerd a huge supply of cranionetric measurements and to control for diet and environment to show this.

        Feed balochis the high carb, low protein, high salt and sugar nearly vegan Guju diet, have them do less labor, and make them shave their beard and suddenly they will look less “robust.”

        A lot of robustness is body fat. Look at people’s jaw lines before and after losing weight.

      2. Lol. In which century are Shattuara, Wasashatta, Tushratta etc cool names?

        With names like these they were probably bullied at school and decided not to speak their mother tongue…;)

    2. Really, it is not a big deal. Greek colonists copied the mythology from indigenous Serbs. Aphrodite was thousands of years younger copy of the Serbian goddess Priya. The later, Roman copy was Venus. Dionysus was actually the first Aryan leader, although Greeks do not (want to) believe that Aryans existed at all because it was before their time. Indo-Europeans (Greeks were not IE!) in this (and any other) context were Serbs, who were Greeks’ only neighbours on the west, north and east and who gave to Greeks the mythology and literacy. In Iliad, ‘Greeks’ Mt Olympus gods are involved in the battle. They, however, haven’t seen Mt Olympus in the following 1000 years. What a falsification.

  28. @Jatt_Scythian

    For your amusement 😉

    And technically 5’7.5 115, when I started. I was a full two inches shorter, when I first picked up a weight at 17. I trained like an idiot till I was 20, but I’ve trained properly for the last 5 ish years or so. I kept growing till I was 21. Everyone in my family has constitutional delay in growth, so we look super young till like our early 20s lol.

  29. Also, just want to put it out there that Gujaratis, my ethnic group, tend to have the worst diets of the subcontinent. Rural Gujarati diet is still somewhat okay. The urban one is pure trash.

    It is this sugar, salt, and oil laden diet with tons of carbs and unhealthy fits with minimal protein. It results in obese and skinny fat people. The complete lack of physical culture, besides some farmer (some non-urban Patels), Rajput (only groups to contribute, until recently, to Indian army), and tribal groups (ie. Bhils), is also a hallmark of Gujarati life. It is really bad.

    So yeah, bloated faces where facial aesthetic potential is so far from maximized it is pathetic, is quite common. Look up two things. Just look at the effect of losing fat and having a thicker neck (correlates with muscular hypertrophy) on facial attractiveness. You will see an insane difference.

    Browns, in general, but especially total fails (on a group level at least) like Gujaratis must eat better and be more physically active. Sitting on your ass and watching Kasam Se, while snacking on Jalebi and Fafra, will make you look like a Jalebi at best and a Gulab Jamun at worst.

  30. I don’t know much about genetic history so forgive me if my questions sound stupid. But how far would have to travel back in time to find a pure AASI? Did any potentially AASI community survive into historical times? Do we have any AASI sample?

  31. Hey Razib, is there any evidence that AASI is significantly connected to Australo-Melanesians/Oceanians? This is often claimed by many people present on anthrofora but I’m not able to find any indication that they are closely related.

    The closest thing to AASI found in the subcontinent seems to be Andaman Islanders and after that it is roughly equally distant to east asians and Oceanians based on G25 and FST.

    But I’m wondering if the reason Oceanians are as far as they are is due to the relatively large amount of Denisovan admixture they have? We know a small amount of any diverged human ancestry can pull a group far away from another one which lacks that ancestry. I would estimate that ancestry from an archaic hominid could have an even greater effect even if small.

    When Australian aborigines are run through gedmatch when there is no Oceanian component, they get more South Asian than anything else with substantial amounts of east Asian as well.

    I’ve also seen basic global PCA plots which put Oceanic populations firmly close to east asians, along with native americans. Im not sure what to make of this, other than perhaps this is just looking at the 3 most basic components(SSA, east eurasian, and west eurasian).

    1. 1) archaic inflates difference
      2) also bottleneck means they are more drifted (long branch on tree)
      3) the ‘attraction’ btwn s. asians and oceanians (australian aboriginal ==> s asian) is because in the ‘east eurasian’ clade the ‘southern branch’ of AASI, hohabinians, and oceanians have deep affinities 40K BP.

      1. So is that southern branch of the East Eurasian Clade called “ancestral south Eurasian”? This appears on one of the gedmatch calculators.
        I wonder if it’s possible to isolate the non-Denisivan segments of Oceanians to see how close it is to us. Papuans seem much further from mainland Asians than aborigines are and I can only guess this a result of their higher Denisovan compared to Australian natives.

  32. Is there WSHG ancestry in South Asia like the ancient Central Asian samples had?

    Also interesting that both Q1a and Q1b were found in SWAT.

  33. Actually, I need to make a correction, we do have CHG skeletons, well more like one CHG skeleton (the older one was only a temporal fragment). And it doesn’t look more robust than the way Hotu cave females were described. I’ll try and see if I can find it’s measurements.

    The jaw seems less massive than the Hotu female’s.

      1. I respect thewarlocks perspective but I’m not yet convinced Iran_N was this facially and skeletally gracile population compared to more facially and skeletally robust Anatolian farmers. And slightly offtopic I’m not buying that Iran_N was super dark and hook nosed and ANatolian Farmers were blue eyed with straight noses.

        1. The point is that Iran_N is irrelevant to south Asia’s modern day phenotypes since it isn’t an important ancestral component, Iran HGs are and you can have a look at the Hotu cave examples to see what they might have looked like. The faces were well built with massive square jaws and a well defined chin, the Hotu 2’s chin and ramus look more defined than those of CHG KK1. Furthermore, Hotu 2’s cranial measurements are similar to those of the Sidelkino EHG in terms of raw size and this is in spite of Hotu 2 being a female and Sidelkino being a male. Sidelkino was a male bordering on a female form while Hotu 2 was a female bordering on a male form, in both cases, the final determination of the sex was based on the hip form.

          The following view of Hotu 2 looks very much like one of the Mahadaha Ganga mesolithic skulls:

          I am not sure about straight vs convex noses, Hotu 2’s nasal profile angle was 65 degrees vs the 56 degrees for Medieval Norwegian females. However it didn’t look as high as cro magnon 1’s. In general, the Hotu females had lower than average nasal heights (vs the cro magnon average) so it doesn’t leave a lot of room for a very convex nose. GD was heterozygous for the fair skinned genes (not fully selected the way Anatolian early farmers were) and IDK about the Hotu details or if they have ever been published at all.

          1. That was an awesome comment. The idea that the population of the IVC was likely a 2 way mixture of AASI and Hotu and not Iran_N totally went over my head.

            Could you simply your point about noses? I’m still pretty dumb when it comes to describing phenotypes. Thanks. It sounds like you’re saying Iranian HGs did not have convex noses so I wonder where the feature came from later.

          2. I am not saying that Hotu HGs did not have convex noses, it is just that it wouldn’t be as convex as those of people like copper age Iranians and modern Balochis. One of the factors is nose height: if you have a tall nose then there is a greater distance over which the nose bridge can have a noticeable upward or downward slope. Heck with very tall noses you could even have a wavy profile where it goes both upward and downward along the way.

            With small nasal height however, you would would have a less extreme profile, closer to the straight nose. At most there would be a small upward slope because the nasal profile angle is only ~9 degrees higher than Norwegian medieval females along with a shorter nose than the said group. Heck it doesn’t even have to result in a convex nose, it could have been a more strongly forward projecting type instead because of the short nasal height.

            One of the aspects of the copper age Iranians was gracialization in comparison to the HGs- that is – a longer and narrower face along with a longer nose, this could have allowed for a more convex growth in the copper age Iranians.

  34. Also can gracilization be reversed? I know Southern Europeans are making massive strides and in heights and Eastern Europeans today are getting longer headed.

    1. You seem to have a confused understanding of what is and isn’t gracile, though you aren’t the only one with the confusion. What people often describe as sharp features would be gracile- narrow and long face along with narrow and long jaws with a noticeable angle. A robust face on the other hand would be bulkier and lower with a nearly horizontal jaw profile.

      Likewise- simply gaining height alone will make you more gracile if you don’t have additional width as well. I am assuming that you have read about eastern European changes as well, have you also read how now there is a selection for a more ectomorphic type- taller and leaner, in other words, a more gracile body shape.

      As far as head length is concerned, that alone doesn’t tell you much, you can have a longer, more gracile head or a longer more robust one depending on the details. For instance, both Gracile Mediterranid and Faelids have mesocephalic and chame-orthocranic skulls with curved occiputs as far as non-metric ratios are concerned, but in spite of the similar basic ratios, Faelids are much more robust than Gracile Mediterranids.

      1. Thanks. I fully admit to being a novice to this. I know steppe MLBA were often sterotypes as Corded or East Nordid. Isn’t Corded relativley gracile to CM influenced phenotypes?

        1. Yes, Corded and Nordids in general are intermediate between (heavily agriculturally selected) Mediterranid and (non-agriculturally selected) cro-magnid.

          1. And steppe MLBA were exclusively Corded or did they have Med proper and CM admixture?

            What about Yamnaya/Steppe EBA?

  35. I know Steppe_LBA (Yamnaya) was dark but isn’t Steppe_MLBA (Sintashta-Andronovo) mostly associated with light hair and eyes?

    Was it also true they had Dinaromophic/Iranomorphic tendencies?

    Also from the source you posted those are some low ass nasal indices no?

    1. Razib ran the Sintashta genomes through hirisplex and the results weren’t as light as modern northern Europeans. This is also true for early bronze age Baltic populations, the selection largely ramped up in the later parts of the bronze age and the iron age.

      The nasal indices aren’t very low considering that these are dry skulls.
      Ranged from leptorrhine to mesorrhine.

      IDK about any Dinarid tendencies.

      1. I mean that’s to be expected though right. They were still located further South than most of Northern Europe. But I doubt they were swarthy- probably in Central European Range I guess.

Comments are closed.