Varna is Indo-European and jati is Indian

A casual comment…most Indo-European societies seem to have originally had some sort of occupational caste system. I’m talking here of Dumezil’s trifunctional hypothesis, warrior, priest and commoner. But only the Indian subcontinent has jati.

I was thinking about it when reflecting on work to come out soon from David Reich’s lab on ancient Pontic steppe ancestry in Bronze Age Greece. There is no stratification by class when it comes to steppe ancestry. From the talk:

In the Balkans, we reveal a patchwork of Bronze Age populations with diverse proportions of steppe ancestry in the aftermath of the ~3000 BCE Yamnaya migrations, paralleling the linguistic diversity of Paleo-Balkan speakers. We provide insights into the Mycenaean period of the Aegean by documenting variation in the proportion of steppe ancestry (including some individuals who lack it altogether), and finding no evidence for systematic differences in steppe ancestry among social strata, such as those of the elite buried at the Palace of Nestor in Pylos.

So why is India so different? One hypothesis that some make is that the Indo-Aryans were racially so different from the indigenous people. But I do not that that is the issue. Instead of bringing strict endogamy to the subcontinent, the Indo-Aryans adopted indigenous forms. There’s genetic differences indicating strong endogamy across South India among non-Brahmin groups. There is also a ‘mystery’ in terms of how the IVC was organized sociopolitically. I think I have a possibility: jati obviated the need for central political authority.

36 thoughts on “Varna is Indo-European and jati is Indian”

  1. //I think I have a possibility: jati obviated the need for central political authority.

    Yes the Jati Matrix is a system with strong inertia and though hierarchy is baked into the Jati system, it can be easily imagined that this hierarchy would not manifest in monumental architecture or other tel-tale signs of hierarchy seen in archaeology.

    Even Kalash who are clearly Indo-Aryan do not have jAti like endogamy ASAIK – clearly they were the closest to RG Aryas – more so than Persians, Medians and various Iranic groups

    Lot of things start to make sense if we assume some sort of loose JAti like matrix being in place in IVC – be it craft specialisation which continued post IVC collapse and many other things which.

    I had touched upon this in a piece of Jati Varna a year or so back.

    1. Hi

      Checked out your earlier post. Even if it is speculation it is fascinating and persuasive. Could you mail me the references, please?

      1. Hoju:
        My piece

        Raghu Cidambi:
        Specific references : (non famous)
        Johannes Brockhorst’s work – Greater Magadha, his next book other papers etc.
        All of Witzel’s open works.
        The Snake and the mongoose
        Naturally Original MB
        Razib Khan’s blogs and substack – other books no genetics – Reich/ Tony Joseph
        Ambedkar’s book on Shudras and especially his book on Gita.
        Irawati Karve’s original work on Hindu society

        Have read the mainstream works on Ancient history by mainstream historians like Thapar, Sharma, Singh etc but nothing specific about from those books apart from broad strokes.

  2. That too probably took a while to re-emerge, unless there were jatis prior to 2000 ybp in the Indo-European era which were more fluid. Speaking of which, can the current evidence distinguish between:
    A- A Situation where the Indo-Europeans first near totally supplant the jati system with a varna system around ~3500 ybp, and this change gets eroded until finally around 2000 ybp jati becomes important again. (No role of jati for ~1500 years scenario).

    B- A situation where the Indo-Europeans introduce varna, but jati remains a thing, albeit changed in its rigidness (lowered inertia) for some time, and so between ~3500 ybp to 2000 ybp both existed. Then by 2000 ybp, jati regains its inertia and becomes more fixed as it was before (or long before) 3500 ybp. (weak role of jati for ~1500 years scenario).

    1. @ DaThang
      The 1500BCE to start of CE is usually the timeline given when the so called ANI-ASI did undergo some degree of mixing, leading to the current base makeup of what a South Asian/India is.

      So it’s possible Steppe migrants reaching certain scale & socio-political relevance lead to a Class-Mixing (like what used to happen in China every few centuries when Dynasties would collapse & rise, leading to a relatively homogenous genetic mix).

      Maybe South Asia had far stronger cultural traditions of Class-stratification (jati-varna whatever) and hence even the 1500 year lull could only do so much and Caste system came back with a vengeance.

      Sort of like 1000 years of Foreign occupation post 10th century CE and the genetic mix holding steady, even though still having some mixing, but relatively not as one could expect.

      jati obviated the need for central political authority.

      This also brings up questions of Mythos of IVC. Like there must have been some powerful ideological pathos to hold society in check otherwise there would have been civil wars since why would people in Lower Classes just accept to get short end of the stick across generations.

      1. That is the gist but my question is whether the current layer of endogamy is masking another layer during the time when steppe and local populations are thought to have mixed. I do not meant during the ivc period BTW, in both these scenarios IVC could have had endogamy bit that isn’t what I am getting at. Like can the current data rule out a scenario of a quick mixture, followed by endogamy before 2000 ybp which was masked by restructuring of endogamy in a different pattern at 2000 ybp.

        Basically quick mix at 3500 ybp + early endogamy (some time between 3500 and 2000 ybp) which is masked by a later endogamy starting at 2000 ybp-> can this be ruled out based on genetic data?

      2. // why would people in Lower Classes just accept to get short end of the stick across generations. //

        They didn’t as region back then was populated with nomads & groups {clans, chieftans, tribes etc.} & even with their regional affiliations with urban / proto-urban centres many groups preferred to remain independent {i.e. beyond walls of state protection} as a means to challenge the newly formed polity of ‘states’ or to maintain the hold in regions they presided over {as tributaries}.

        Artisan guilds {Shrenis} used to move to diff. kingdoms if they were not satisfied by patreons or treated properly by royal elites. There are sources where kings not only donated land to Brahmins but to the guilds as well.

        E.g. –

        Land and Society in Early South Asia Eastern India 400–1250 AD {Book}

        This is not to say that certain groups or sections of society did not suffer but rather to suggest that Indian society was dynamic, with it’s core focus on interdepence rather than morality & it lacked centralized ‘state’.

        E.g. –

  3. This is the distribution of jati and varna in the Vedic literature (Source: Uma Chakravarti).

    Rgveda – Varna (23), Jati (0)
    Samhitas – Varna (97), Jati (0)
    Brahmanas/Upanishads/Aranyakas – Varna (63), Jati (3)
    Vedangas – Varna (98), Jati (99)

    In the Vedangas, the term jati is used in the sense of “species” as is the context across all IE languages – Latin (gens), Avestan (zato). Jati is frequently used as an identifier for different kinds of animals – as in cows, horses, sheep.

    The word “gotra” (lineage) is far more frequently used than Jati as a marker for people (jana).

    In spirit and word, the word jati used in Ancient India has absolutely no organic relevance to the colonial term as we use it today….. zero linkages with either commensality or connubiality.

    The propensity to project stratified endogamy (colonial interpretation of jati) into the Bronze/Iron Age of India is not backed by historical or textual evidence.

    1. I mean the genetic evidence, semantics aside. Can the current evidence distinguishe between a scenario of endogamy -> loosened/no endogamy -> return of endogamy and endogamy-> varied endogamy-> return of solid endogamy.
      The current research shows endogamy started at most 2,000 years ago, but according to the data at hand is it possible to glean if there is a more ancient layer of endogamy that is being masked. Do you have info on that?

    2. // This is the distribution of jati and varna in the Vedic literature (Source: Uma Chakravarti). //

      Can you provide source details like book name or paper’s title ?

  4. I have been suggesting it but i attribute ‘endogamy’ to regional tribal formations & negotiations over time instead of ideology, prejudice etc.

    Temples, Rituals etc. emerged as mediating / negotiating ground {even if Brahmanical in their orientations} where connections among elites were formed while commoners remain entrentched in regions.


    Twitter thread with theoritical musings along with academic resources {Papers, Video presentations & academic discussions} –

    Related –

    Other aspects which people fail to discuss when they discuss Caste is ‘Language’ & ‘Region’ –

  5. Are you saying the incoming Aryans learnt “caste based society” from the IVC?

    What evidence is there is for racial stratification in the IVC please? Is it the fact that groups like Lohanas and Kalash have very low AASI levels while others from the region have significantly higher levels (despite living alongside each other for presumably long periods)?

    Do you think this stratification happened more in the epicentre of the Indus civilisation (present day Sindh), but less so in other parts of the subcontinent?

    To make your argument stronger, the Aryans must have mixed with Turkic or E.Asian populations in the Kazakhstan region, but we see no race based caste system or stratification there.

    It’s all quite confusing – if the Aryans just adapted to the indigenous society, how did they have such a massive influence on religion, language, culture etc.

    1. ‘ if the Aryans just adapted to the indigenous society, how did they have such a massive influence on religion, language, culture etc.’

      Like i dont wanna brag, but we are the best… 🙂

      1. @Saurav We? Really dude, are you pure Aryan? 😂

        @dathang What were the two main phenotypic groups in the IVC, and any evidence to back this up?

        1. There were 2 common groups: one was similar to (though not the same as) Mediterranids, let’s call this group 1. The other has been variously classified as proto-australoid or eurafricanid or cappelid among others, let’s call this group 2.
          Groups 1 and 2 were the most common groups and group 1 was more common than group 2 at least in the western IVC zone. Group 1 had dolicho-mesocranic skulls which were ortho-hypsicranic as well. Group 2 had more dolichocranic skuuls (lower cranial index) which were low orthocranic (lower height length index). Group 2 skulls were more robust and also had a higher cranial capacity than group 1.
          I do not know about their relative frequency in the eastern ivc zone. My hypothesis is that group 1 is indicative of chalcolithic migrant ancestry from Iran while group 2 has less of that and is more of an Iran hg + local aasi descent. I base this on the fact that chalcolithic western ivc shows a deviation in dental traits while Neolithic western ivc, mesolithic Ganga cultures and the chalcolithic regions far south and east of western ivc cluster together. Although one critical link would be to check the dental traits of group 2 and see if they too cluster closer to the neolithic-mesolithic cluster in comparison to group 1 and I haven’t seen that being done yet.

    2. “What evidence is there is for racial stratification in the IVC please?”
      IDK about racial stratification/hierarchy, but if you are talking about presence of groups with distinct appearances, then there were 2 main ones and a few minor groups.

      “Do you think this stratification happened more in the epicentre of the Indus civilisation (present day Sindh), but less so in other parts of the subcontinent?

      “It’s all quite confusing – if the Aryans just adapted to the indigenous society, how did they have such a massive influence on religion, language, culture etc.”
      Remember that the impact of Aryans on DNA is minor, so they weren’t an outnumbering element. Once the original input has happened, certain things get carried on, while others might get syncretized

    3. According to recent studies, it seems that caste as a system of social satrification(by profession) was developed indigeniously in India and was present before the arrival of Indo-Aryans.

      The invading Aryans interjected the varna or a system of color/race into it by placing themselves in top of the already present caste system.

  6. Doesn’t the genetic evidence point to strict endogamy by jati arising long after contact between Indo-Aryan and indigenous South Asian people?

  7. Also worth mentioning that the general framework of an IVC substrate influence on IE concepts that make it into Hindu religion and society in a way that is uniquely South Asian and different from other branches of IE influenced societies is a worthy one whether or not it is supported in this particular instance.

    For example, one probable IVC influences include the very detailed exposition of the concept of *Prana*, compared to other IE societies with a vaguely similar but far less specific concepts like Greek *pneuna* or German *od* that at that level of vague generality is found in many non-IE cultures as well.

    Another example is the existence of Hindu gods that are not basically human in appearance, unlike other IE gods where are basically human in appearance except possibly for size.

    It feels to some extent like the IE religious concepts were retconned into the existing IVC religion that never really died, rather than the reverse.

    It is also very plausible that Hindu dietary prescriptions are primarily substrate influences.

    It is also conceivable that pre-IE culture had something like Jati but more like Medieval guilds that were not endogamous, or like Middle Eastern clans that are lineage based but without an occupational association and not so cousin marriage oriented (possibly influenced by IE patrilocality and spouse exogamy), and that the endogamy feature of Jati was a later local innovation.

    But the comment of @Ugra re the timing of literary mentions of the term, and the gotra v. jati distinction, is quite persuasive.

  8. If just a few % of marriages occurred across jati lines, that could have been enough to preserve the institution socially while masking it genetically. It’s easy to imagine a system of otherwise strict endogamy that allowed rare exceptions: marriage alliances, fostering/adoption, tolerating the occasional eloped couple. The real innovation around 1500-2000 years ago could have been the full anathema of exceptions.

  9. @Razib

    Jati is surely IVC. There is no analog of jati west of the Indus. No similar word in other IE languages west of the Indus. Glad you reached this conclusion. I have been tweeting this for a while.

    IVC was larger in population than Mesopotamia and Egypt combined. The Indo-Aryans who entered the subcontinent had no choice but to adapt. They were heavily outnumbered.

    However Jati in the mid 2nd millennium BC did not have the characteristic of strong endogamy. It was rather a social organization based on occupation. Surely there was a tendency to endogamy – but it was not mandated by religion.

    The IVC leadership (traders, priests) mixed with the IA invaders and over the centuries a new brahminical class arose aligned with the IA. That class was a scourge on the population by its demands. Buddha and other Sramanas rebelled against the brahminical culture and became very successful under Ashoka. The brahmins reasserted their hegemony with Pushymitra Shunga’s murder of the last Mauryan monarch. Only then did the strict religion supported endogamy of jatis start as codified in the Manusmriti. This was the way for the brahminical revolt to ossify their ascendence, by taking advantage of a jati based culture and converting the jatis into a system of graded inequality using a varna overlay.

  10. If Brahmins were buried their graves would have the least amount of goods 😉

    Higher in Social hierarchy + High steppe doesn’t correlate to their economic standing.

    Varna are classes
    Jati are castes

    Vedic had Rajanaya

    Post Vedic many Kshatriyas Indian and Iranian were absorbed into the Indian social structure.

  11. This book, “The Creation of Inequality” argues that unranked but extremely endogamous social groups (i.e. the functional equivalent of “jatis” without a hierarchy or a well developed sense that one group is superior) are common throughout tribal peoples in the cultural “frontier” zone where India, Myanmar and Tibet meet today.

    The groups have different origin stories, different religious traditions and tutelary deities, and strict rules on endogamy. The authors don’t use the term “jati”, but it’s interesting to me that they’re geographically next door to the core Indian cultural sphere although not really part of it. Maybe they offer a clue to how jati & caste more generally originated?

    1. Lots of Indian people in Plano, hence the Mexican lady had a chance to develop some prejudice.

Comments are closed.

Brown Pundits