The rise of Indicus!

By Razib Khan 25 Comments

A few years ago an ancient DNA paper on cattle was published, Ancient cattle genomics, origins, and rapid turnover in the Fertile Crescent. It’s a pretty good paper with interesting results. The paper confirmed pretty strikingly that there was a punctuated and massive expansion of indicus ancestry across the Near East between 3,500 and 4,000 years ago.

The interesting aspect of cattle is that there are really two species that intermix. Using mtDNA researchers estimate indicus and taurus diverged 300,000 to 2,000,000 years ago. But the main thing you have to remember is cattle generations are about 20% as long as human generations. So 300,000 cattle years are equivalent to 1.5 million human years. And, for technical reasons (smaller effective population size) one should probably assume mtDNA underestimates the divergence.

Ancient cattle from the Near East are all taurus. The PCA plot shows that most of the variance is on PC 1 which separates indicus and taurus (a secondary dimension is PC 2, between African and Near Eastern/European lineages). The figure at the top of this post shows that there is a massive jump in genome-wide indicus ancestry across the Near East between 2000 and 1500 BC. As the authors note this can’t be diffusion; the jump is too sudden and sweeping.

So what happened during this period? As noted in the paper: Bronze Age civilization almost collapsed around ~2000 BC. More concretely, after 2000 BC is when we see evidence of Indo-Europeans in the Near East. The Indo-Aryan Mittani show up in Mesopotamia in ~1600 BC. The Indo-European Hittites, the Nesa, are known from Anatolia a bit earlier.

This is also the period that small, but detectable, levels of “steppe” ancestry show up in some ancient samples.

Before this paper, I would have leaned to the position that the Mittani Indo-Aryans migrated directly from the Sintashta homeland without much contact with Indian Indo-Aryans. These data are too suggestive of a widespread zone of expanding agro-pastoralists that existed between western South Asia and the Near East between 2000 BC and 1500 BC.

One of the things we know from the barbarian period during the Fall of Rome is that barbarian groups had strong channels of information flow. For example, a group of Saxons arrived with the Lombards in Italy in the second half of the 6th century. But, through various channels, these Saxon warriors learned that their co-ethnics had established dominance in what was to become England, and there are texts which allude to the reality that they decamped and crossed the Alps, presumably on the way to what was going to be England. The point here is that there was a “Saxon international.”

Aside from the Mittani the evidence of Indo-Aryans in the Near East is tenuous, though some of the Kassites of Babylonia may have had Indo-European affinities. There is not nearly as strong a genetic imprint of steppe in the Fertile Crescent as in Northwest India. The Hittites were very different from Indo-Aryans, who seem to have the closest relationship to the Slavic language family.

The indicus breed is adapted to tropical dry climates. It seems plausible that the Indo-Aryan international facilitated the spread of this breed in the centuries before 1500 BC.


25 Replies to “The rise of Indicus!”

  1. Really sad Indians didn’t manage to domesticate Bos Gaurus in the deep east & south – those would’ve been some heavy duty cattle 🙁 . Maybe even WarGaurs

    1. Those places like south India were outside of the range of domestication for a long time. The neolithic spread to south India around 5000 years ago and the Dravidian languages of south India coalesce around 4500 years ago.

    2. @GauravL
      The gayal/mithun (Bos frontalis) is a domesticated gaur. But they were domesticated somewhere between eastern Bhutan through the Arunachal Pradesh.

      1. Is it confirmed? I haven’t followed the latest literature? I was under the impression that there is still some debate whether it’s domesticated Gaurs or crossbred

    3. Domesticated gaurs certainly exist, although I don’t think they’ve ever really taken off on a large scale.

      Most of the wild cattle species across Asia, and there are a lot, seem to have been domesticated at one time or another.

  2. Aside from the Mittani the evidence of Indo-Aryans in the Near East is tenuous

    To be honest, it is a stretch to call Mitanni Indo-Aryan in the first place. The whole claim rests on just one word – the word for numeral one: aika (contrasted with PIr *aiva > Av aeva, whose cognate survives as “eva”, lit. only, in Skt). And aika is the same as in Vedic. Classical Skt lost the diphthong to get eka, whence all words for one in modern IA languages (Hindi-Urdu ek, Kashmiri akh, Punjabi ikk etc)

    Calling Mitanni “Indo-Aryan” is therefore generalising from a sample set of (ironically) just one data point. Makes the claim quite tenuous in my opinion.

    I think it is equally if not more likely that Mitanni blokes were a more archaic Proto Indo-Iranian stage split-off. In other words, it makes sense to use the phrase “Indo-Aryan” only for the stage of language that we attest in (Greater) India. India is the cauldron that cooks Indo-Aryan just the same way England is the vessel that created Old English (from Saxonic, Kentish/Jute and Anglic dialects).

  3. Finally this paper is being discussed!

    Razib, what do you have in mind when you say “Western South Asia”? Which archaeological complex? You have come this far, but are you willing to go a bit further!

    IVC is the only ancient civilization in that time period between West and South Asia that has attested pictorial depictions of the Indian Zebu in great detail. The general consensus in the IVC community that the seals were trade/tax/transaction tokens. This point was further buttressed by Kenoyer when he established that the seals were always found in great quantities by the entrances to the Indus cities and also in the Main Street, implying a point of entry transaction and a consumption terminal.

    More recently, Bahata Ansumali’s excellent paper goes further in analyzing the seals to its logical end.

    The “Indo-Aryan International” that you speak of – if it was a commerce network, then the IVC Zebu seals could only have been the signature of this trade. This hypothesis coherently brings together an archaeological find and a genetic conclusion.

  4. seems these cattle were distributed at the tail end of the “IVC” period. look at the plot. it’s in the interval 1500 to 2000 bc. i don’t think this was trade. these were male cattle stolen and deployed elsewhere. and yes, they’re from the ivc originally

    1. @Razib

      IVC zebu and bison seals have been found in Ur, Mesopotamia, Persian Gulf, Farmana, Kachi and Quetta – this is basically a breadcrumb trail for the trade.

      Many of the archaeological papers regarding this trail were written in the 60s onwards to 90s – Hallo & Buchanan, Kjaerum. Mitchell et al. More recently, Shinde and Konasukawa posit the same. The genetic paper confirms the trade from the other end.

    2. Are you trying to imply that the invading olive skinned and chiseled jawlined Aryans with their spoked horse ridden real chariots performed war (Gavisti) on the darkies of the Indus, stole the cattle and then sold elsewhere? 🙂

  5. IVC zebu and bison seals have been found in Ur, Mesopotamia, Persian Gulf, Farmana, Kachi and Quetta – this is basically a breadcrumb trail for the trade.

    these predate the gene flow. he figure is clear that the gene flow happened btwn 2000 and 1500 bc. this is when ivc trade with the west stops and ivc itself declines/collapses

    1. The abstract says the process of mixing started 4.2 kya.

      Such long distance trading could only be sustained by highly efficient contractual and mediation networks patronised by the local kingdoms. There is also this onomastic paper from 2019 (Simone Gentile) that implies that the Mitanni kings took on Indo-Aryan names that are based upon the Rgveda and Avesta.

      The archaeological artifacts, genetic conclusions and king names are strongly suggestive of a well established cattle trade that had begun to influence local kingdoms in order to sustain the network. Somewhat like the English traders in India.

  6. In this writeup, I would like to document facts related to Mitanni. What they mean — and how strong they are — is for the reader to decide and verify:

    First of all, the claim that the IA designation of Mitanni is based on one word is incorrect:

    Quote from the paper [1]
    The discovery of ‘Aryan’ looking names of [Mitanni-] princes on cuneiform documents in Akkadian from the second half of the second millennium B. c. (chiefly tablets from Bogazk6y and El-Amarna), several doubtlessly Aryan words in Kikkuli’s treatise in Hittite on horse training (numerals: aika- ‘one,’ tera- ‘three,’ panza- ‘five,’ satta- ‘seven,’ na[ua-] ‘nine’; appellatives: uart- tana- ‘circuit, course [in which horses move when being trained],’ asua- ‘horse’ 3), and, finally, a series of names of Aryan divinities on a Mitanni- Hatti and a Hatti-Mitanni treaty (14th century B. C.)

    Secondly, the order and names of gods on the Mitanni Treaty correspond to Rig Vedic hymn 10. 125. lbc [1]. The Rig Vedic Book 10 is accepted to be one of the 2 youngest books that were composed. There are other similarities too with Rig Veda that are mentioned in [1].

    Thirdly, according to Talageri [2]:
    If the classification of the Mitanni culture as representing a pre-Rigvedic stage of “Indo-Aryan” culture is correct, then the elements common to the Rigveda and the Mitanni (and also the Avesta) should be found in the greatest numbers in the Old Books of the Rigveda (which would be closest in time to the connection with the proto-Mitanni Indo-Aryans) …… However, we find exactly the opposite case, i.e. a huge mass of common elements is completely missing in the Old Hymns in the Old Books, found a few times in the Redacted Hymns in the Old Books (which were edited in the period of the New Books), but found in abundance in the New Books and in post-Rigvedic texts and later literature…..

    1) Thieme, P. (1960). The “Aryan” Gods of the Mitanni Treaties. Journal of the American Oriental Society, 80(4), 301.
    2) The Rigveda and the Aryan Theory: A Rational Perspective THE FULL OUT-OF-INDIA CASE IN SHORT REVISED AND ENLARGED 20/7/2020

    1. All this comment shows is a bucket load of confirmation bias and little else. There’s nothing specifically Indo-Aryan about:

      tera < PIIr *trayas (cf Nuristani tra/tre)
      panza < PIIr *panča (cf Iranic Bactrian panzo)
      satta < PIIr *sapta (cf Nuristani sat/sot)


      asua < PIIr *Haćwas (cf Iranic Saka *aśwā, Khotanese aśśa, Old Persian asa)

      The claim thus rests on aika, where there's a bimodality of *Haykas and *Haywas in PIIr, and no non-IA branch attested (not even Nuristani) uses the reflex of *Haykas for numeral one.

      1. @Slapstik

        You are a knowledgeable person regarding linguistics; I think much more than me. I agree there was a large controversy in designating Mitanni as IA. I am also writing this comment to learn something new.

        Firstly, since you consider that Mitannis were proto Iranians, won’t it be better to use Avestan to prove your point? It is closer to the source after all. Quoting from later languages can simply be passed off as borrowings or coincidences since you quote multiple branches. Various Indian languages such as Pali, Prakrit, Bangani, etc. can work as proofs too then. If you think I am not correct, do enlighten me.

        Secondly, it was the Mitanni treaty that also influenced the designation of Mitannis as IA, as discussed in the paper.

        1. @timepaas

          Firstly, since you consider that Mitannis were proto Iranians, won’t it be better to use Avestan to prove your point?

          The first rule of a sensible argument is to be able to express the other person’s position in a way they agree with. I never claimed that Mitanni chaps were proto Iranians and certainly do not think so.

          I am saying the following three things:

          a) there is very little linguistic data about Mitanni – just some loanwords in a horse-training manual written in another IE language (Hittite)

          b) linkage with IA ultimately rests on just the word for numeral one and is very tenuous (i.e. I see a +ve payoff in placing a bet against it with bookies). BTW I think they were not Iranic speakers either and all features of what we now call Iranic probably hadn’t even developed by then.

          c) Indo-Aryan is a meaningful term of the subset of Indo-Iranian languages situated in India (including present-day Pakistan), not outside – unlike Iranian which can be applied as easily to languages of people who never lived in (what’s now) Iran and/or never were Zoroastrian etc either.

        2. @Slapstik

          Your point regarding only some evidence being present is valid. I nevertheless disagree: The language category is created by linguistic evidence and not by its geographical distribution. Here itself, your premise is wrong. Also, it is the quality of evidence that decides rather than the quantity; though more evidence undoubtedly is better. Furthermore, how can later languages influence an ancient language? To arrive at the correct conclusion, you should analyze the data using Avestan and Vedic Sanskrit rather than later languages; I will certainly accept what you postulate then. Right now, your analysis is wrong and invalid. You don’t even consider all proofs, and neglect Mitanni gods and people’s names as if they don’t exist.

          Here, let me show you since you are so shy of utilizing Avestan, and resort to linguistic absurdity like using later languages:

          tera < Ved. tráyas, Av. θrāiiō
          panza < Ved. pañca, Av. panca
          satta < Ved. saptá, Av. hapta

          With regards to your present results, even if I take what you say at face value, IAs and Iranians had a common culture in the beginning; calling Mitannis as not IAs and Iranians, or whatever, won't change that. The historical records of them demonstrate their similarities with Vedic Sanskrit and Avestan. Names of Mitanni kings, their gods, etc. all show a strong Rig Vedic and Avestan influence. Do you think all this is a coincidence?

          Besides, IA languages are not limited to India. Firstly, you should know Dardic languages are spoken outside India too. Secondly, Romanis originated from Central Asia. Thirdly, IA languages have also spread to China, Japan, etc.; this was a later development though, and has no bearing on the present debate. Stop writing absurd stuff.

  7. The abstract says the process of mixing started 4.2 kya.

    this is why you get on my nerves. you just look for what confirms whatever you believe.

    in table s2 of the supplements it is explicit

    1) 4200 years ago the sample has 0% indicus
    2) 3800 years ago the first sample with indicus shows up.

    so the mixing happened between 4200 and 3800. they picked that date because it correlates with climatic change, but they don’t know when (i suspect it’s after 4000 years for various reasons i suggest above)

      1. i don’t care what dan says offhand in a newspaper article.

        the raw results are in the paper!!!! i gave you the date.

        yeah yeah could be trade. stop selectively interpreting papers tho. READ THE WHOLE PAPER

  8. The evidence for the Mitanni being closer to IA than Ir goes beyond the number one. Colour terms, the names of their rulers, and the deities names are closer to their IA counterparts than anything attested among the Iranian people.

    That is not to say it is certain, but the evidence we have now points to a stronger IA connection.


Leave a Reply

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

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