On the historicity of the Vedas

BPer Mukunda and I were having a discussion on Twitter, which I want to elevate and push to the blog, because it’s somewhat important.

When I was young (20th century) I read stuff about how the Indo-Aryans described the natives of the subcontinent as dark and “snub-nosed.” That their arrival in some ways was a meeting of two different races.

In the 2000’s I read other books and works that suggested that actually, these descriptions were metaphorical. Terms like “dark” in other words reflect an ideological or tribal conflict, with the descriptions pointing to tropes that signal which side is evil and which side is good. This is not a crazy view. The anthropology is clear that a certain level of fictitious dehumanization occurs with inter-group conflict.

So I accepted this view and moved on with my life.

But in the 2010’s things changed. I am now convinced that 3,000-4,000 years ago a people who resembled what we would term “white” expanded within the Indian subcontinent. If modern Armenians are white, then the Indo-Aryans were white. At least initially. In the subcontinent, they met a variety of people. Some of them, such as in Sindh, were of brownish complexion. Others, to the south and east, would have been considerably darker. I also assume that the Vedas were constructed in situ in the Indian subcontinent. That is, they reflect a milieu of people who were encountering the northwest of the subcontinent, and had recently traversed through BMAC (Indra may actually be a BMAC diety).

What’s the upshot here? I know think that the metaphorical view of the physical descriptions should be set next to the literal view. The reality is probably a mix. But the fact is that groups with very different physical appearances did interact in ancient India. The Aryans were almost certainly very light-skinned, with “sharp features”, in comparison to many of the people they encountered. Though one can construct hybrid scenarios, where Indo-Aryan enemies were described in inaccurate ways precisely because those tropes were associated with tribes and peoples the Indo-Aryans had conquered.

Someone who has deep knowledge of the Vedas in Sanskrit and genetics needs to look into this. That’s obviously not me.


54 Replies to “On the historicity of the Vedas”

  1. You got one part of it correctly when you stated that a person with deep knowledge of the “Vedas in Sanskrit” and genetics needs to dive in. Because the total set of the “white people meeting dark people” tropes originated with European translations. Sayana, for example, did not bother to talk about races or pigmentation in his 14th century commentary on the Vedas.

    There is one person who does meet your criteria – that is Manasataramgini. However that person (I know the identity) has been analysed threadbare by Koenraad and Talageri. His darshana of the Vedas is in the Shruti tradition. Talageri pointed this out back in 2013 – he seemed unable to assess the Vedas in a historical light and rather veered quickly to the linguistic (IE trad) angle.

  2. Commerntary on vedas within Hindu tradition is hoary and Sayana (https://en.wikipedia.org/wiki/Sayana) is the best known today. That is as saskritic as you get get Sayana does not mention different ‘races’ while doing commentaries. For many Englsih speaking commentators , word ‘krishnatvac’ where Indra defeats krishna only means Indra defeated black skinne dpeoples. But that is not the case within Sayana or other commentators themselves . For Sayana krishna was a demon mentioned in the vedas who was killed by Indra, no race connotation implied


    Western indologists invariably see only races , usually antagonistic races, in history. The sanskrit commentators never did it. Indologists project their own racial categorisations on any word

  3. May be useful here, a thread describing colors in the vedas:




    Text points out if we take it literally then many Vedic gods have golden eyes/red hair. Probably more so figuratively if we look at it in totality. Lots of sun/fire worship vibes and exaltation of associated colors.

    Also more *ritual*/effort to beget a dark skinned son (reminiscent of Ram/Krishna) than a light skinned one according to the Brihadaranyaka Upanishad –


  4. Because the total set of the “white people meeting dark people” tropes originated with European translations

    this could be right. what do i know?

    but it’s a bit curious that the light-skinned aryans didn’t comment on the physical differences when iranians and arabs later copiously did so btwn themselves and the natives of India who were dark and differently featured

    1. @Razib

      Some part of the Vedas indeed throw light on historical events but a large part is invocation by suggestion. For example, Time is “Kala”, but it is also a stand-in for black and Yama, the God of death is black-skinned. So Time -> Death -> Inevitability – all are underpinned by a single colour.

      I doubt whether the Vedic composers would have been so reflexively provincial and trippy at the same time. To have colours as proxies for philosophical concepts and skin pigmentation in the same frame.

  5. I would not attach much weight to the commentaries on the Vedas that were written many centuries after the composition of the Vedas themselves. Even the commentaries on the Koran that were written only a few centuries after the composition of the Koran have many inaccuracies. Do not see a reason why it would be much different in Hinduism.

  6. It’s entirely possible that early indo aryans looked like ‘Armenians’ as you put it and they spread into the northwest of the subcontinent. A minority of this population probably had more pronounced euro features like light hair and eyes like you see amongst kalash, pamiris and nuristanis. This population integrated with a local culture in the Punjab/Haryana region and then spread out again via subsequent empires like the mauryan empire.

    Where I disagree and see things differently is the nature of the relationship between the various parties. I think it was far more integrative rather than combative. This is where western scholarship sort of reads the tea leaves wrong in the vedas. They assume an outright racial conflict because that’s their history. I imagine the process in the subcontinent was more elite dominance and integration of elites into a new group. As you’ve pointed out there is a significant non-r1a component amongst Brahmins that suggest this group was an amalgamation.

    The major gods of Hinduism as it evolved are not Vedic gods and many of them including the major ones are clearly dark in complexion. A conquering force and imposed religion is much more clear about this – see blonde haired and blue eyed baby Jesus imagery all over Indian and Sri Lankan churches. Or the clear demarcation of Mohammed as from Arabia. What you see in Hinduism is the opposite.

  7. A lot of being centred on Sayana – but two things about him: a) He is writing from the Vijayanagara period and b) He sees the Vedas as externally ritual.

    The Vedas themselves have a strongly reenactive element to them, and in their actual ritual and recited form require the priest to effectively “be” and “act as” the “character” within certain hymn or ritual sections. To someone who understands the language, as an observer, the whole thing is a dramatic reenaction. The priest becomes a performer who is actually “acting out” the very origin of what he is doing.

    When we think about it in this sense, the whole racial element breaks down. The Panca-jana aren’t fighting outsiders, as much as they are fighting amongst themselves at set times. If we are looking at the “Vedic” period, in sensu stricto, then the attested tribes are no further than the headwaters of the Ganga, and they seem to migrate between the Himalayan foothills in summer to the edges of the plains in winter.

    There is little possibility of them interacting with a population that looks substantially different to them in this geography. Post-vedic & Pre-Janapada is a different matter, but with the tri-Vedic core you are dealing with outsiders on the Northwest, not the expanding frontiers to the South and East.

  8. “Western indologists invariably see only races , usually antagonistic races, in history.”

    While this has some merit, this is being blown out of the proportion. People need to throw their biases out of the window.

    It is very clear that there is some outright racism in the Rig Veda targeted at the higher ASI % population, call them Dasa or Dasyu.

    Brahmins should not feel bad about this, most of their DNA is from the native IVC people anyway (the Dasa).

    The earliest layers of RV seem to have these racist verses, but the later layers show a hybrid population, and we have darker skinned ‘arya’ rishis popping up.

    I recommend people read the following essay, if they cannot be bothered to read the Rig Veda:

    Religion, Skin Colour and Language: Arya and Non-Arya Identity in the Vedic Period (2019 edition)
    David N Lorenzen

    1. Yes but Karan you could easily make the opposite argument. It’s well understood that dasa refers to the dahae confederacy and that asura actually are gods from rival indo Iranian tribes. In fact Indra is a minor demon in Zoroastrian texts.

      Every other Indo European culture worshipped asuras (ahura in Iran and aesir in Europe) but Indo Aryans worshipped Devas and Indra.

      Whatever the integrative process, it wasn’t just a plain and simple racial invasion and war.

  9. Everybody seems to be talking about genetics. So here are some googlies:

    1. Bustan_BA, Dzharkutan1_BA, Sumbar, and Parkhai_LBA — Indo Aryan regions — have no Sintashta ancestry till about mid 2nd millennium and that too in one sample (I6667, Parkhai_LBA_o, 1497-1413 BCE)! They also don’t show R1a; I suppose you must be arguing for female mediated invasion starting from 15th century BC. Astonishingly, Dzharkutan is the region from which Zoroastrianism descended. How do you explain this?

    2. Kangju that are mentioned to have invaded India can be a source of Steppe ancestry. According to Razib, “admixture models” show earlier ancestry of the Steppe in India. This is perfectly consistent with “ancient DNA” that shows widespread Indian ancestry in BMAC starting from the 3rd millennium BC. The admixture between Steppe and Indian ancestry population happened around 14th century BC and thereafter not in India but in BMAC. This is what the “admixture models” are picking up. So, the mystery is now solved.

    3. Not a single R1a has been found in Middle East until the 15th century. Historical records show that Mitannis were present in the middle east since the 18th century BC.

    4. While Razib continues on with the debunked “color theory,” he seems to not wade into the historical and linguistic analysis done by Elst and Talageri. This is what we call selective pickings.

    5. Rig Veda is a Bronze age text. And iron age in India started in the 16th century.


    1. 1.Why exactly would these BMAC sites have Steppe Ancestry before the second millenium BC when BMAC WAS NOT ORIGINALLY EVEN ARYAN TO BEGIN WITH?

      2.So Kangju on invading,decided that they will mix only with higher castes and the higher castes said “Sure,Why not?”.

      3.You seem to conveniently ignore Alalakh Outlier.

      4.Even if you ignore colour,genetumics are starkly indicative and Talageri has to put genetics out of question in order to push his theory.

      5.So what

      1. 1. Do you know anything about Indo Aryans? BMAC is the only place the archaeologists agree to be indo aryan from 2nd millenium BC. I think you are unaware that the places that I linked are Indo-aryan; they also have zero steppe ancestry.
        2. You know Kangju were founding population of Kushans, and were also part of white Huns. Why not does not work as we have historical texts attesting the invasion. You know rajputs are said to have central asian descent. But let us ignore all evidences to fit a theory. Are rajputs not elites? Do you expect an invading population to become serfs and indentured labourers? Did mughals do so? Did British do so? Why would an invading population not become elite? Forget India, did it happen with Normans, Vikings, etc? Please explain. Furthermore, there have been Indo greek, persian invasions. There are multiple sources of Steppe ancestry, you know. And the proportion is just 15% of population as calc. by @Razib.
        3. Alalakh outlier happened to be a female sample in the late 15th century BC. Mitannis were present in Middle East since 18th century BC. No two ways about it. 300 years difference, you know. If you are arguing for a female mediated invasion starting 15th century BC, I am all ears. Also, go ahead and report an R1a sample from Mitanni controlled areas. That area must be teeming with R1as. Fun Fact: 0 R1a find until now.
        4. I suppose you are unaware that Iran_N/CHG ancestry is shared by all IE groups including Harappans.
        5. If Rig Veda is composed before 16th century BC, this implies its composition could only start from 19th century BC at the latest. Further, Mittanis were associated with late Rig Veda and yet they were present in Middle East since 18th century BC.

        If all this does not convince you, nothing will.

        1. 1.Ever heard about Vaksh Culture,Bishkent Culture and other Fedorovo sites of Andronovo Culture?
          2.So you mean the Z122 Subclade Kangju automatically fathered L-657 Indo-Aryan children?
          3.Ok my bad!
          I should have said Meggido Outlier.
          4.How much CHG/Iran_N do the Baltic and Slavic people have?
          5.Who says that Mittani are associated with later Rig Veda?

        2. 1. Are your arguments a joke? The accepted Indo Aryan sites in BMAC predate Steppe ancestry. You can quote 10, 000 cultures, they amount to nothing when the accepted Indo-Aryan sites don’t have Steppe ancestry.
          2. Meggido Outliers have R* group. No confirmed R1a. It is speculative. Also, they dated to 16th century BC. Even if I accept your unverified claim, still 200 years missing.
          3. I was quoting Talageri.

        3. The claim in Reich’s paper was that no Steppe group exists in Central Asia that can be a source of Steppe ancestry for India. There are groups that exist; mentioned in the same paper itself. And moreover, no L657+ has been detected ever in ancient samples. Who knows where they arrived from.

      2. To explain point 1 in simple terms for @AMS 212: Even without Steppe ancestry, BMAC is Indo-Aryan starting from 2nd millenium BC. Steppe ancestry appears only around 15th century BC. Furthermore, BMAC shared culture and belief as well as spoke a related language to Harapans according to archaeologists.

        1. Early BMAC has nothing resembling the Vedic Culture.
          Even the paternal lineages are like J,L and other Middle Eastern lineages.

          Because BMAC was only later occupied by Andronovo.

          Speak on Andronovo,the ORIGINAL INDO-IRANIANS.

  10. @Karan,

    The Dasas/Dahas survived into the historic age, so they are not some unknown ASI group. To the contrary, they were a Central Asian group that were attested until close to the common era.

    Applying colour terminology to the Rig Veda makes little sense as the core geography is basically a triangle from the Himalayan foothills near Chandigarh, to modern Jammu, and down into Multan. This was within the IVC range, so the likelihood of high AASI makes little sense.

    Also, there is a lot of polysemy in the Vedas, because there is a strongly reenactive character to the samhitas. When looking at its deixis (language of time, place, person) the colour terminology takes on a very different character.

    Let us just consider that as late as the 1920s you could describe someone as “glowing white”, “shining with cleanliness”, or “dark and grimy” without that person changing their race. Light and colour were strongly associated with cleanliness and grooming, in a way that we tend to associate with race today.

  11. “Someone who has deep knowledge of the Vedas in Sanskrit and genetics needs to look into this.”

    >>> This is exactly a framework for research which we accepted by acclamation, i.e. discussing genetics, linguistics, mythology and toponyms. Toponyms are maybe the most interesting part which none here wants to hear. This can be also discussed within the linguistics topic. Anthropology can be a good complementary topic, too. Right now, we are officially in a middle of linguistic discussion.

    It is funny that we still haven’t absolved the basics, e.g. the meaning of the names ‘rg’ and ‘veda’ (old Irish or Armenian? Pls give me a break). Recently, I provided a quasi-computer program to find out in which language they originated but so far pundits could not deliver. A half-dozen of red almost got frozen in my cellar. Back to square one – what does it mean ‘veda’ and from which language it came?

    Or, speaking about colors…It would be useful to know beforehand if for e.g. Yamnaya people (R1b) were white? This would make the discussion about colors at the time of Aryans migration more interesting and much simpler. But, it seems some are afraid to say anything about the color of Yamnaya people (ancestors of future westerners). Maybe this is just a straightforward thing but in this case someone can just say that this question is baseless or stupid. Is it?

  12. BPer Mukunda and I were

    Is Mukunda Raghavan there on BP? I only see a couple of podcasts, and would very much like to read if he has articles here.

    Aside: if any Indian classical music lover is around, check out Mukunda’s co-podcaster Krishna Parthasarathy’s youtube channel – nice combination of really traditional stuff and creativity.

  13. I doubt if the early indo-aryans resembled the Armenians

    I guess, they came in waves, of fortressed settlements across riverbanks, with associated cattle & copper tools

    If the multiple settlement waves is true, initially they would have resembled Kazakh/central asian & then later Afghan/Iranian

    Also Dasa can be anyone who is defeated by the victors (Aryans, noble ones). If we accept that winners get to write history, then Dasas could be the “aryans” who settled earlier. I believe some defeated dasa/aryans were chased away to Assam

  14. If you look at history of how a minority of Europeans overran India, it is because they were invited to take part in the partisan conflicts: the british by the bengalis & the french by the carnatic rulers.

    Something very likely could have happened after breakdown of IVC, the different Aryan settlers could have taken sides in internal conflicts.

    Since the winner writes the history, you don’t get the other side of the story

  15. Is it a given that Steppe people brought Vedic religion to India?
    If yes, how does one explain Vysyas?

    For Narasimhan et al. Vysyas sample was about 43 samples, with estimated admixture event between approximately 2600-2000 BCE and have about 5% Steppe (less than Dalit castes of same geographical region). We can argue that Vysyas are early IVC people with small founder population who came to Andhra region (Deccan).
    However, Vysyas are also strictly endogamous and are actually “dwijas”, i.e., fall into vedic varna status of merchant class (not just jati status).

    There can be two points to derive from this then:
    1. To be a vedic varna has nothing to do with steppe proportion, i.e., people can “adopt” vedic religion without having to have proper genetics.
    2. IVC had the vedic religion already before steppe people showed up.

    Despite all the talk about Brahmins and R1a proportions, Narasimhan et al. data set shows same proportion of (~30%) R1a in Brahmins (Brahmin_Vaidik 5 out of 15) and Dalits in Andhra (Yerukali – 6 out of 18). Granted these are very low sample sizes, but I don’t know if anyone else has a good size data sets of southern dalits to be confident about Brahmin-only spread of high-percentage R1a. (perhaps someone can point to that latest reference data? despite strong statements about Y-haplo in Silva et al. 2016, their n is pretty low for each group let alone representation of the enough number of deccan dalit groups).

    I am not so sure if historicity of vedas is even warranted without squaring these other details.

    1. Despite all the talk about Brahmins and R1a proportions, Narasimhan et al. data set shows same proportion of (~30%) R1a in Brahmins (Brahmin_Vaidik 5 out of 15) and Dalits in Andhra (Yerukali – 6 out of 18). Granted these are very low sample sizes, but I don’t know if anyone else has a good size data sets of southern dalits to be confident about Brahmin-only spread of high-percentage R1a. (perhaps someone can point to that latest reference data? despite strong statements about Y-haplo in Silva et al. 2016, their n is pretty low for each group let alone representation of the enough number of deccan dalit groups).

      it’s pretty clear that “high castes” have more R1a on the whole. but there are some “low caste” and “tribal” groups that have high R1a. part of this is probably drift, as these groups tend to have small effective population sizes

      the wikipedia on indian genetics collates y

      1. I looked at wikipedia entry but sample sizes are low if divided by caste, and not enough groups are sampled (e.g., just OBC+SC in AP are over 100 castes, “n” for all Andhra Tribals is 29, popular reference is Trivedi et al. 2006 that sampled about 1000 out of 80 groups, so the sample size per group isn’t any better than Narasimhan et al. 2019). I also suspect actually grouping of castes as “middle” and “upper” since the same caste is categorized as SC in UP and OBC in AP. (e.g., Kumhar vs. Kummari – Potters)

        Telugu Brahmins are about 5% of 100million population in Telugu states (not even split between Vaidik and Niyogis). SC populations of different castes are of comparable size (e.g., about 5 million just for Mala). Without samples any of those individual castes, it is hard to say they would all show “drift”.

        Brahmins_Karnataka have R1a at 2 out of 17, which Naidu (upper caste from Andhra) have 2 out of 18 (Narasimhan et al 2019). This indicates “Brahmin” lumping isn’t appropriate without considering sub-castes of Brahmins themselves. There is quite bit of “lower caste” lumping happening without distinguishing the actual jati endogamy.

          1. @Ugra,
            I looked at this while looking at wikipedia references. But notice that sample size per group is about 20 samples. Narasimhan et al. are using this data as referenced in Table s5, col K in their supplements.

    2. I thought this is a good place to get feedback about multiple hypothesis from Vysya scenarios. (Razib please feel free to not entertain this much speculation from me)

      Scenario 1: Vedic religion can have converts
      Hypothesis 1.1: Initial caste spread of steppe (or R1a) is homogeneous among all castes but for some reason R1a/steppe couldn’t thrive Indian infectious disease load without social-distancing/isolation (Brahmins and Chandalas are either volunteers or victims for this). Squares with North-South variation since disease load in tropical South is greater than North.

      Hypothesis 1.2: Brahmins were the most open at absorbing/adopting new comers to vedic religion and re-converts from no-jati Buddhism as long as they attended Nalanda/Taxila universities. Re-converts from Buddhism gets excessive representation in Brahmin castes and also most closely aligned with Buddhist principles (aside: I always wondered what happened to descendants to loads of Buddhists in upper deccan around 1CE, not all of them were monks if one looked at Ajanta/Ellora caves lol).

      Scenario 2: Vedic religion is known to IVC people
      Hypothesis 2.1: IVC was multi-cultural and multi-lingual (ref. Kenoyer). Vedics are people outside city-gates hating on “unicorn-seal” people for their customs and taxes to enter city-gates. When the “unicorn-seal” people fell in 1900BC, Vedics now have more free-trade opportunity between different IVC centers rather than suffering under uniformity imposed by “unicorn-seal” people from Harappa. They thrive and call their cousins from steppe to prosperity. They out-compete other cults of Pasupathi, Sakthi, tantrics etc, and absorb them over the time.
      In this case, timing of composition of Vedas has no relation to the fall of “unicorn-seal” people that triggered a population event and not necessarily language first introduction event (depending on where PIE homeland ends up and if this falls under IVC trade network periphery). Also squares with multiple waves arrival.

      In any case, it is a hopeless case to consider we will have representative aDNA samples of IVC people at all since Kenoyer says that number of burials vs. size and occupation length of Harappa indicates that IVC people mainly cremated.

      Any suggestions to falsify the hypothesis or steel man them are much appreciated!

      1. @Violet – a few attempts at knocking down your hypotheses

        Scenario 1 (both hypotheses) – The Rgveda Old Books do not have a reference to Varna or Jati. The first reference is in Book 8 and then Book 10. These are the late Rgveda. Old books are completely silent about any form of social structure. In fact, the old Books talk about Rajan (kings) without employing the term Kshatriya. No point in superimposing caste calculations onto them.

        Scenario 2 – Vedics were people outside the gates. This is rather implausible from a material perspective. The composers of the Vedas were poets and priests. Right through the old books, we see references to kings (samraj) and grhaspati (householders) inviting them to their lands/houses with material gifts. The Rgveda was an elite production, both from a content and style perspective. Jamison and Brereton make that very clear.

        1. @Ugra,

          1. Caste social structure is not needed in the vedas for both of these hypothesis. social structure can be an outcome of IVC people converting to vedics and keeping their jati. (e.g., how Indian muslims and christians have castes, or jain vanias).
          I am not commenting on presence or absence of varnas specifically in IVC or vedas, since varna is the standard IE social structure all over. Being “twice-born” is taken to be a primary ritual for being “Arya” (ref. Manusmriti). If IE people == Vedics, then people practicing “twice-born” ritual with low “steppe-genes” implies conversions of non-steppe to “Arya” status (i.e. to explain Vysyas).

          2. Good point on vedics having poets, priests, and kings. However, people can have all these in rural kingdoms and hate on urban people. Vedas have references on destroying cities but protecting cattle herds. Remember Viraata king of Mahabharata. He was a “king” with pandavas in his employ, but wealth is mainly dependent on cattle and horses not “trade” of crafted materials which was the lifeline of IVC (the so many seals are all for parcel tracking 🙂 ).
          Also curious fact that 2600-1900BC is estimated as the pre-dominance of “unicorn-seal” people. Look how the dates match the confidence interval on “steppe” admixture to Vysyas.
          (i.e., IF Vysyas were hating on “unicorn-seal” people and travelling to deccan to escape, assuming that Vysyas got their “twice-born” status from original steppes heavily admixed with local In-Pe pops but BEFORE the additional steppe population wave following the collapse of “unicorn-seal” people)

          1. @Violet

            It is not parsimonious. First of all Kenoyer is not looking into traditional Indian astronomical system (Jyotisha) to explain the unicorn seal. This is because it would imply that the IVC was IE. And due to this he misses out on a very valid explanation.

            Every year in March, the Sun climbs (apparent movement) into the Northern hemisphere. This is known as the Vernal Equinox or the Spring Equinox. When it does so, in the night sky there is a constellation in that position. Ok?

            Now because of precession, very slowly the night sky constellation changes over hundreds of years. Today in 2021, it is the Pisces constellation at the time of the Vernal Equinox.

            But between the years 4000 BC to 1700 BC, it was the constellation (Rishabha), the Bull or Taurus at the time of vernal Equinox.

            So the unicorn seal represents the constellation, not a real bull. And it was used to measure the agricultural produce brought in March – when harvest festivals peak in India.

            After 1700 BC, earth’s axis precessed and a different constellation took the place of Taurus. So the unicorn seal faded away.

            And in the Rgveda, the most frequent term used to describe Indra is Vrsabha (Rishabh or Bull). So we have a tri junction of thema – that the unicorn seal represents the vernal Equinox in March, it is also the constellation in the night sky and also refers to Indra.

          2. @ugra,

            Given what we learned about prehistory so far in searching for PIE homeland, we can discard parsimony and occam’s razor as tools for the hypothesis testing as long as we can find evidence for all the bits in the evidence-chain.

            I know about precession of equinoxes and the movement of “ages” between taurus, aries, pisces (and more recently people going on about coming of aquarius age). If I wanted to know more about archeo-astronomy there are a whole bunch of people from Nilesh Oak to Manish Pandit going on into “date of Mahabharata”. We don’t need Kenoyer for that. He is good at what he does.

            Kenoyer made no claims about anything but simply provided data on what are the different types of seals, when they showed up in the archeological records, and how frequent they are. Unicorn seal is exactly that – “unicorn”. There are bull seals that are distinctly bulls (with hump and two horns). It’s not like IVC guys didn’t know how to carve a bull and accidentally carved an unicorn instead. There are even tiger seals, and elephant seals (which had no constellations assigned), and before you go on, tigers are different from lions and we can’t argue tiger seals are showing leo.

    3. Brahmins just group castes into different varnas based on the occupation.. When they change the occupation their varna change.. So, just stick with sub castes and dont confuse it with varnas…

  16. I read a claim on OIT Twitter that the haplogroups R1a-M780+ and R1a-L657+ haven’t been detected in aDNA samples anywhere across the IE areas. Is this true?

    For the most contrarian view on IE prehistory, I recommend the work of Robin Bradley Kar published in the University of Illinois Law Review in 2012. He breaks up OIT into three distinct possibilities – Out of Eastern Iran, Out of Indus Valley and Out of IVC/Bactria/E.Iran. It is the most scholarly piece of work on OIT I have ever read. It’s a 250 page work – close on to a novel.

  17. Sintashta is around 52°N, North bank of the Caspian Sea is ~47°N, Ferghana is ~40°N, Swat Valley is ~35°N, Punjab is ~31°N, Rakhigarhi ~ 29°N, Mohenjo-daro at ~27°N, Sindh is around ~26°N, Dholavira site is ~24°N, Lothal ~ 22°N.

    Isn’t this sort of self-obvious though that a population group having 1000s of years of continued existence at a certain latitude of the planet is going to develop certain melanin type body response?

    The widest range is ~30° From Sintashta to Lothal, or the average to NW India can be termed half that at the gross minimal extreme. Is there a place/region in the world where there is little to negligible skin pigmentation differentiation over that much of Latitudinal range?

    Maybe there wasn’t much difference (arguendo for this Indo-Aryan example) but it would be an outlier of monumental proportions and such an extreme outlier then must have pretty stark evidence somewhere eventually, esp. in genetics since if the debate is around skin/eye/hair color genetics is the pre-dominant field one should be looking for answers, which in the general domain at large (AIT, etc etc) there is a bit more balance with archelogy, genetics, human sociology, etc all having a significant say.

    1. Isn’t this sort of self-obvious though that a population group having 1000s of years of continued existence at a certain latitude of the planet is going to develop certain melanin type body response?

      The time scales for this would have to be much larger than those required to explain IE-speakers’ peregrinations, no? Think of the Inuits: from what I can see, their skin color is significantly darker than Northern Europeans (even most Southern Europeans?) despite them inhabiting Scandinavian (or even further north) latitudes. Their skin color seems to reflect their Asian origin, despite the fact that they migrated ~12k years ago.

      Are there proven records of light-skinned people spontaneously darkening after migrating south, or dark-skinned people spontaneously lightening after migrating north? I’m quite skeptical of evolutionary explanations for changes happening since the Holocene; I believe cultural preferences/selection, migrations, and population-mixing are necessary and sufficient.

  18. vedas are religious text and reflect on what matters to the believers, since it was an oral tradition for long, it might not reveal other less important aspects. It would be like expecting quran or bible go into details about color of skin etc. Religious books are about beliefs and perhaps might not give a complete view of social commentary and attitudes that people might have but are not important in terms of belief to be worth recording?.

    1. The development of clear genre divides between secular non-fictional historical accounts, mythology, and religious writing is a recent one. In most of the world, Bronze Age and early Iron Age storytelling (and some Medieval storytelling as well) is “legendary history” that blends all three in accounts that are memorable, that has some touchstones in historical reality, and have some aspects that are religious and/or mythical, or added simply to make the story better even if it isn’t what happened and doesn’t have a strong religious lesson to tell.

      There is really no way within the four corners of ancient texts like the Vedas to parse these components out from each other. But archaeology, genetics, linguistics, analogies from similar events documented in a more factual way, and other tools can help inform our understandings of which components are which. On the whole, we are in a phase where we are coming to appreciate that there are historical components to these legendary histories, while also learning to take them with a grain of salt by recognizing that they aren’t comparable to modern non-fiction historical accounts. Recognizing that a lots of ancient encounters involved visible racially distinct groups (in the layman’s sense of the word, and sometimes between groups, neither of which exist in pure form any longer) can help us understand one more dimension to those encounters.

  19. Do u think that they looked similar to Steppe_MLBA(75%Brown/25% blue eyed, Tall, Fair skin, Northen European faces, 79% Brown haired/21% blond haired) ? Like Fatyanovo-andronovo-sintas-Abashevo ?

  20. some of you have low IQs. the reason i don’t address/talk as much about nongenetic stuff is I DON’T KNOW IT WELL.

    if i said “these archaeologsits don’t talk about genetics, they are so biased!” you would probably say “what a moron.”


  21. Religious books are about beliefs

    this is false. religious books are often mostly about ritual.

    please say true things if you want to be taken credibly on other issues

    1. i said it is about what matters to believers first and only then did i say about belief. And i am agnostic about this issue. I always try to say things that are true. I might be mistaken, but do not lie.

      1. qs in general.is it wrong to use “beliefs” here in sentence that “religious books are about beliefs” if i began with vedas being religious text and record ” what matters to believers”. ?

  22. I don’t have any strong opinions on this matter but I appreciate Razib willing to go against the grain even if he understands he will get pushback from his readers

    Any author willing to alienate his own audience in the pursuit of truth is worthy of respect. It shows that intellectual rigor is the highest ideal, even if at the cost of monetary gain and social approval.

  23. To close the loop on sample size issues with R1a frequencies, here is some data collated from
    Moorjani lab in 2017 (“pop estimate”): https://www.biorxiv.org/content/10.1101/047035v2
    Narasimhan et al. 2019 (“y-haplo-n”, in conjunction with Sahoo et al. 2006, and “admixture-n”) and Trivedi et al. 2007 (“caste-label” – the main source of R1a categorization in South Indian castes, and co-authors of Sahoo lab)

    group name pop estimate y-haplo-n admixture-n caste-label
    Brahmin_AP 2E+06 15 NA upper
    Brahmin_Vaidik 3E+06 15 27 NA
    Chenchu 7E+04 20 NA tribe
    Koya 6E+05 27 9 NA
    Lambadi 2E+06 18 5 tribe
    Naidu 2E+07 18 8 lower
    Naidu 4E+06 18 8 lower
    Reddy_Telangana 2E+07 12 8 middle
    Yerukali 5E+05 18 7 NA

    1. some population estimates are suspect since combined AP and Telengana populations are 93 million, but pop estimate totals exceed 100 million in supplements. So, clearly some populations are double counted or overestimated (see the discrepancy in “Naidu” given with two different pops with order of magnitude difference).

    2. Reddys in total are estimated at 8% of combined telugu population (this was a huge factor in how AP was split – about 7-8 million is a reasonable estimate). So, estimation of 20 million for just Telangana (with 39 million total pop) is way overestimate. (they also estimate about 12 million Shia_Iranian_Hyderabad in Telangana, you would think most Telangana is just reddys and iranians – to be clear I am grateful for open access by authors, but these type of details are easy enough for a lay person to check and evaluate evidence accordingly)

    3. Narasimhan et al. maps “kapu naidu” of sahoo et al. to just “Naidu”, but they are labelled as “lower” caste by Sahoo et al (I kept same number for both pops of “Naidu” from Moorjani lab). This is obviously wrong. Trivedi et al. do it consistently with labelling also “kamma” as lower caste and “reddy” as middle caste. How can anyone trust the narrative about “upper”, “middle” and “lower” castes, as reddys have 0 out of 12 R1a and naidu at 2 out of 18?

    4. I am stuck by obvious very low power of R1a deductions not just in AP samples but brahmin sample size for UP, Bihar, Maharastra, West Bengal and Madhya Pradesh. All of these states have population from 200 million to about 100 million. If we conservatively estimate Brahmins as 5%, that is about 5 to 10 million population per state. But max y-haplo-n from these groups is about 40-50 to distinguish proportions between 10-15 different y-haplo groups.

    If we are going to run with this narrative with not even 0.1% of total group random sampling (i.e., 0.1% of 5 million = 5k) for frequency estimate over about 10 different y-haplo groups the confidence intervals over these would be so wide as to be meaningless. Perhaps we don’t need 5k since it can be with random sampling of only independent households, and not individuals. But entire “North” sample size of unrelated individuals is 180 in Trivedi et al. (2007).

    It is reasonable in day-to-day stats to merge samples for larger sample size and improve confidence. But are Brahmins in UP really mergeable with ones from AP, Maharastra and West Bengal? Also, caste labelling is so suspect (naidus are one of the two ruling castes of AP, along with reddys, and are labelled as “lower”), I don’t know if any of it can be taken on high confidence.

    My purpose to posting them here is to see if anyone from UP, Maharastra and West Bengal can post their data review of caste status assignment and relative population sizes in their states. Perhaps most of these authors being North Indians, they did a better job for those states?

    My purpose for picking Telugus is it is the most populous dravidian state pre-split. If we picked geo-variance among the top 5 populous Indian states (north, east, west) at least we get bit unbiased understanding of “overall” trends instead of relative frequency inference based on sampling bias.

  24. Im following BP From quite a time .I think it(genetics) is an amazing approach in understanding South Asia demography.I thank everyone Who is commited to bring truth to Us with having no personal agenda.
    Q-Do we have any sample with pure steepe ancestry in Subcontinent?,As I think we might have Pure AASI.

  25. So many speculations and corrupted taxonomy. Too much focus on genetics disconnected from linguistics and others. Why not having a holistic view, as we agreed above, and crosscheck the facts on the spot?

    Again, let’s explain the meaning of ‘Veda’, from which language it came, is still in use there (this was already done but without mentioning the name of the language). Similar for ‘rg’ (for this we haven’t done anything so far). This single and simple clarification will uncover so many things (for example who were Aryans) similar when you press the right button in ‘minesweepers’ which causes an avalanche and supersedes so many unnecessary dilemmas.


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.