How much “steppe” ancestry is there in South Asia? (Indian subcontinent)

By Razib Khan 109 Comments

Since this question always comes up at some point, I decided to do a rough back-of-the-envelope calculation of the % steppe across the Indian subcontinent. The way I did it was by taking Pakistan, Bangladesh, and India, and estimating the average percentage from the caste breakdowns (e.g., UP is 20% “upper caste” and 20% “Dalit” and 60% neither, with fractions of steppe/Sintashta about 30%, 10%, and 15%, respectively).

So the final number I came back is that 14% of the ancestry in modern-day South Asia is from the steppe in the form of people descended from Sintashta pastoralists. That is about 220 million human beings worth. You can judge whether that’s significant or not. Additionally, it looks like closer to 20-25% of the Y chromosomes are derived from these people.

I’m not “showing my work” because I think no matter how you estimate it, you’ll get a number in this range. Perhaps 12%. Perhaps 16%. But what difference does that make?

7+

109 Replies to “How much “steppe” ancestry is there in South Asia? (Indian subcontinent)”

  1. It sounds good to me. Based on my sources, two years I mentioned 16% on average. It is above the threshold for pulling heads out of sand. Many significant conclusions can be made from this. I may later put some specific perspectives on this.

    0
    1. I am not getting the answer, so there is another one or two – which haplogroups “steppe” brought to SA and were the ‘steppe’ – white?

      0
    2. For liars like @Milan Todorovic:

      1. Haplogroup Y-DNA I % does not matter as Hittites — with zero steppe lineage — have J2 haplogroup and derive descent from Iran_N/CHG ancestry.
      2. Mycenaeans were Greek speakers and have been present in Greece at least from the 17th century BC.
      3. Also, Vincans — primarily having Anatolian descent — have the following Ydna haplogroups:
      G2a2a1, G2a2a1a, G2a2a1a2a, G2a2a1a, G2a2b2a1a, H2

      There is no I haplogroup in PIE. Your duplicitousness to prove Serbia as PIE homeland by using I haplogroup and Vinca does not work.

      8+
  2. I get these same numbers using the Eurogenes G25 PCA. What method did you use to get Sintashta scores for South Asia?

    1+
  3. “Additionally, it looks like closer to 20-25% of the Y chromosomes are derived from these people.”
    Isn’t R1a more than 50% of all South Asian Y chromosomes? It will even increase in future because Northern South Asians have higher birth rate.

    The lesser Autosomal steppe DNA, the better it is.

    How much autosomal steppe-derived ancestry do Bangladeshis and Biharis have?

    0
  4. I have a couple of outside questions –

    1. What is the fit between Steppes aDNA and modern Iranian DNA?

    2. Based on the 12 IVC periphery cline individuals and the Rakhigarhi woman, what is the contribution of the ancient Indus farmer to modern day populations of Kazakhstan, Uzbekistan and Turkmenistan?

    0
    1. 1. What is the fit between Steppes aDNA and modern Iranian DNA?

      lots inf central asia (tajik) and eastern iran. lot less in the west. i’ve done a blog post on it, do search

      Could you possible do a breakdown by region?

      just use search. i’ve posted so many qpadmins

      i estimate ~20% in pakistan, ~15% in the north, ~5% or less far south

      1+
  5. Makes sense.

    I am Y H and mt K1a

    My own ratios are like 20% steppe, 45% iranic related, and 35% aasi.

    Groups like Rors might be 40% steppe, 45% iranic, 15% aasi.

    The most non steppe shifted panniyas would be like 70% aasi 30% iranic related 0% steppe

    Average Indian would be like 12.5% steppe 45% iranic and 42.5% aasi

    As one of the protagonists of the Brownpundits Mahabharata:
    https://imgur.com/a/aupLKX2

    May the Junglefather give you strength
    May the Rivermother bless you
    May the Skyfather smile upon you

    0
      1. Can you run them? I haven’t really gone through much. I am estimating my percentages off S Indian Brahmin numbers from Narsimhan, since I cluster with them on GED match calcators

        PC1,PC2,PC3,PC4,PC5,PC6,PC7,PC8,PC9,PC10,PC11,PC12,PC13,PC14,PC15,PC16,PC17,PC18,PC19,PC20,PC21,PC22,PC23,PC24,PC25
        Warlock_scaled,0.047806,-0.045699,-0.143306,0.112405,-0.080938,0.06275,-0.00047,0.007846,0.029656,0.018953,-0.001949,0.005395,-0.000149,-0.008533,-0.004479,0.008353,0.018906,-0.001394,0.004902,-0.003252,0.000998,-0.000989,0.007765,0.001807,-0.003353

        0
          1. Thanks. Feel free to run it through diff stuff. Curious how I come out and which groups I cluster with, especially other Jain samples

            0
          2. 1. IVCp is a good source for Iran HG ancestry for south Asians because using Hotu as well as IVCP results in 0% Hotu. There will be some south Asians who will need shahr ba1 or maybe even BMAC but that isn’t the case with warlock. I’ve tried using BMAC and Iran HG and it makes no difference. Forcing them makes a worse model. The reason for this is because for warlock pretty much all of his Iran HG ancestry is routed through shahr ba2.

            2. Simulated aasi may not be perfect especially not NW aasi. But southern aasi was made such that it being combined with the Iran component (along with othets) results in paniya and other high aasi southern samples. Sure this isn’t 100% accurate and sure the aasi ancestors of paniya aren’t going to be the same as warlock’s aasi ancestors but it is still a much better choice than a group that split tens of thousands of years ago and has bottlenecked like crazy since then. Why do you want quantifiably worse results? I can try Andamanese later on today, as can you and I will post the results as well.

            3. Yes and those results are less accurate. See how far those results are from the target of what they are being modeled as. Try using just anatolian and papuan next to get even worse results.

            4. The only later ‘near eastern’ ancestry in IVCp is anatolian and that is present in shahr ba1 (warlock’s model did not pick this up at all and neither do regular gujaratis) but pretty much absent in shahr ba2. Also none in the old ba2 and old BA3. If there was a more recent input after the mesolithic Iran admixture then it would be a neolithic iranian admixture prior to any anatolian influx and all of that is accounted in warlock’s shahr ba2 ancestry.

            0
        1. You didn’t pick up NW AASI or Shahr BA1. I did a run on Gujaratis and Gujarati Brahmins sometime ago, however I didn’t save it, but I can recall that both overwhelmingly preferred S AASI over NW AASI. Also none of the regular Gujaratis had any Shahr BA1 (Brahmins did iirc). So without the steppe ancestry, it is Shahr BA2 + S AASI.

          https://imgur.com/a/HnL90SD

          NOTE: Shahr BA2 refers to the average of the following samples: I11456, I11459, I11466, I8726, I8728.

          0
          1. Thanks man. Can you roughly translate this into Narsimhan paradigm? Looks like I fit best with Tambrahms again just like Harrapa? Just curious about your interpretation of my results, in the context of my background (Guju Jain Vania-paternal from Gujarat proper and maternal from Saurast)

            0
          2. @INDTHINGS
            Those are more distant sources, it is better to use sources that are temporally closer to the target if you want a more accurate result. Hypothetically, someone could undertake the project of making a time-map of the contribution and continuity from paleolithic to mesolithic to neolithic to post neolithic to modern populations, but currently there isn’t enough information to fill all of it. We could make such a map for Europeans from the mesolithic to the present, but in south Asia, the oldest DNA is from the bronze age only. This year, a lot of east Asian DNA has been published, so in their case a map can already be made from early neolithic (or even very very late paleolithic/mesolithic) to the present. Unfortunately, the rate at which south Asian DNA is published is very slow.

            I have tried to model Shahr BA sources using older ones but the fits aren’t very good because the inputs are only so-so approximations of the actual ancestors of the InPe group and by extension the majority of south Asian ancestry. The lack of an actual AASI sample doesn’t help. Here is my attempt from some time ago: https://imgur.com/a/Gtu9IfT
            The fit for BA2 is equivalent to ~4.58 on other global25 calculators, a good one would be less than 3 and a really good one would be between 0 and 2.

            0
          3. @thewarlock

            So comparing it to Narasimhan results: Shahr BA2 is InPe. Sintashta + Tyumen is steppe. S_AASI_SimAv is a simulated southern AASI component, akin to Narasimhan’s native south Asian hunter gatherer group.

            The shortest distances to you are Tamil Brahmins and regular Gujaratis. Though your distance to Velamas is as much as your distance to UP Brahmins, and more than your distance to West Bengal Brahmins.

            0
          4. So would I be roughly 16% steppe, 34% AASI (let’s make 25% of IpE as AASI), and 50% iranic related?

            So essentially, similarly west eurasian: east eurasian as tam brahms approximately but bit more iranic related shifted within the west eurasian?

            Intermediate distance between valemas and UP bramins is fairly congruent with caste positioning as well. Do my conclusions make sense? I realize I am oversimplifying

            0
          5. Dathang,

            Using more proximal sources gives you better fit values but can ruin the admix results.

            Modeling South Asians as IVC-Steppe-AASI leaves out Neolithic related Near-East ancestry and puts it in Steppe, which gives really weird results for some Indus Valley populations (Sindhis getting 22% Steppe, Brahui getting 27%). Breaking it out into Hotu (Iran hg), and Barcin (neolithic near east) seems to fix that.

            Hotu seems to be a pretty decent proxy for the basal Iran-related ancestry in India per the Rakigarhi paper, so it just makes sense to use that rather than trying to approximate it haphazardly with IVC. Andamese as a proxy for AASI also seems better than trying to model AASI ourselves as it cuts cleanly away from Iran HG much better. Using Andamese is only an issue when trying to tease out Austronesian (and maybe SE Asian) admixture from pure AASI when looking at tribal or certain bengali pops.

            1+
          6. @INDTHINGS

            You are right about possibly missing recent admixture from places not covered like Natufian type of ancestry from Arabs or the more recent Iranian type from Iranians.
            But you are wrong about using Andamanese and using Hotu to not “approximate it haphazardly with IVC”. If you are worried about missing BMAC, then that is a legitimate concern. However in this case, he wasn’t even picking up Shahr BA1, so BMAC is a long shot. I have tried another run with Gonur and it results in no difference. The Shahr source is supposed to approximate all of the Iranian ancestry that he gets through south Asia. Using an extra Hotu source is going to be unnecessary since Shahr BA2 already descends from a Hotu-like source among others, and its not like there is an extra additional Iran mesolithic pulse unlike AASI. And just to confirm, I have used Hotu as an input and it makes no difference at all. Hotu is already accounted in Shahr BA2. Trying to force in Hotu by eliminating all Shahr and using Hotu as the sole Iranian input worsens the fit to a 4.5 because now you are using the equivalent of a ‘low resolution’ source (Hotu + now even more AASI) so to speak to replace the more accurate Shahr BA2 input. Modeling yourself as your parents will be better than modeling yourself as your great-great grandparents.

            Using Andamanese instead of simulated AASI results in worse fits and this is because the simulated S AASI is based on high AASI groups like Paniya. NW AASI is based on NW Indian groups and it doesn’t usually work well outside of a few targets. Simulated southern AASI isn’t real but it is made from an actual south Asian population as opposed to a distant southeast Asian cousin.

            0
          7. @thewarlock
            It is hard to split the ancestry of Shahr BA2 in well defined proportions as of now, but if we insisted, then it would be about 2/3rd Iran related and around 1/3rd AASI (based on my old but not so reliable run that I have posted earlier in the second imgur link). Since you end up with 2/3rd of this in total, that is (2/3)*(2/3) = 44.44% Iran related; (2/3)*(1/3) + 17.4 = 39.62% AASI and the rest is steppe. The AASI in InPe might be 25% in total, but your InPe is the more AASI heavy Shahr BA2. Your estimated aren’t far off from mine however, so I don’t think I will propose a well defined alternative, especially since the proper composition of Shahr BA2 with a good fit isn’t established yet. For it to be more well defined, we would need DNA from India’s mesolithic and neolithic.

            0
          8. DaThang,

            1.) IVC-p (Shahr BA2) is not a good proxy for Iran Hg because its only about 66% Iran Hg. The less Iran Hg the reference sample has, the more false signal is being kicked into steppe to compensate. You can test it out right now using different IVC-p samples as your source for Iran Hg. Shahr BA 3 has only about 50% Iran Hg, and gives significantly higher Steppe values than Shahr BA2.

            2.) Stop using simulated AASI they are just splitting the Iran HG signal (or splitting themselves into Iran HG).

            3.) Using Hotu+Andamese rather than IVCp+Sim AASI gives much different results. Warlock’s Steppe goes from 16% on your run to 6%. Warlock before you start emoting, this makes everyone’s steppe go down.

            4.) I disagree on this but per Araingang the Near-East ancestry in the Indus (excluding Iran HG) arrived before the Aryans, and there’s been very little input since.

            0
          9. @INDTHINGS
            1. IVCp is a good source for Iran HG ancestry for south Asians because using Hotu as well as IVCP results in 0% Hotu. There will be some south Asians who will need shahr ba1 or maybe even BMAC but that isn’t the case with warlock. I’ve tried using BMAC and Iran HG and it makes no difference. Forcing them makes a worse model. The reason for this is because for warlock pretty much all of his Iran HG ancestry is routed through shahr ba2.

            2. Simulated aasi may not be perfect especially not NW aasi. But southern aasi was made such that it being combined with the Iran component (along with othets) results in paniya and other high aasi southern samples. Sure this isn’t 100% accurate and sure the aasi ancestors of paniya aren’t going to be the same as warlock’s aasi ancestors but it is still a much better choice than a group that split tens of thousands of years ago and has bottlenecked like crazy since then. Why do you want quantifiably worse results? I can try Andamanese later on today, as can you and I will post the results as well.

            3. Yes and those results are less accurate. See how far those results are from the target of what they are being modeled as. Try using just anatolian and papuan next to get even worse results.

            4. The only later ‘near eastern’ ancestry in IVCp is anatolian and that is present in shahr ba1 (warlock’s model did not pick this up at all and neither do regular gujaratis) but pretty much absent in shahr ba2. Also none in the old ba2 and old BA3. If there was a more recent input after the mesolithic Iran admixture then it would be a neolithic iranian admixture prior to any anatolian influx and all of that is accounted in warlock’s shahr ba2 ancestry.

            1+
          10. Okay I have tried modelling warlock using the populations INDTHINGS mentioned in some other post, and here it is:

            Target,Distance,IRN_HotuIIIb_Meso,Onge,RUS_Sintashta_MLBA
            Warlock_scaled,0.11339447,50.6,42.2,7.2
            Average,0.11339447,50.6,42.2,7.2

            And WOW it sucks, that distance is ~11.34, yup, a good reminder not use a model for south Asians using Onge.

            Even replacing S_AASI with Onge in the original warlock model is bad:

            Target,Distance,IRN_Shahr_I_Sokhta_BA2,Onge,RUS_Sintashta_MLBA,RUS_Tyumen_HG
            Warlock_scaled,0.04600810,79.2,12.0,5.8,3.0
            Average,0.04600810,79.2,12.0,5.8,3.0

            Target,Distance,IRN_Shahr_I_Sokhta_BA2,Onge,RUS_Sintashta_MLBA
            Warlock_scaled,0.04662308,79.2,12.8,8.0
            Average,0.04662308,79.2,12.8,8.0

            0
          11. I have tried to post the alternate models using INDTHING’s suggested groups, but akismet did not publishing it. Here is a summary: the fits were bad and it wasn’t surprising. Not extra Hotu when Shahr BA1 and BA2 are already there, at least for warlock’s coordinates. Might be one or two percent for extreme NW Indians who are on a far extreme of the south Asian distribution.

            EDIT: Here is the test for extra Hotu in the extreme populations:
            Target,Distance,IRN_HotuIIIb_Meso,IRN_Shahr_I_Sokhta_BA1,IRN_Shahr_I_Sokhta_BA2,NW_AASI_Sim_modern_Av,RUS_Sintashta_MLBA,RUS_Tyumen_HG,S_AASI_SimAv
            Ror,0.00983453,0.0,7.2,43.6,6.6,35.6,3.6,3.4
            Jatt_Pathak,0.01788128,1.6,10.2,36.0,7.0,35.0,4.8,5.4
            Kalash,0.01803924,0.0,38.4,21.8,7.6,22.6,6.4,3.2
            Sindhi,0.01207524,0.6,29.6,35.0,6.8,19.0,0.0,9.0
            Average,0.01445757,0.5,21.4,34.1,7.0,28.0,3.7,5.3

            With Gonur as well:
            Target,Distance,IRN_HotuIIIb_Meso,IRN_Shahr_I_Sokhta_BA1,IRN_Shahr_I_Sokhta_BA2,NW_AASI_Sim_modern_Av,RUS_Sintashta_MLBA,RUS_Tyumen_HG,S_AASI_SimAv,TKM_Gonur1_BA
            Ror,0.00982906,0.0,4.6,43.6,6.2,35.6,3.6,3.6,2.8
            Jatt_Pathak,0.01788128,1.6,10.2,36.0,7.0,35.0,4.8,5.4,0.0
            Kalash,0.01761336,0.0,15.8,25.8,6.8,20.8,6.8,2.8,21.2
            Sindhi,0.01203462,1.0,21.0,36.8,6.6,18.4,0.0,8.8,7.4
            Average,0.01433958,0.7,12.9,35.6,6.7,27.4,3.8,5.1,7.8

            Only 1% or 2% extra Hotu in the extremes.

            0
          12. I don’t remember what the barcin, andamanese etc, fits were but they were like 11+ for warlock and 6+ for regular population averages. Pretty bad on both accounts.

            0
    1. warlock:
      Rocking that blue sherwani 🙂
      BTW, the other guy in the photo reminds me about a pet peeve: Gujjus don’t do headgear well at all. It looks manufactured, tacky even 🙂 Don’t mind but you guys can learn a bit from your Marwari/Rajasthani cousins just up north on this one. Don’t believe me then look up Marwari safa 🙂

      2+
  6. (Part – 1)

    I have been thinking and researching for a historically attested people/class of people that could fit the Steppes DNA influx into India. First the background, the rationale and my assumptions.

    A section of Indians (the composers and bards) historically have been virtuously recording every anthropological and geophysical events that have impacted Bharat in the last 5 millennia. These remembered anthologies (Smrti like the Puranas), oral histories (Shruti-Vedas) and historical-kernels-with-fiction (Itihasas – MB/Rmyna) have been preserved and handed down generations with a fidelity unparalleled in the Indo-European world. The entire historical literature of some IE branches will run to barely less than a tenth of the Mahabharata!

    It is also true that in the last millennia Indians have been rather lackadaisical to their own history. Perhaps it is the impact of two consecutive imperialisms – one Asian and the other European – both of which uprooted and impoverished the traditional book-keepers of histories. Physical destruction also extinguished accumulated knowledge to an extent that is unimaginable. Nalanda’s books provided enough fodder for a fourteen day bonfire – so humungous were they. One shudders to think what historical data was lost!

    The etymology of Bihar is a corruption of the word Vihara. In early Sanskrit and Pali, Vihara refers to the act of classification and distribution – a grand human library. Nothing – almost nothing survived. Even the treatise (Arthashastra) of the greatest Magadhan administrator, Chanakya was found in Mysore, Karnataka far away from the ravages that extinguished the intellectual lights of North India.

    I firmly believe that ancient Indians also recorded the arrival of the Steppes people into Indian society and their integration there-after. The following hypothesis is sited within the autochthonous framework namely that IVC was IE-speaking, the successors of the IVC were a series of smaller urban agglomerations, that they coalesced into the Magadhan super-empires almost a millennium after the last Harappan city died out.

    1+
  7. (Part – 2)

    One of the enduring mysteries of Indian history writing is that of the identity of the Yonas or Yavanas. They are mentioned in the Puranas, Mahabharata, almost all Buddhist lores (Mahavamsa included), works of linguists (Panini) and astronomers (Aryabhata). After the works of William Jones and Prinsep, it was universally agreed that they represented the Greeks and Seleucids – the word being the equivalent of Ionians. The edict of Ashoka seemed to confirm this (Yavana kings).

    For some time, this conjecture held until the word started to surface in other works such as Panini who is conclusively dated between 400-700 BC or thereabouts. Of course there were no Greeks in India during this period. And of course, at this point the Buddha was not yet conclusively dated.

    The Buddha makes a very specific point about the Yonas and Kambojas (which will become important to the genetics later). In the Assalyana Sutta, the Buddha tells his disciple that the Yonas have only two classes – Masters and Slaves. “Why cannot we be like them?” – a very obvious reference to the Indian caste system. At the time of the Buddha, there were absolutely no Greeks in the Indian subcontinent or in its Northwest.

    So ancient Indians indeed referred to a class of people as Yonas even before the Indo-Greek interactions. How did these Yonas behave? Lets look into the sources.

    Gautama Dharmasutra says this in the context of Yonas – Kshatriyas who do not follow the Dharma become the Yonas. In another place, it explicitly states that a Yona is the offspring of a Kshatriya father and a Shudra mother.

    We can infer several things by now – ancient Indians have classed a set of people as Yonas much before the Greeks arrive on the Indian scene. One of them (Gautama Buddha) admires their lack of caste, another traditional Hindu treatise castigates them for the same – explicitly stating that they do not adhere to the codification in traditional Indian society and also refers them to be a product of two distinct varnas.

    1+
  8. (The final Part – 3)

    Now in this final part we will see the etymology and the roots of the words Yona, Yavana etc. Sanskrit is a fantastic language – the roots of the words formed themselves throw light on what they are describing.

    There are two separate explanations.

    Yauti mishrayati vaa mishriibhavati sarvattra jaatibhedaabhaavaat iti yavanah – meaning that the stem Yu implies mixing or mixed people and hence the word Yavana for mixed peoples.

    In Vedic texts, Yu is used to refer to the act of moving quickly. The word Yavan actually means a man riding a horse!! Also it is used to referred to velocity or speed. The suffix-yavan is used to form compound words with the intent of expressing motion or movement or act.

    The Majjhima Nikaya (Buddhist scripture) expresses surprise that the Yavanas call themselves as Aryas. Here the circle is complete – in the past the Buddha admires the Yavanas for lacking caste. Hindu scriptures castigate the Yavanas for mixing. And finally the Majjhima Nikaya (3rd BCE to 2nd AD) is surprised that the Yavanas are now fully Aryanised.

    In so many ways has the Steppes influx been fully recorded by the ancients in myriad ways – on their social structure, their lack of adherence to the Arya ways and finally a kind of integration.

    The Yavanas, Yonas were the Steppes people who came to India. It fully explains the genetic data in more ways than one. It also provides a secure terminus post quem for the mixing and the reaction it triggered in Ancient India.

    I have a post for the sociological impact of the Yonas which might have hardened caste structures and endogamy. That’s for later.

    0
  9. I wrote briefly before about Yavans (in correspondence with Anan) and explained that Yavans were not Greeks than Serbs. The above comment indirectly confirms because Mahabharata mentioned them at the time when Greeks did not exist (and btw Greeks never had kings nor states, they had only cities-polises). I will make another writing about this (it would require at least 1-2 pages). For now, just to quote one the prophet Daniel, who in the Bible calls Alexander the Great: “Sar Javan” what literally means – King of Yavans. This is quite correct because Alexander the Great was the king of the whole Helm (Balkan) peninsula, that is the Yavan land (FRANCOIS LENORMANT: De Populo Iavan, Journal des Savants, Aout 1882, p.481).
    Yava (i.e. Java) is a Serbian word (=daylight or reality i.e opposite of dream) and Yavan was a god of daylight. Much later, his replicas were Roman Jupiter and Zeus (btw, Greeks never used this name!)

    In my sketchy writing about Alexander the Great, I proved that he was not a Greek what even a common sense confirms. I listed all members of his 500+ years long dynasty (Greeks never had dynasties, states, kings nor tsars) and the list of Macedonian (Serbian) tribes. There was not one Greek king nor one tribe amongst them. Greek cities were Macedonian vassals. Their participation in Alexander’s army was symbolic as vassals, where sons of rich Athenians were semi-hostages to be preserved a peace back on Balkan while Alexander was in Asia.

    Accordingly, even without additional knowledge everyone can conclude who were Alexander’s dukes and top generals. They were Serbs and many of them were together with him in a princes’ school during their teenage days. After Alexander’s death, which was loudly celebrated by Greeks (btw this Serbian name for them was widely used only after Roman conquering in the 2nd c.BC – it means that terms ‘Minoan or Mycenean Greeks’ are oxymorons), his dukes divided the empire. (P)Tolomey Lagić took Egypt and established his dynasty which last member was Cleopatra. Kasander, his school friend took Balkan while Seleukić took Mesopotamia.

    I will write about Yavans and Alexander in more details. After we removed DUMBAS from our list of outstanding tasks, remain only the meaning of RG (I am glad to see that eNeM is back and I expect his confirmation that BP-oit cannot resolve this task), RB, PIR, Knowledge Zero and now Yavana(plus Alexander).

    1+
  10. Modeling South Asians as IVC-Steppe-AASI leaves out Neolithic related Near-East ancestry and puts it in Steppe, which gives really weird results for some Indus Valley populations (Sindhis getting 22% Steppe, Brahui getting 27%). Breaking it out into Hotu (Iran hg), and Barcin (neolithic near east) seems to fix that

    baloch/brahui don’t fit well with IVC-steppe-AASI model. trying to do that is trash tbh as you saw in previous qpadmin runs. way worse fits than pathan. sindhi probably is marginal on that…but seems to work for me.

    0
    1. IVC method puts Sindhis at 22% Steppe, same as Punjabi Brahmins. Using Hotu+Barcin puts it down to a more realistic 13%. Same issue is seen with Punjabi groups. Kambohs, a very West Asian heavy Punjabi group, come out as identical in Steppe to Jat Sikhs (who have middling West Asian but high Steppe) with IVC method, because the West Asian is being shoved into Steppe to compensate for poor reference samples.

      By the way, we know Sindhis don’t have the same steppe as Punjabi Brahmins, or Kambohs as Jat Sikhs, because other calculators show this. Whether Gedmatch calculators (like Harappa), or even using G25 with modern reference samples (Lithuania, Pania, etc). Its only when using the IVC method do you get these weird results.

      Andamese+Hotu+Barcin C+Sintashta+Kazakh EMBA+Devils Gate is the way to do it.

      1+
  11. Hello Razib,

    I noticed that when discussing the “ingredients” of Europeans, we use 3 different groups: Neolithic Farmers from Anatolia, Yamnaya, and H&G.

    For South Asians, we have an analogy and have 3 “ingredients” as well: Zagros Mountain Farmers, AANI Steppe, and some H&G groups that are distinct from the H&G of Europe.

    How similar are the AANI of South Asia to the Yamnayas who form one of the principal components of Europeans?

    Also, I noticed that the subreddit that was created is not being used right now. Anything I can do to help?

    0
    1. How similar are the AANI of South Asia to the Yamnayas who form one of the principal components of Europeans?

      the yamnaya+ (~25%) neolithic farmer => sintashta => steppe

      looks like ‘steppe’ is reflux from europe itself along the forest-steppe conveyor belt

      the reddit isn’t hopping but ppl post there. so idk

      0
      1. What was exactly this reflux? Which haplogroup carried Yamnaya? What was the ‘steppe’ – R1a and/or R1b? What was the relationship between them? Who brought (if they brought) the Sanskrit to SA or, otherwise, which language ‘steppe’ spoke? Thanks.

        0
  12. @Indthings
    Emotive? Whatever truth is it is. Doesn’t matter. I could be 16%, 6%, or .6%. Just curious how my individual proportions work out. I roughly know which groups I cluster with. From there I just use their percentages to ballpark. It also gives me a sense of other people’s proportions based on what fraction mine compare to theirs. But I don’t get into the weeds. I leave that to Dathang.

    My impression is that you think Narsimhan overestimates steppe? I just got my rough proportions from there. I’ll modify my paradigm with new info.

    0
    1. Don’t know if he’s written about it anywhere. I’m in one of his Twitter groupchats with a bunch of Pakistanis and we go back and forth about this stuff (history, genetics, politics).

      0
      1. tell him he needs to write about it so i can check it out.

        otherwise you are passing off whatsapp wisdom and that’s useless.

        (i did some PCA analysis the other day to look for west asian in pak pops and i didn’t have the power so i want to know what tests he did)

        0
  13. @thewarlock

    Good sir, I hope that you find this to be of some interest:

    55.4% Iran_HG-related
    32.7% AASI
    11.9% Eurasian_Steppe (Sintashta + Dali_EBA)

    So you were pretty spot on about yourself.

    For comparison, Tamil Brahmins:

    55% Iran_HG-related
    31.7% AASI
    13.3% Eurasian_Steppe

    Again, you were right; you’re very similar to Tamil Brahmins.

    For comparison, Rors subject to the exact same setup:

    43.4% Iran_HG-related
    39.5% Eurasian_Steppe
    14.9% AASI
    2.2% BMAC/Oxus_Civ

    Yup, same old expected stuff (40% Steppe, 15% AASI)

    The Paniya in the same setup:

    70.4% AASI
    29.6% Iran_HG-related

    Closest living population to AASI (only 30% West Eurasian, with no steppe)

    Now, this is me below as a point of comparison to South Asians (I’m being subjected to the exact same reference populations)

    Commentator/Seinundzeit:

    42.2% BMAC/Oxus_Civ
    25.3% Eurasian_Steppe
    18.8% Iran_HG-related
    8.6% AASI
    5.1% East Asian/Siberian (MNG_East_N)

    ^ I’m a son of Bronze Age Central Asia; Oxus Civilization represent! 😂

    But yeah, this is why I would tell Apthk that Pashtuns and Haryana Rors/Jats are different.

    I’m 25% Steppe; Rors are 40% Steppe. The largest single stream of ancient genomic ancestry for Pashtuns is something very closely related to the Bronze Age civilization of southern Central Asia (Tajikistan, Uzbekistan, Turkmenistan, etc); the Rors are quite clearly Iran_HG shifted, with almost no BMAC affinity. I’m 8%-9% AASI, while the Jats/Rors are 15%. And I have 5% Siberian/East Asian-related admixture, while the Jats/Rors have none.

    Below is a Pashtun with different tribal and geographic origins, and an even stronger BMAC affinity:

    58.9% BMAC/Oxus_Civ
    24.3% Eurasian_Steppe
    7.2% AASI
    6.2% Iran_HG-related
    3.4% East Asian/Siberian

    And just to show how diverse Pashtuns are, here’s a Pashtun (speaks Pashto, identifies as Pashtun… the whole deal) with roots in a region historically characterized by extensive intermarriage with Muslim Punjabis:

    39.4% Iran_HG-related
    23.9% AASI
    21.8% Eurasian_Steppe
    13.2% BMAC/Oxus_Civ
    1.7% East Asian/Siberian

    This Pashtun is genetically indistinguishable from a generic man of the Pakistani Punjab (despite his IRL distaste for Punjabis 😁).

    ^ Note also his much higher AASI in relation to Rors and Hindu Jats… but a noticeable level of BMAC, unlike Rors. This is typical of Pakistani Punjabis (more AASI… but the West Eurasian heritage is less “Indian”).

    3+
    1. Obviously this was done with your coordinates (nMonte).

      But my work with qpAdm and qpGraph have yielded identical results for the other samples.

      ^ So, you can be confident that you’d get the same sort of results with qpAdm.

      Speaking of qpAdm though, I’ve taken a break from tinkering with that sort of analysis, due to this preprint:

      https://www.biorxiv.org/content/10.1101/2020.10.12.336628v1

      The technical details provide much food for thought; it’ll take quite some time before I can reassess my methodology with qpAdm.

      2+
    2. Thanks. Pretty much the ratios I thought. I fall right within the realm of the typical subcontinental. It explains my broad passing look. I will look South shifted in some places and North in others or East or West. But I can pass among the diversity of a lot of places. Of course not all

      3+
    3. At this point, I would say “less indus.” But yeah we can say the more aasi shifted end of IVC spectrum= Indian look.

      Groups like Patels and Reddys are the best living relics of the IVC people.

      Yeah makes sense. I think by the modeling I clusted fairly well among generic N Indians out of the big groups discussed. And I look very “Indian.”

      0
    4. Yeah caste cline makes s lot of sense. A vaishya of N Indian should pretty much cluster with a Brahmin of South. I just like to jokingly flaunt my east euraidan Y H DNA and west eurasian maternal K1a. I am the anti cuck 😉

      #junglemasterrace

      1+
    5. hello @Commentator/Seinundzeit

      can you write a brief tutorial about how to calculate these population percentages ourselves., using oft-quoted tools like G25, qpAdm etc. i am thinking of digging deeper into this field of population genomics during this thanksgiving break. i took a look at some qpAdm related articles on the internet, and it can be a bit overwhelming for outsiders. also, i wonder how to obtain reference “left” and “right” datasets.

      i can figure it all out myself eventually, but that requires greater activation energy. i was hoping to get some pre-cooked meals. 🙂

      0
      1. also, i wonder how to obtain reference “left” and “right” datasets.

        reich lab has all data on their website. go there. the left/right pops are usually noted explicitly in the supplements if not always the papers. i like to use what is in the papers since then there is an apples to apples comparison

        0
  14. Wiki says (I believe about reflux): “The genetic data suggested that these cultures were ultimately derived of a remigration of Central European peoples with steppe ancestry back into the steppe.”

    >>> If this is the truth, it would mean that these guys came to Europe only to learn the language, came back to grassland and later continued to SA. To me – it is pretty unlikely the idea that so-called (P)IE language(s?) originated in Yamnaya, came to Europe, somehow imposed them on all mysterious (!) indigenous people (plus – their own northern R1a neighbours) who preserved this language(s?) while original IE people have forgotten the language(s?) which they brought to Europe – has less and less support (personally, I think it is ridiculous). We don’t know who ‘original’ IE people were, R1b or R1a, or both, what was the relationship between them (I know that they were fierce enemies) and who of them came back to grassland. There is much more convincing version of all of this, I already presented its key elements.

    0
    1. For fraudsters like @Milan Todorovic — an extremely deceitful person:

      1. Yamnaya (R1b haplogroup) is not PIE as Hittites have J2 haplogroup and 0 steppe ancestry.
      2. Mycenaeans have J2 haplogroup; they also maintain genetic continuity with Minoans. Linear Script B decipherment places the presence of Greek speakers aka Mycenaeans in Greece to at least 16th century BC.
      3. Speculation for PIE: According to Reich, only Armenia and Iran qualify for PIE as both primarily descend from Iran_N/CHG ancestry. AMT believers go with Armenia as they don’t believe Harappa to be IE speakers; while OIT believers go with Iran as PIE because Harappa also has Iran_N/CHG ancestry.

      I am fine with Armenia or Iran. I just want the truth.

      3+
  15. question for TheWarlock/DaThang/Razib/Commentator/ApTHK

    Tony joseph in his book says AASI is the single largest ancestry component in India.
    Unless I am mistaken Razib said the same in BP_E03.

    Though now this thread seem to indicate b Iran_HG-related is the single largest ancestry component just ahead of AASI.
    Is it due to different data or different methods of calculation ? OR
    Is is a mistake due to higher weightage given to Upper-Middle caste weightage as proportion of population. UP has 20% so called Upper caste. MH has around 5% ?

    About the 21 IVC periphery samples.
    Are they 50- 70% Iran_HG-related & 30- 50 % AASI related – I seem to recollect this.

    This may be covered elsewhere on the blogpost (Gnpx or BP) but if you get time answer this

    And are South and West Indians more IVC shifted while East and North are more Stepp/AASI shifted

    2+
    1. @GauravL

      “Though now this thread seem to indicate b Iran_HG-related is the single largest ancestry component just ahead of AASI.”

      I don’t know much (if anything) about Indian demographics.

      ^ But I think approximately 40% AASI for a large segment of Indians (but certainly excluding upper castes and Adivasi) makes sense.

      “Are they 50- 70% Iran_HG-related & 30- 50 % AASI related”

      Two of the samples are undoubtedly West Eurasians; in comparison to Iran_N, they seem to be only 10% AASI. But they might be even less.

      In fact, it could very well be that Iran_N is a “southwestern-shifted” spin on a similar stream of ancestry (just like how CHG can be construed as a “northwestern-shifted” spin on Iran_N). So again, if so, the two IVC samples might very well be even less than 10% AASI (the AASI is inflated since they’re being compared to a population that has affinities to the Levant and Anatolia, affinities that are completely absent in IVC_periphery).

      By way of contrast though, another sample seems to be around 40% AASI. Most of them seem to hover around 20%-25%.

      The big question though involves “AASI” itself. We need Mesolithic (and Paleolithic) aDNA from South Asia.

      I’ve played around with qpGraph in relation to Eurasian phylogeny, and I should note that sometimes things get really weird. No substitute for real aDNA.

      “And are South and West Indians more IVC shifted while East and North are more Stepp/AASI shifted”

      This isn’t as airtight as people think. One of the confounding factors is the IVC-related ancestry. It isn’t totally homogenous in affinity towards ANE. For example, in South India the West Eurasian heritage seems slightly even more ANE-like than IVC periphery! But only ever so slightly.

      Still, it’s enough to inflate Steppe percentages for some South Indians. Besides Brahmins, I don’t think many South Indians have more than mere trace amounts of steppe.

      For example, I assume that the Velama are typical of South Indians who aren’t Brahmin yet also not “lower caste” or “tribal”, and this is what I get for them using the same setup as above:

      59.5% Iran_HG-related
      38.7% AASI
      1.8% Eurasian_Steppe

      Like 1-2% steppe vs 13% steppe in Tamil Brahmins.

      1+
      1. The Chamars of Uttar Pradesh are interesting when compared with non-Brahmin South Indians:

        48.1% Iran_HG-related
        47.5% AASI
        4.4% Eurasian_Steppe

        UP Chamars are nearly 50% AASI! But unlike the Velama, the steppe is no longer a mere trace. Exactly as expected (Indo-Aryan vs Dravidian populations).

        1+
        1. That makes the Chamar community the most perfect Indian.
          Interestingly, in G25, the closest group to them is Maratha, a West Indian group.
          Chamar
          0.01482675 Maratha
          0.01856177 Dusadh
          0.02303645 Sakilli
          0.02627744 Pallan
          0.02662626 Yadava

          0
          1. @Son Goku

            The Maratha samples in question are essentially identical to these Chamars:

            50.4% Iran_HG-related
            46% AASI
            3.6% Eurasian_Steppe

            1+
          2. @ son goku
            That maratha samples(7 samples) were showing huge diversity, almost like Reddy/Kamma type to Mala/Madiga(South Indian Dalit) types, i think some of those samples were mislabeled(especially those who were identical to South indian dalits), Check MT08 and MT34(Maratha) samples, who were quite AASI shifted and were comparable to South Indian Dalit populations and also check MT-10 whose AASI levels were comparable with Reddy/Kamma populations.

            2+
          3. I would say their steppe is still a tiny bit too high. Patels, Valemas, and Reddys I think are the closest.

            The also carry the most flagship “Indian” look abroad, IMO. Granted, Patels can be a big exception. Their range is very big like marathas.

            0
        2. @Razib
          @Commentator
          @Dathang

          Do you know enough about Sri Lanka, Bangladesh, Nepal, and Pak to place me there genetically among a group? India is big but those places are smaller. I have had people from all of those places ask me, if I am from their country.

          0
          1. “I must visit Dhaka to meet all my brudas. I just hope it doesn’t become more Islamist over time.”
            @thewarlock It won’t. In a traditional Bangladeshi middle-class family, the parents often start the morning listening to Rabindra sangeet after Fajr prayer. The youths celebrate Bengali (non-Islamic) festivities like Pohela Boishakh with higher enthusiasm than Eid. Muslim-Hindu (or other religion) friend hood is very common among students. BTW You’re very welcome to visit BD. Alongside Dhaka visit Razib’s ancestral town Comilla as well. You can taste the best Rasmalai ever there. 🙂

            @Dip
            I agree with you. It doesn’t change the insight about haplogroup distribution in BD, anyway. Bangladesh is just like India in this case, both y H and y R1a are strongest, followed by substantial R2 and J.

            0
      2. Velema are closest living relics of IVC people along with patels.

        60 40 iranic: aasi with barely any steppe falls onto the cline.

        0
        1. Check out kannada Kodava. From the % given in Razib’s article they seem even better relics. Of course Patels, reddys and velamas are close behind them.

          1+
    2. Iran HG related is the biggest, followed by AASI, followed by steppe. Most extreme west shifted groups have a steppe approaching Iran HG but never above it on average. The other end has groups w/ more than double AASI compared to Iran HG related.

      A good average for subcontinentals would be someone like me.

      0
  16. @Razib

    Based on your back of the envelope calculation, if I do a retrospective of the incoming Steppes people and how did they mingle with the pre-existing elites of Indian society, then it comes out to be –

    40% of incoming Steppes population mingled with the pre-grouping of UCs
    40% of steppes population mixed with the pre-grouping of OBC/Others
    Remaining 20% of steppes newcomers went with the pre-grouping of SC/STs

    It looks like the Steppes admixture was caste-blind and more or less uniformly spread through the Indian population. There is a overcompensation for UCs but the majority (60%) did not contribute to the UC pool.

    2+
  17. @Razib,
    Just curious, How much genetic diversity does great apes have(Gorillas and Chimpanzees) compared to humans?, I know they have lot more diversity than humans but Does genetic divergence between eastern and western chimpanzees comparable to modern humans and Neanderthals?
    I know this question is not at all related to this thread lol

    0
  18. “Check MT08 and MT34(Maratha) samples, who were quite AASI shifted and were comparable to South Indian Dalit populations and also check MT-10 whose AASI levels were comparable with Reddy/Kamma populations.”

    @Akhilesh True. MT08 and MT34 have higher AASI and MT10 looks quite steppe shifted in comparison to other samples. I assume except for these 3 samples the rest are typical for most Marathas?

    0
    1. @son goku
      The remaining 4 samples are fairly closer to each other.
      I don’t know but, i think marathas are not a homogenous group, so we definitely need more samples to confirm whether they are typical for marathas or not,
      Any maharashtrian can correct me here, it seems like Marathas constitute almost 30% of maharashtra’s population(110 million), so they are large group, For large groups, we need more samples.

      2+
  19. “User avatar
    level 7
    Sakastani1
    7 days ago
    India (Republic of) didn’t exist before 73 yrs ago.

    Hell no, we are not Gangus. We are IVC descendants, and you are an alien culture.

    We look nothing alike. You are dark, and we are light-skinned.

    Go claim Bangladesh, Burma, or Sri Lanka.”

    HAHAHAHAHAHA

    The trolling is fantastic from the Indus Larper gang

    0
  20. Have you thought of making a calculator that can do this w/ 23andMe data? I’m sure you could get people to pay you ($10-$20) for each run, similar to G20 coordinates.

    2+
    1. @Razib Do Y-DNA/mt-DNA haplogroups impact personality, immunity and cognition? Or only autosomal DNA?

      And, How much steppe, Indus, East Asian and AASI do average Biharis, Bangladeshis and Rohingyas have?

      0
      1. “And, How much steppe, Indus, East Asian and AASI do average Biharis, Bangladeshis and Rohingyas have?”

        Razib mentioned Bangladeshis as 85% Patel + 15% Tibeto-Burman. That means Bengalis are both 7% less Iran_HG and 7% less AASI than Guju Patels. He also said the steppe is 10-15% among Bangladeshis just like most North Indians.
        Rohingyas could be regular Bengali + extra east Asian?
        Bihar has many castes. The agricultural Bihari castes would be similar to Patels?

        “Despite genetically not being very South Indian, it is interesting that many South Indian Brahmins look quite Aboriginal. There is a Black Tamil Brahmin in this site who doesn’t even look Indian: He looks very aboriginal(Negrito or something)
        OTOH average South Indians look very East Indian and quite Caucasoid.”

        The term Caucasoid is vague. A Scandinavian like Erling Haaland looks nothing like the Mediterranean like Gianluigi Buffon.

        1+
        1. Caucasoid-looking South Asians look closer to the Mediterranean/West Asian type but on average North Euros aren’t very different in terms of facial features compared to other West Eurasians:
          Norwegian crowd: https://www.tnp.no/newsimg/norwegian_wood_festival.jpg
          German crowd: https://cdn-prod.opendemocracy.net/media/images/PA-38304769_70rkCCZ.width-800.jpg

          Life-style factors might be the reasons why many beggars and lower-class females in Bangladesh look so Australoid shifted although lower-class males & Bangladeshi labourers/Rickshaw-pullers usually look Caucasoid-shifted.

          Many Australoids(including Paniyas) also inherently have sharp-enough features(especially at the lower part of the face) which can contribute to a pseudo-caucasoid look.
          Or may be the West Eurasian-derived genes responsible for facial features are very dominant.

          Paniyas:
          https://i.pinimg.com/originals/3d/6c/e0/3d6ce0fda8cfa41f8a56332723bc3817.jpg
          https://www.gutenberg.org/files/42996/42996-h/images/pl6-081.jpg
          https://www.gutenberg.org/files/42996/42996-h/images/pl6-105.jpg

          0
    2. The numbers are above. I am Jain Vania and literally cluster with them on all calculators

      Me
      55.4% Iran_HG-related
      32.7% AASI
      11.9% Eurasian_Steppe (Sintashta + Dali_EBA)

      For comparison, Tamil Brahmins:
      55% Iran_HG-related
      31.7% AASI
      13.3% Eurasian_Steppe

      Indus is Iran_HG-related+AASI

      Steppe is steppe

      S Indian Brahmins have numbers comparable to N Indian mid caste vaishyas.

      1+
      1. Despite genetically not being very South Indian, it is interesting that many South Indian Brahmins look quite Aboriginal. There is a Black Tamil Brahmin in this site who doesn’t even look Indian: He looks very aboriginal(Negrito or something)
        OTOH average South Indians look very East Indian and quite Caucasoid.

        0
        1. I mean these averages. There is likely a range. And also phenotype doesn’t equal genotype. Phenotype is composed of relatively few genes relative to one’s entire genome. Therefore a “biased” look towards one aspect of autosomal ancestry is certainly a possibility and does occur not infrequently. This is why siblings from the same Indian family can often look so different in terms of Western pull of facial features and skin tone.

          2+
        2. @Dip
          \Black Tamil Brahmin in this site who doesn’t even look Indian:\
          I am a TB who is black. OTOH, you know very little of India or Indian society when you say ‘not even Indian’ . I don’t which part of the world you are from

          3+
          1. @Dip

            have been using the same pic over many months. I don’t have any pic of leaf underwear with spear in hand for fishing.

            4+
  21. Another thing – How good a Proxy is Steppe for the Vedic Aryans or their ancestors ?
    We dont really know when they started picking up Non Steppe ancestry in India. Maybe early Rgveda itself which has non IA names in prominence.
    What i am speculating is Rgvedic Aryans themselves could be Steppe + Indus people ?

    0
    1. Possibly, Vedic Aryans were genetically mostly Indus and had minority steppe. Maybe everything Indo-Aryan came from these Indus-people.

      0
      1. Possibly. But I think proto-sanskrit was likely spoken on the steppe. Indo european religion does not equal Hinduism and may not even really equal ancient Hinduism. It all depends on composition timing and location of Rig Veda. But it’s hard to say. Also it is controversial what dasyu and such actually refer to. So the racialized aspect of it is hard to interpret well.

        Regardless, the Hinduism we know today is certainly a mix of pre and post steppe migration beliefs. Hence why the pagan indo european faiths of Europe are quite different.

        Was caste present in the Indus Valley? That is the next major question. Do Kalash have it? I don’t think so. What caused it to develop? Was it present before or did the interaction between the steppe and non-steppe accelerate its creation. And regardless, why did it cement only really starting 2000 years ago. What prompted that move by the Gupta Empire?

        0
        1. So much guessing…

          Indo-Aryans were Indus people? But ‘steppe’ spoke proto-sanskrit? Who spoke, where PS originated? Indo european religion? What is this? Who are Indo-Europeans? Rig Veda? What is the meaning of these words? The pagan indo european faiths of Europe are quite different? Says who? It will be the next discussion topic after linguistics. Was caste present before or did the interaction between the steppe and non-steppe accelerate its creation? Is there a theoretical possibility that they came with ‘steppe’? And, who are these bloody ‘steppe’?…

          In sum – ‘steppe’ came with some faith different from Hinduism, speaking proto-sanskrit which locals took from them and produced Vedas? Theoretically – not impossible.

          0
        2. @thewarlock

          Do you think that it’s possible to fill in the candidate for the IVC prima lingua without considering Sanskrit or a form of Proto-Sanskrit? The one thing that Rakhigarhi DNA has falsified is the candidature of a Dravidian language. More samples should remove this possibility totally. Munda is already ruled out, then what remains? Were they mute while building cities – Mohenjodaro is five times the size of the walled Vatican.

          Caste was already present in the Indus Valley. There is ample archaeological evidence to prove this. In the open thread we have a comment by @principia on the spatial analysis of residential areas to asses effects of caste. The same methodology has been used by archaeologists to assess residential areas in IVC ruins. Caste had architectural foundations. Read up on Iravati Karve’s work.

          The people who brought the Steppes DNA to India were the Yonas – starting from 1000 BCE onwards. They did not have caste (but they had slavery) and Indian traditional sources (Buddhist and Hindu) record this as well. They married at will – perhaps due to paucity of willing mates. They performed the role of Kshatriyas – mercenaries and cavalry. The migration was totally male-mediated. Hindu shastras call the union of such kshatriya men and shudra women as breaking taboos.

          Sociologically this perfectly describes the reaction of a well settled society to the actions of migrants entering its confines and breaking taboos. You can see the same reaction all across Europe today – the hardening of traditional camps. By the turn of the millennium, between the Mauryas and the Guptas – endogamy had set in completely with rigid formalisation.

          Again if you assume the Steppes migrants brought caste and Sanskrit to India – you can neither explain the past and the later – without being able to explain the cracks. The theory with the least discontinuities is the most elegant.

          0
    2. Have been thinking on these lines for long time. If one follows current consensus of arrival around 1500BC and almost immediate writing of rgveda, there would be some reference to such migration event. Is it possible to push back the arrival of steppe people to 3rd millennia BC / mature IVC?

      Other question is for linguist if a civilization can start writing their ideas/memes using language they picked up from migrants? If yes then the connection between IA language use and exclusive steppe origin of its users is untenable.

      Talageri answer to this question is that Vedic people (composers of Rgveda) are a subset of IE (Indo European) language speakers living in Haryana/western UP. There are more such IE peoples on their east as well as west. Since the area was already inhabited by IVC people they are one and same. (or mixture of two?) ‘Aryan’ according to him is just a honorific name for one’s own group and doesnt exclusively refer to Steppe migrants.

      0
  22. i have proposed that the vedic aryans or indo-aryans when they entered into were not ‘pure’ sintashta. this means we under-estimate their numbers/contribution.

    e.g., if they were already 50% non-steppe when they entered south asia (mixing with indus periphery without AASI in Afghanistan?) that doubles the estimates

    2+
    1. 50% non-steppe of what kind? I don’t think that it could be BMAC because the BMAC input is pretty low to non-existent in south Asians. Maybe Shahr BA1 instead, that is possible, but even then, 50% sounds too low.

      0
  23. Anonymous Says:
    This comment by “ugra oogra ogreess) is jhoot fraud patent falsehood.
    Ugra
    NOVEMBER 24, 2020 AT 10:52 AM
    @thewarlock

    Do you think that it’s possible to fill in the candidate for the IVC prima lingua without considering Sanskrit or a form of Proto-Sanskrit? The one thing that Rakhigarhi DNA has falsified is the candidature of a Dravidian language. More samples should remove this possibility totally. Munda is already ruled out, then what remains? Were they mute while building cities – Mohenjodaro is five times the size of the walled Vatican.

    Caste was already present in the Indus Valley. There is ample archaeological evidence to prove this. In the open thread we have a comment by @principia on the spatial analysis of residential areas to asses effects of caste. The same methodology has been used by archaeologists to assess residential areas in IVC ruins. Caste had architectural foundations. Read up on Iravati Karve’s work.

    The people who brought the Steppes DNA to India were the Yonas – starting from 1000 BCE onwards. They did not have caste (but they had slavery) and Indian traditional sources (Buddhist and Hindu) record this as well. They married at will – perhaps due to paucity of willing mates. They performed the role of Kshatriyas – mercenaries and cavalry. The migration was totally male-mediated. Hindu shastras call the union of such kshatriya men and shudra women as breaking taboos.

    Sociologically this perfectly describes the reaction of a well settled society to the actions of migrants entering its confines and breaking taboos. You can see the same reaction all across Europe today – the hardening of traditional camps. By the turn of the millennium, between the Mauryas and the Guptas – endogamy had set in completely with rigid formalisation.

    Again if you assume the Steppes migrants brought caste and Sanskrit to India – you can neither explain the past and the later – without being able to explain the cracks. The theory with the least discontinuities is the most elegant.

    My response

    Absolute fraud – yonas janas
    Janas is common west Aryan (*nonSemite•euro) name Jan= piously Godly man – vegetaryan morality
    Dharm unequivocally condemns all forms of slavery .
    * Casteless (pseudo*pretensions of caste ) hindoo lacking Both lineage and character revel in slavery .- polygyny polyandry etc- no divorce in hindudom but there is Sati. Fire 🔥 brahminical element.
    Sikh religion direct descendant of RigVedic Aryan religion is the world’s only warrior religion and Sikh religion Alone condemns all forms of slavery .
    Source — sikharchives.net now defunct
    Wretched existence in perpetuity for enemies of Afghanistan.
    Loy Afghanistan
    Khalistan Zindabad
    Jeevai jeevai Pakistan home aumsweethome to k2 penultimate Himalya peak.

    2+
    1. That’s the girl talking. ‘Steppe’=Yonas???? At least, it is original. Why it was hidden until now? I asked already – who are these bloody ‘steppe’? Go girl and please do not ignore VV’s hat and my questions.

      1+
      1. On the second thought…

        If Yonas are Yavanas (or Ionians – although it is a later term, they are not Greeks, who did not live on Ionian Sea) that is actually correct because Yavanas were Serbs. Only, it was few hundreds years earlier and Serbs never practiced slavery.

        1+

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.