Notes the emergence of “Indic civilization”

Note: This post is a supplement to the podcast below.

People get hung up on particular words a lot. This post is to clarify some terminology from my own perspective. It needs to make clear here that I am a semantic instrumentalist. Words don’t have power or meaning in and of themselves but point to particular concepts and patterns. If we disagree on words while agreeing on the concepts and patterns, the disagreement is semantic.

To give an illustration about the “power of words,” I have read works on “Western history” which begin the narrative in Egypt and Sumeria. As the centuries proceed, the focus moves north and west, and eventually, the Near East is excluded from the West. Clearly, most people can agree that the Near East is, and became, very distinct from what we term “the West,” but if our history is to deal with Northwestern Europe, it will start with the Roman period, and its roots clearly owe something to the earlier Near East. The reality is that the West that the histories outline developed much later (arguably after the fall of the Western Roman Empire), but its roots are diverse and broad, inclusive of Near East antiquity.

When I use the world “Indic,” please keep in mind that I am focused in particular on the civilization which had crystallized by the Gupta period across South Asia. The civilization which gave rise to concepts which form the basis of the Dharmic family of religions. Moving forward, and moving backward, this is the reference cluster of characteristics.

In the podcast below I made the assertion that the precursors to Indian, Indic, civilization from the north to the south are rooted in prehistory. That the adoption of a set of religious ideas and identities promoted by Brahmins across much of South India several thousand years ago was not a coincidence, because South India was connected in some deep ways to North India in the prehistory period after the fall of the Indus Valley Civilization (IVC) 4,000 years ago.

Shortly I will have a book review in India Today on a work by Tony Joseph which tries to synthesize archaeology, linguistics, mythology, and genetics. Though I knew many of the details in the book, Joseph brings the strands together which clarified and solidified some general intuitions I have developed.

First, I do think that the most likely point when the Indo-Aryan eruption into South Asia occurred is around 1500 BC. Earlier I had been very vague about the timing because there had not been any ancient DNA. But the presence of people with a “steppe” genetic profile, which is ubiquitous in much of Northern India today, seems to date to the period after 2000 BC. Additionally, 1500 BC is when Indo-Aryan rulers (albeit, culturally assimilated) seem to become dominant in the Upper Euphrates are of Syria.

One can push the date a few centuries earlier (there is a possible Indo-Aryan text in Syria dated to 1750 BC), but not too much earlier. And, Joseph claims that the dominant metal in the earliest Vedas was copper, not iron. Since the Iron Age in India starts around 1200 BC that definitely puts a lower limit on when the Aryans arrived (though the earliest genetic samples in the Swat Valley with “steppe” ancestry shows up around 1200 BC, so we already knew this).

So we can put the ethnogenesis of the Indo-Aryan Indic component in Northwest India in the centuries around 1500 BC.

But they did not come into an empty subcontinent. The Harappan society, the IVC began to go into decline around 1900 BC. There are various debates as the regional continuity of this civilization down to 1300 BC, and as to why it declined. Though more archaeology needs to be done, here is my own personal position as of now:

– The IVC was likely fragile and in decline before the arrival of large numbers of agro-pastoralists

– A good model for the arrival and cultural domination of the Indo-Aryans may be the situation of post-Roman Britan or the post-Roman Balkans, where large numbers of barbarians arrived and assimilated peasant cultures whose elites had disappeared

– In the Near East the Indo-Aryans, and later the Iranians, show no overwhelming ideological reason to destroy civilizations that were existent prior to their arrival. On the contrary, agro-pastoralist elites aim often to take over elite positions in those societies so that they can extract rents and become wealthy. The problem is that sometimes the arrival of these people destabilize weak or declining societies, and the social order rearranges and regresses.

We don’t know the details of how the IVC was organized, but its relative uniformity of architectural design and layout indicate large-scale elite coordination of some sort. The societies and subsequent to the IVC seem simpler and less indicative of large-scale coordination. This sort of “unwinding” was common to earlier societies.

But we also need to move beyond the focus on the Indus Basin the Doab.

The Neolithic, indicative of a transition from hunting and gathering to pastoralism and farming, began in South India after ~2500-3000 BC (depending on the source). While the IVC was flourishing, a zone of agro-pastoralists seems to have pushed south along the western edge of the Deccan. Joseph (and others) reports that placenames of Dravidian origin seem to be common in both Maharashtra and Gujarat. Curiously, he notes that Dravidian placenames do not seem to exist in Bihar or Bengal.

Recent phylogenetic work on the extant Dravidian languages indicates diversification beginning 4,500 years ago. This is 2500 BC. A possible beginning date for the South Indian Neolithic.

When the IVC was a mature civilization it seems that some farmers pushed the frontier further. While the IVC was unwinding in the northwest of the Indian subcontinent, I believe that their “country cousins” in South India continued with their small-scale society.

The “Ancestral South Indians” (ASI) have now been shown by the Reich group to be about ~25% “West Eurasian” in ancestry. This is almost all “Iranian Farmer.” That is, the ancestry derived from western Iran before there was much mixing between Iranian farmers and Anatolian farmers (which occurred before the rise of civilization). Today there are tribal and low-caste groups in South India whose ancestry is nearly all ASI.

I believe that this ASI population formed during the South Indian Neolithic when ash mounds were common through the fusion of people out of the IVC with indigenous people, Ancient Ancestral North Indians (AASI).

But South India is not all ASI. There is in fact caste genetic stratification between Dalits and Adivasis on the one hand, and local non-Brahmin peasant and elite groups (e.g., Reddy). The latter has more “Ancestral North Indian” (ANI), specifically, Iranian farmer.

I believe this is a function of the second wave of continuous migration out of northwest India, likely via the lower Sindh and Gujarat. These people assimilate and conquered the earlier agro-pastoralists and likely absorbed the remaining hunter-gatherers. The relatively deprived position of the original  ASI populations is indicative to me of their long-term marginalization by the new arrivals.

But these people were not simply “Dravidian traditionalists” who were fleeing their homelands. A quick survey of Y haplogroups shows that R1a, associated with the Northwest, and upper-castes, is found through India. Though in far lower frequencies, it is found among Dravidian peoples, even Dalits, and Adivasis in the South.

To me, this is suggestive that the Dravidian-speaking populations that moved south along the western coast of India were, in fact, a synthetic people who were expanding out of a hybrid cultural zone. Some of the populations, tribes, in this hybrid zone, were Indo-Aryan. Likely the dominant element was. But some of them retained their Dravidian language, though they assimilated some Indo-Aryan groups in their mix. The linguistic diversity may have been greater than this, as indicated by the presence of Burusho, a linguistic isolate.

The later expansion of polities such as that of the Maurya to the south on the edge of history then is not an expansion into alien territory, but a conquest of a related set of peoples, distinct, but nevertheless connected.

Obviously, the push east out of the Doab, and down the Ganges, would have resulted in an encounter with Munda people. The people who were moving eastward were Indo-Aryans, who had possibly assimilated Dravidian peoples in the region of Harappa (most of the ancestry even of Brahmins from the Gangetic plain remains non-Indo-Aryan, as is the case for the Kalash of Pakistan). The Munda seem to have arrived after 2000 BC, and admixed with another AASI population. This suggests that Northeast South Asia was not touched by the movement of Dravidian peoples that affected the Deccan. Additionally, the extant Munda tribes lack Indo-Aryan ancestry.

The final point I’d like to make then is that the prehistoric roots of what we term perceive to mature into Indic civilization by the period of the Gupta as its Classical form has its roots in the synthesis between 2000 BC and 1000 BC. First, as the IVC as a civilization defined by widespread coordination and uniformity unwound, and Indo-Aryans enter into the matrix as a hegemonic agro-pastoralist group who introduced a way of life accessible to many local non-Aryan elites, who were assimilated into their culture. Meanwhile, other groups remained Dravidian speaking, while assimilating into themselves some groups of Indo-Aryans, and following a southward trajectory that had been pathbroken by cultural relatives centuries earlier during the IVC.

3+

131 thoughts on “Notes the emergence of “Indic civilization””

  1. From the Dravidian paper:
    “His main concern is that reconstructed vocabulary terms from the South I and II subgroups suggest a society that is advanced beyond anything that has to date been found in the archaeological record of peninsular India, and thus whether there is a mismatch between linguistics and archaeology.”

    The vocabulary is a lot less surprising if the Dravidian languages were brought by migrants from the IVC. Also possible to speculate that South I and II Dravidian were brought south by the later wave of migrants that were mixed with Indo-Aryans. The times seem to match up, and that wave would probably be more likely to include people from the (formerly) more developed core of the IVC, as opposed to the earlier migrations of peasants and herders on the fringes who might not have carried vocab related to the more advanced parts of IVC culture and technology.

    2+
    1. // Also possible to speculate that South I and II Dravidian were brought south by the later wave of migrants that were mixed with Indo-Aryans //

      This is more than mere speculation I’d have thought. E.g. I’ve always believed that Indo-Aryans lacked the organizational structure to create the different schools for composition, redaction and dissemination of Sanskrit metrical poetry and development/standardization of grammar – a *stupendous* feat in a purely oral[*] culture. Vedic work stands out in this regard compared to all IE bardic traditions on how well structured and copious it is. (Greeks are an exception because they adopted writing long ago)

      That sort of organization is the hallmark of an organized priestly class of a settled culture, namely the IVC. And given the sheer size of the lexical and phonetic substrate of Dravidian in the oldest IA (and innovations thereon), it is very likely that core Brahminism is essentially Dravidian not an IA imposition. It is almost like the (migrating/invading) Indo-Aryans were converted to a new hybrid religion by a priestly class of primarily Dravidian extraction – obviously with some inter-mixing with the new steppe arrivals.

      And this process repeated itself again and again as new waves of the steppe “invaders” came in .. starting from Old Iranics themselves, maybe fleeing Zoroastrian reform in the BMAC, absorbed into the latter-day atharvaveda schools (cf. athar “fire” in *not* attested in any Sanskrit nirukta-s, i.e. etymologies, but atar is a valid and well-used word in the Avesta). Obviously shaka-s, kuSANa-s, hUNa-s came in wave after wave and were kSatrIya-fied by the brAhmaNa-s.

      [*] I do not think IVC were literate and the IVC “script” does not code phonemes or logograms (though some Dravidian experts, like Mahadevan, disagree)edic

      4+
        1. “And given the sheer size of the lexical and phonetic substrate of Dravidian in the oldest IA (and innovations thereon), it is very likely that core Brahminism is essentially Dravidian not an IA imposition.”

          Very interesting hypothesis. Would you know anyone who’s written more on this?

          @Zack
          They could have adopted/adapted extant Dravidian culture without mixing genetically to the same extent as other groups.

          0
          1. @Prats @Zack

            // they don’t show the spike in Dravidian ancestry //

            Correction: don’t show the spike in Dravidian paternal ancestry.

            I am not suggesting that the Brahmins do not have Indo-Aryan fathers (or primarily so), but that their Brahminism is better explained by their non-Indo-Aryan ancestry. Proto-Indo-Aryan shamans and bards who were assimilated into a Dravidian priestly order, but with language replacement.

            After all, the “kafirs” of Nuristan and Kalash do not have the Brahminical foot-print and probably never did. Brahmins like kalhaNa actively called them mleccha-s even in the pre-Islamic period. No Indo-Aryan sympathies whatsoever!

            Besides, Dravidian is a linguistic/geographical marker not genetic (draviDa is a hyper-correction of damiLa; a reference to language and Southern geography in Sanskrit)

            9+
        2. I agree this is a puzzle, the vedic people alone couldnt get this done, especially if no other set of their cousins could do this much either. However, I am not sure of this being product of IVC, I dont know the reason, people can always speculate of course.

          I think there is a curious sentence in rig veda, praising Indra for relieving the oppressed dasa or something like that. That is followed by the question of human sacrifice, another passage that first invokes it and later rejects it. There is a recurring theme of human sacrifices and goddess worship. With one ivc coin being interpreted as a case of human sacrifice. Ancient societies around the time had this and I suspect this is true in case of IVC as well.

          0
        3. “Brahmins are not Dravidian / they don’t show the spike in Dravidian ancestry.” — i remember one “priestly” group in narsimhans pre-print labeled as gaud_karnataka(probably the gaud saraswat brahmins) had barely ~4% Steppe_MLBA with mostly InPe and remaining AASI. So,i am not sure if what u are saying applies to all the south indian brahms groups . Even the nearby living brahms groups in TN (vadama vs brahacharanam) didn’t intermarry with each other till recently.
          Besides,Didn’t the north indians brahms (particularly the UP ones) received recent(roughly only 65-70 gens ago) additional ANI admixture as per Moorjani’s 2013 paper ?

          0
        4. What do you define ‘Dravidian’ as?

          Historically Brahmins from Tamilakam (ancient and medieval Tamil Nadu and Kerala) have referred themselves as ‘Dravida’. Adi Shankara being a case in point.

          Dravida is a sanskritised form of Prakrit Damila (Tamilar = literally ‘Tamils’ in Tamil language).

          No doubt it was a regional/linguistic identity and not a caste exclusive one.

          4+
      1. redaction and dissemination of Sanskrit metrical poetry and development/standardization of grammar

        Wasn’t that done in Panini’s time? Or do you think Panini was just channeling his predecessors?

        0
        1. paNini was standing on the shoulders of giants.

          A single human understanding and redacting all of Vedic linguistics, coding-up a generative parser for it and inventing a meta-language to write it in and composing mnemonical poetry in that meta-language, all without knowing how to write, beggars belief.

          0
      2. The earliest Rig Veda hymns are clearly foreign to the IVC/indigenous Hindu religion which finds its earliest expression in the later scriptures (e.g. Vedanta). The lingam worship of the dasa is denigrated for example.

        One more recent parallel is how today many native ‘indians’ are not reciting koranic arabic after coming under Mughal/Arab dominance.

        A similar dynamic happened to the IVC people under the Aryans.

        2+
        1. Aryan worship with its fire rituals are clearly contrasted with the Dravidian water rituals. The core of Hinduism is Dravidian/IVC worship with a Aryan superstrate (Vedic hymns).

          Siva linga worship is more popular than Agni/Indra worship.

          1+
          1. in the later vedic age there are clear mentions of some brahmins being the progeny of Dasi women. The Aryan-Dasa religious conflict culminated in an eventual hybridisation.

            Difference in religion was the main beef the Aryans had with the native Dasa going by the Rig Veda.

            They are denigrated as riteless heathens in the earliest hymns. As well as being dark skinned.

            2+
          2. // They are denigrated as riteless heathens in the earliest hymns. As well as being dark skinned //

            Shit like this gets my goat. Can you point to a single reference to colour commentary on dasyu-s in the Vedic scripture?

            Secondly, being a riteless heathen is a fairly common slur all the way until the time of the Mahabharata and applied just as well to darada-s (Indo-Aryans), cIna-s (Sino-Tibetans) and pArasika-s (Persians) – all of a fairer skin tone than the average Brahmin.

            0
          1. There are numerous quotes from the Rig Veda, I suggest you re-read it.

            “Blowing away with supernatural might from earth and from the heavens the swarthy skin which indra hates”

            “Driving the black skin away. Quelling the riteless Dasyu, may we think upon the bridge of bliss, leaving the bridge of woe behind”.

            It is not at all surprising after seeing the DNA results at Rakhigarhi. There would have been a lot of black skinned people in the IVC prior to Aryan invasion and admixture.

            0
      3. > given the sheer size of the lexical and phonetic substrate of Dravidian in the oldest IA (and innovations thereon), it is very likely that core Brahminism is essentially Dravidian

        Per “Substrate Languages in Old Indo-Aryan”, M.Witzel, there is no Dravidian substrate in the oldest parts of RgVeda. Dravidian influences only begin in middle and late part of the corpus. Basically, retroflexion is not unique to Dravidian language families.

        I don’t think there is any basis for “sheer size of the lexical and phonetic Dravidian substrate”.

        Given that strict endogamy in the subcontinent set in in the Gupta age, I’m not surprised that contemporary Brahmin Y-haplogroups aren’t overwhelmingly of steppe-origin.

        0
        1. The Aryans may have had first contact with pure AASI people before Dravidian speakers, hence the later arrival of a huge Dravidian substratum in the later Vedic period. Witzel should stop using the stupid term Para Munda, it is confusing and inaccurate. Munda was a later arrival too.

          0
  2. it’s an interesting issue as i’m positing here that IVC complex emerged, while non-state populations of some sort moved beyond it and introduced farming. in europe almost all the continent had become agriculturist by the time a state had emerged….

    0
    1. I think that the best analogies are to Egypt and Sumeria, which like the IVC were united in a very fertile major river basins and were separated by substantial distances from any plausibly “near peer” adversaries. The IVC and Egypt also both adopted agriculture at roughly the same time and took 2000-3000 years before they developed genuine cities. All of these areas developed state civilizations much sooner than Europe did which didn’t have a unifying and isolated river system and had a Neolithic revolution that in many places started as much as a couple of thousand years later than Egypt or the IVC, and about four thousand years later than the Fertile Crescent.

      0
  3. @Razib, so at what point did the Dravidian and Arayan hybrid population or Arayans who absorbed Dravidian people of Harappa migrate to the Northeast to mix with the Munda + AASI hybrid? And why do you think this was, because the Indus Valley dried up or just out of curiosity?

    0
    1. so at what point did the Dravidian and Arayan hybrid population or Arayans who absorbed Dravidian people of Harappa migrate to the Northeast to mix with the Munda + AASI hybrid? And why do you think this was, because the Indus Valley dried up or just out of curiosity?

      IVC ppl seem to have done some rice cultivation, but it was clearly not their forte. the arg is that move to deccan easy for their basket of crops. the indo-aryan matrix = more rice adoption + iron.

      0
      1. “http://discovery.ucl.ac.uk/1347997/1/2011GL050722.pdf” among other papers discuss the impact of monsoon change after 4200 year BP, and the response of early Indians to aridification in NW India. Permit me to reproduce in entirety

        “The significant aridification recorded after ca. 4,000 years ago may have spurred the widespread
        adoption of sedentary agriculture in central and south India capable of providing surplus food in a less secure hydroclimate [Asouti and Fuller, 2008]. Archaeological site numbers and the summed probability distributions of calibrated radiocarbon dates from archaeological sites, which
        serve as proxies of agricultural population, increase markedly after 4,000 BP in peninsular India (Figure 3; see auxiliary material) for cultures of Southern India and the Deccan Plateau. In contrast, the same process of drying elicited the opposite response (Figure 3) in the already arid
        northwestern region of the subcontinent along the Indus River (Figure 1a). From 3,900 to 3,200 years BP, the urban Harappan civilization entered a phase of protracted collapse. Late Harrapan rural settlements became instead more numerous in the rainier regions at the foothills of the
        Himalaya and in the Ganges watershed [Madella and Fuller, 2006]. During the Iron Age, after ca. 3,200 years BP, adaptation to semi-arid conditions in central and south India appears to have been well established with 60% of sites in areas with <1000 mm of rainfall today and a significant
        number of sites (18%) in areas with <600 mm (see auxiliary material). Later, the rapid increase in rainwater harvesting structures that occurred after 1,700 BP in the semi-arid regions of south India [Gunnell et al., 2007] points to an expansion of the cultural adaptation to an additional increase in aridity (Figure 3). Although tentative, these correlations between hydroclimate and cultural changes in the Indian subcontinent suggest distinct societal responses to climate….."

        This suggests some interesting hypotheses; it might mean the InPE people wandered north and south; mixed with in coming Steppe-MLBA/INPE people (now in the Himalayan foothills) relocating there due to 4200 year BP events elsewhere; and AASI people in south and Deccan, cultivating a new Sorgum/millet/pulse cultivation. Rice kind of receded into background, until the SW monsoon strengthened around 3200 BP. This still does not explain R1A extent everywhere, but suggests climate change as a cause of admixture.

        Yes, I continue to trupet geology and climate variability as a cause of geneticc admixture everywhere today.

        1+
        1. No question re climate change, migrations and civilization.

          It appear the collapse of the 11th century Sinhala polity in the North Central Province (Rajarata) was possibly due to climate change and south indian Invasion.

          2nd author, Liviu Giosan was a fellow grad student at Stony Brook.

          0
        2. I agree climate change in early periods must have been devastating, 200 yrs (or 900 yrs) with poor monsoons would be catastrophe even now. Climate, tsunami,volcanoes, earthquakes etc. I take them all seriously.

          0
  4. Interesting view of Indian Civilization. Some thing I can nearly agree with on the basis of my readings of history, culture and linguistics.

    That said, I am not a fan of genetics outside medicine and I am underwhelmed by it.

    1+
  5. “But these people were not simply “Dravidian traditionalists” who were fleeing their homelands. A quick survey of Y haplogroups shows that R1a, associated with the Northwest, and upper-castes, is found through India. Though in far lower frequencies, it is found among Dravidian peoples, even Dalits, and Adivasis in the South.”

    Just to play a little devil’s advocate, after thousands of year I see no reason for there not to be. DNA can be easily spread, we only need look at south america where I don’t remember the paper, but an analysis of some supposedly pure natives today are found to at least contain 2 percent or so european DNA with Spanish peninula related Y haplogroups, that was only after a few hundred years. Also another paper came showing somewhere about venezuela or so some of the tribes by the rivers had south asian dna, most probably related to all the shipping around the rivers and south asians being part of some of the crew most likely indo-carribbean. After the thousands of years I would be more surprised if there were no Indo-aryan Y haplogroups than not frankly.

    2+
  6. fwiw, i checked. a substantial minority of brahmin Y chrom in north india (paternal lineages, so gotras) are clearly non-indo-aryan (HG H eg)

    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1380230/

    after thousands of year I see no reason for there not to be.

    you are hand waving. you don’t really know genetic dynamics, i do. gene flow does happen, but very little evidence for this in s. asia over the last 2,000 years. additionally, r1a is not found much in western iran, but it is in south india. it is found in eastern iran. this is indicative of certain cultural and geographic connections.

    genes don’t ‘flow’ automatically with time. geography and history drive it.

    1+
  7. ok…i did some more digging in Y chromosomes of 2000s papers.

    the ‘indo-aryan’ ancestry of north indian brahmins probably in the range of 20% of their total ancestry. looking at the Y chromosome probably 50-75% of their y chromosomes are indo-aryan (about 50% is r1a, but hard to assign j2 and others definitively to aryan or dravidian).

    2+
    1. Razib just listened to the podcast and the comments on the Sinhalese which are inaccurate. The Sinhalese cluster closer to South Indians genetically than Bengalis and Gujaratis. The bulk of Sri Lankan ancestry derives from the spread of agriculture and civilisation from Neolithic South India (circa 1000 BC). They were likely Dravidian speakers who mixed with the indigenous Veddahs. Later long distance migrations of North Indian elites before 500 BC resulted in language replacement paving the way for the Sinhala language. The Maldives have a similar history, genetically closer to South Indians but linguistically Indo-Aryan.

      Interestingly, the majority of Jaffna Tamils are genetically closer to the Sinhalese than to Indian Tamils. Language replacement occurred again in the North following the 150 year long Chola occupation. Most Jaffna Tamils are descended from Tamilised Proto-Sinhalas.

      0
  8. I’d agree with most of your analysis although I’d tweak it a little. For example, I fully agree with the observation that: “The Munda seem to have arrived after 2000 BC, and admixed with another AASI population. This suggests that Northeast South Asia was not touched by the movement of Dravidian peoples that affected the Deccan. Additionally, the extant Munda tribes lack Indo-Aryan ancestry.”

    I think that Cemetery H cremations and some isolated archaeological instances of non-meteoric iron suggest that Indo-Aryans were probably present in far NW India closer to 1900 BCE, even if it took them some time to expand. The 1500 BCE time period may have even been a second wave of Indo-Aryan expansion.

    The sources I’ve seen for the South Asian Neolithic suggest 2500 BCE on the old side and possibly a few centuries after that before it really took hold.

    With regard the comments on a Harappan source for Dravidian, I still find that to be highly unlikely although I do think that the Dravidian languages likely ultimately had their origins in the South Indian Neolithic. (Among other reasons, right up until the demise of the IVC culture ca. 2000 BCE, trade between South India and the IVC was very thin and channelled through a couple of IVC frontier trading posts more or less in the vicinity of modern Mumbai). There is also basically no evidence of use of Harappan seal symbols in the Dravidian area in that time period. But, the case for the contemporary Dravidian language family having a most recent common ancestor 4500 years BP are very weak. I would suggest that the amount of linguistic diversification in the Dravidian language family is closer to that of Germanic, perhaps closer to 3000 years BP than to 4500 years BP.

    This can make sense in light of the analysis of modern South Asian population genetics that show that the last date of steppe admixture is younger in North India than in South India, contrary to any plausible single wave of advance theory, but people with steppe ancestry clear did admix with at least high caste people in Southern India.

    I think a more likely scenario is that a first wave of expansion by people with steppe ancestry and a transition to Indo-Aryan languages reaches all of India except one little last redoubt roughly around the area from modern Hyderabad to the eastern coast of the Deccan Peninsula which is roughly where proto-Dravidian should be if you trace back the linguistic phylogeny to a likely source (the South Indian Neolithic’s epicenter, in contrast, is closer to modern Bangalore), wiping out a diverse family of Dravidian languages in the process. Then, in sort order, the sole Dravidian holdout region retakes most of the Dravidian region from their Indo-Aryan rulers, establishing its sole remaining Dravidian dialect across the region which begins to diversify at that point in lieu of early Sansrkit, and leaving assimilated Indo-Aryan priests in place while casting aside secular Indo-Aryan rulers. This re-established Dravidian region then successfully rebuffs a second wave of Indo-Aryan expansion which impacts only North India. This reconquest would have to take place sometime after the inferred steppe ancestry admixture date in South India, but before the inferred steppe ancestry admixture date in North India (the method used to determine those dates is based towards the most recent wave of admixture when there is more than one wave of it). This would explain why a language family which should have come into being ca. 2500 BCE to 2000 BCE, from a linguistic perspective (particularly if judged by the lower level of linguistic diversity seen much later when the first Dravidian inscriptions are available) would support in addition to explaining the counterintuitive dates of admixture with steppe people in South v. North India.

    0
      1. Isolated Indus Valley symbols have been found in both Tamil Nadu and Kerala. And many places name from Sindh, Gujarat have Dravidian roots.

        1+
        1. I don’t disagree that places like Sindh and Gujarat with Dravidian place names were probably part of a Dravidian linguistic region for some extended period of history. If I were to guess, I would estimate in the period from ca. 2000 BCE (when the South Indian Neolithic has spread to close to its full extent) until about 1100 BCE plus or minus a century or two (when genetic data indicates that steppe ancestry and pre-steppe South Indian ancestry admixed), and that these regions were part of the former Dravidian geographic range that were not reclaimed in a reconquest of its former range sometime not long after 1100 BCE plus or minus a century or two, and almost surely, before 200 BCE (probably at the older end of that range since in the most recent periods in that time range, historical records establish that these regions were not then Dravidian linguistically and it would have been hard for an isolated Dravidian linguistic region to bounce back to most of its old range after many centuries of Indo-Aryan rule). I’d also guess that the reconquest may have been related to the climate event that caused Bronze Age Collapse in the Mediterranean and Europe, since the climate trends in Indian and those regions tend to be in synch with each other.

          “Isolated Indus Valley symbols have been found in both Tamil Nadu and Kerala.” The key word is isolated, when they would have been pervasive if the Harappan language, more accurately called “Meluhha” (because one of the few things we know about it for sure from Sumerian records of an IVC ex-patriot traders colony in southern Mesopotamia, is the proper name of this language) was widely used in the region. Even thin trade between the IVC and the Dravidian region (which is archaeologically attested) would be enough to cause some Indus Valley symbols to be present in Tamil Nadu and Kerala. In the same vein, many modern shipping containers in the U.S. have Chinese characters on them even though the Chinese written language is used by few people in the United States, and even fewer who are employed in the U.S. rail industry that carries those shipping containers. I suspect that there are similarly isolated instances of Chinese characters in evidence in India even outside ex-patriot Chinese communities there.

          0
          1. You realise that Sindh and Gujarat were part of the IVC proper? With Harappan script found in both, like at Dholavira.

            0
      2. “An estimated generation is roughly 29 years.” — Don’t you think the generation time is too high ? Something around 23-25 would be more likely.

        0
        1. The 29 year estimate has been empirically validated multiple times.

          Roughly speaking, average generation length corresponds to the median age at which you give birth to your median surviving child. It is not, for example, the average age at which a woman typically has her first child.

          In pre-historic times, a typical woman would have an average of about 8 children per lifetime (similar to the least developed countries today and pre-industrial fertility patterns in places that are now industrialized, like the Colonial era of the U.S.) of whom 2-3 would survive to have children of their own (otherwise, populations would have grown much faster).

          The onset of puberty was a few years later than it is today, in part due to inferior diets to those we have today. If a typical woman married on her sixteenth birthday (half up to a year or two earlier, half up to a year or two later, on average), got pregnant while still age 16 (it takes an average of six months without contraception), had her first child at age 17 and had her last child at age 42 (i.e. roughly her entire period of biological fertility in that era) with 8 children per lifetime implying spacing of about 40 months between children (which would be extended compared to what is possible today due to the imperfect contraceptive effect of extended breast feeding at a time when formula didn’t exist and baby food didn’t come in jars from factories), and birth order didn’t strongly impact a median child’s chance of surviving to adulthood, then you get a 29 year generation. An average age at first childbirth just a year or two older and people who have more than the average of eight kids can compensate for deaths in childbirth before menopause.

          If your generation length were 24 years, and the average women first got pregnant at age 16 had her first child at age 17, she’d have to stop having children on average by age 31 without using contraception, and to fit in 8 children, would have to have a child more often than once every 21 months while breast feeding exclusively. Twenty-four years is low mostly because women in pre-modern times kept having children until they hit menopause (typically in the 40s) or died, and, to a lesser extent, because the birth spacing is too tight.

          2+
    1. If the IVC had no Dravidian connection, then how do you explain the Brahui language and the fact that the Brahui speakers are genetically similar to their Balochi-speaking neighbors (which is what I heard in the podcast)?

      5+
      1. “the Brahui speakers are genetically similar to their Balochi-speaking neighbors” This is accurate.

        The Brahui language, based on linguistic data, very likely diverged from other Dravidian languages ca. 1000 CE or later. As the relevant Wikipedia page explains: “it has been argued that the absence of any Old Iranian (Avestan) loanwords in Brahui suggests that the Brahui migrated to Balochistan from central India less than 1,000 years ago. The main Iranian contributor to Brahui vocabulary, Balochi, is a western Iranian language like Kurdish, and arrived in the area from the west only around 1,000 AD.[76] Sound changes shared with Kurukh and Malto also suggest that Brahui was originally spoken near them in central India.[77]
        [76] Elfenbein, Josef (1987). “A periplus of the ‘Brahui problem'”. Studia Iranica. 16 (2): 215–233. doi:10.2143/SI.16.2.2014604.
        [77] Krishnamurti (2003), pp. 27, 142.”

        Brahui speakers are genetically similar to their Balochi-speaking neighbors because they acquired their language primarily via culturally driven language shift from an elite that migrated there, rather than mostly via migration. Brahui folklore also recounts a migration from afar to its current location.

        FWIW, the probable time of the migration of Brahui elites to Balochistan followed by language shift in situ, is about the same time as the time that the European Roma commenced their migration from India to Europe.

        Put another way, Balochi is an example of pre-European colonialism from a remote location in the region. This means, of course, that the historical range of Dravidian speakers did not extend nearly that far, although evidence from Dravidian toponyms suggests that some of the very Southernmost part of what is today part of Pakistan was linguistically Dravidian (the IVC culture was mostly further north in the Indus River Basin).

        A similar case of probable language shift in a North Dravidian population (the Gonds), which may have been roughly contemporaneous, also exists, although in that case, the original language of the Gonds was probably a Munda language. See Gyaneshwer Chaubey, et al., “Reconstructing the population history of the largest tribe of India: the Dravidian speaking Gond. ” (2017).

        Another, contemporaneous, famous case of language shift without significant population genetic change that is well documented historically is the transition from an Indo-European language to the Hungarian language which is part of the Uralic language family in Hungary ca. 900-1100 CE after it was conquered by the Magyars. Less famously, and also contemporaneously, just 10,000 Normans who migrated to England ca. 1066 CE constituting less than 0.5% of the population of England at the time are the source of almost all of the numerous French origin words in the English language and the transition from Old English to Middle English in England.

        The Yiddish language came into being in the 800s CE in Central Europe as Ashkenazi Jews synthesized a new language that none of the members of the community had spoken before from High German, Hebrew, Aramaic, and Slavic languages. And, no one on Earth was a native speaker of Hebrew at the time that it was resurrected as a living language in Israel in the late 19th and early 20th centuries.

        We don’t know precisely why this language shift occurred at this time because the historical record is shorter in Balochistan than it is in Hungary or England, but two of the most common drivers of this kind of language shift historically have been mass religious conversion and secular rule by outsiders (or both).

        1+
        1. Highly unlikely hypothesis that a Dravidian elite from central india migrated and conquered a isolated part of Baluchistan in the post medieval/islamic era. There is no convincing evidence of this. The whole argument based on the lack of avestan words is also tenuous. The genetic evidence suggests that Brahui is a relic and that there is no substantial difference between them and their aryan speaking neighbours.

          0
          1. if this weak Dravidian elite dominance language replacement theory was correct we would expect to see a spike in AASI ancestry amongst the Brahui compared to their Baluchi neighbours. This is not the case. It is much more likely that the Baluchi used to speak prehistoric Brahui.

            0
          2. The linguistic evidence runs both ways. Not only are their loan words from later but not earlier, there are also shared, late arising linguistic features.

            The fact that the same kind of long distance conquest and language shift took place almost simultaneously at the opposite corners of Europe also suggests that it is not unlikely at all. The fact that there is another contemporaneous example of language shift to Dravidian at or beyond the frontier of the core Dravidian territory at about the same time in the Gonds (there from Munda) again is a very definitively example that this is anything but unlikely.

            The genetic evidence suggests with overwhelming force that Brahui is not a relic, because if it was, its populations genetic composition would be very different from its neighbors, and would be different in the direction of the population genetics of all other Dravidian populations. Relic populations are almost always very different from their non-relict neighbors see, e.g., the Hazda, the San, the Andamanese, the Kalash, Greek speaking enclaves on the Black Sea coast of Turkey, the Sardinians, the Druze, the Ainu. Jewish populations in Yemen and Africa have significantly different population genetics than neighboring populations even when they aren’t visibly different. Relict zoroastrian populations in Iran are very genetically distinct from ordinary Iranians. The nearly identical population genetics to neighboring communities which overwhelmingly speak another language is the classic and uniquely distinguishing sign of language shift due to elite dominance.

            There is also a lack of any evidence of Dravidian occupation between Balochistan and even the most remote outposts of Dravidian toponyms or distinguishing archaeological relics. Brahui found where it is bears some similarity to people speaking an Austronesian dialect closely related to that of Borneo in Madagascar (this migration is earlier than the proposed Brahui one but still within 1000 years), although the genetics there tell us that the source populations were crudely half Bantuish and half Austronesian with the Africans experiencing language shift. The hypothetical Brahui colony would have been founded within a few hundred years of the Crusader states in the Levant, again showing that long distance language transmission and conquest was possible in that time frame. But, go back further in time and this kind of thing is much less common.

            0
  9. That said, I am not a fan of genetics outside medicine and I am underwhelmed by it.

    if it wasn’t for genetics we’d still be having stupid about “out of india theory.” though perhaps that’s a feature not a bug for you …. 😉

    5+
      1. No. Linguistic and genetic evidence are really quite convergent.

        I find them especially convergent in the more neglected cultures on the E/NE side of India. E.g. Axom (a Thai group) show clear non-Indic (East Asian) genetic ancestry and their language, though Indo-Aryan, completely lacks retroflexion.

        Retroflexion is the linguistic analogue of the genetic smoking-gun of bonafide South-Asianness. And the Axom lack it thoroughly. Not even a teeny weeny bit.

        ~

        OIT is basically bonkers.

        4+
        1. “Retroflexion is the linguistic analogue of the genetic smoking-gun of bonafide South-Asianness. ”

          You do know that Retroflextion is abundant in Pashto, and scanty in Eastern Hindi, Bhojpuri etc, right? How do you explain this anomaly?

          0
  10. Thanks for answers. More on health perspectives I think will be the future. Also, you just need to learn to pronounce “bharat” properly.

    0
  11. @ohwilleke

    // Harappan source for Dravidian, I still find that to be highly unlikely //

    // Dravidian languages likely ultimately had their origins in the South Indian Neolithic. //

    // right up until the demise of the IVC culture ca. 2000 BCE, trade between South India and the IVC was very thin //

    The above sentences make zero sense to me.

    How do you reconcile the above with the ubiquity of retroflexion – a key phonetic feature of Dravidian – across S Asia and nearby zones on the Iranian plateau? E.g. retroflexion in Pakhto spoken in Kandahar (Arachosia) or Khotan Saka?

    Retroflexion is certainly not a super-strate in Dravidian but a deep and defining feature of this language group. And it is a substrate in both IA and E Iranic and certainly not present in PIE.

    So if the Dravidians are really a peninsular Indian phenomenon and had little trade contact with IVC (speaking an entirely extinct language) how did this defining feature of sub-continental speech occur in both places? Coincidence?

    And if you are going to argue that IVC invaded S India, then retroflexes should be the super-stratum in tamiL and kannaDa. Which they are not.

    Dravidian also has lent a substantial substrate lexicon in the earliest Vedic speech spoken in N Punjab. And the transition from Old Vedic to early and middle prkRta-s also shows clear signs of Dravidian language throughout. E.g. endemic retroflexion, diphthongs > long vowels, simplification of consonant clusters esp with voiced sibilants etc. How did that happen so far up in the North, so early on?

    A simple application of Occam’s Razor shows that Dravidian language family is clearly the most serious contender of IVC speech and almost certainly the linguistic group that was common to a lot of people from Punjab to Karnataka on the Western rim of the Indian subcontinent. It is they who shaped IA in their image.

    6+
    1. Hello Slapstik,

      This is not directly related to the matter at hand but I have been itching so hard to know this about the development of New Indo-Aryan languages and the Dravidian-like grammatical features in them, even in such northeastern (this type of a geographical location might point to Telugu or a Telugu-like language as the influencer) languages like Hindi and Bengali (I read somewhere that there are certain Dravidian-like particles or something in Bengali but I don’t know much further). The Hindi dative case marker ko, for example, clearly seems to be like a result of Dravidian influence, doesn’t it? When did this sort of Dravidian influence creep into the ancestor of Hindi and how? As another question, is there a possibility the earliest ancestors of Apabhramshas, by which I mean spoken Indo-Aryan languages, date to much earlier than usual, existing in parallel with the Prakrits or did they develop according to the usual chronology, in your opinion?

      0
      1. “The Hindi dative case marker ko, for example, clearly seems to be like a result of Dravidian influence, doesn’t it? When did this sort of Dravidian influence creep into the ancestor of Hindi and how?”

        “Mayrhofer identified a “prefixing” language as the source of many non-Indo-European words in the Rigveda, based on recurring prefixes like ka- or ki-, that have been compared by Michael Witzel to the Munda prefix k- for designation of persons, and the plural prefix ki seen in Khasi, though he notes that in Vedic, k- also applies to items merely connected with humans and animals.”
        From https://en.wikipedia.org/wiki/Substrata_in_the_Vedic_language

        This is important for a couple of reasons.

        First, prefixing languages appear in only geographically limited areas (mostly sub-Saharan Africa and island Southeast Asia, with isolated examples in the Americas and in the Ket language of Siberia with is related to the Na-Dene languages of pre-Columbian North America).

        Second, languages which categorize words by categories such as “animate” are relatively uncommon. In some places these systems are called “genders” and in other places these systems are called “noun cases” but they seem to represent the same phenomena.

        Most of the most naively plausible macro-linguistic family relations of Meluhha, which is what the Harappans called their language, are not prefixing and do not have a gender/noun case system that includes candidates like “animate” which are found in some African languages, some South Asian languages, and some Australian languages.

        I suspect that the presence of this prefix in Munda is likely to be either an adstratum effect of Meluhha on Munda, or a substratum effect of a hunter-gatherer language of India, than visa versa, as this feature seems to be absent in other Austro-Asiatic languages from which Munda is primarily derived.

        0
        1. Hello ohwilleke,

          The Hindi ko particle that I’m talking about is a suffix and not a prefix and it is so extraordinarily similar in form and function to, say Telugu ku, which also functions as a suffix and does the same exact function that its Hindi counterpart does: act as the dative case marker. Telugu dative case suffix descends from Proto-Dravidian *kku which is robustly reconstructible to Proto-Dravidian (dative case is one of the two/three few cases that are reconstructible to all the subgroups within Dravidian and thus going back to Proto-Dravidian and many other case markers arose later on in the individual subgroups and languages with postposition-origins) so it goes back as a unit encoding necessary grammatical information in the Proto-Dravidian (and likely Pre-Dravidian) stage.

          1+
          1. “The Hindi ko particle that I’m talking about is a suffix and not a prefix and it is so extraordinarily similar in form and function to, say Telugu ku, which also functions as a suffix and does the same exact function that its Hindi counterpart does: act as the dative case marker.”

            Given the linguistic indications that Meluhha was pre-fixing and not suffixing, it seems likely that neither of those grammatical features were derived from it, and that Dravidian and Meluhha were parts of different language families from each other.

            The evolution of the oldest version of Sanskrit language in South Asia into the many Indo-Aryan daughter languages such as Hindi took many centuries, during which there was widespread Indo-Aryan exposure to Dravidian languages and there were probably lots of language learners whose first language was a Dravidian one in some place other than the old IVC territory.

            I wouldn’t be at all surprised if the suffix feature you describe is a Dravidian substrate in many of the post-Sanskrit daughter languages of Sanskrit.

            0
    2. “// right up until the demise of the IVC culture ca. 2000 BCE, trade between South India and the IVC was very thin //

      The above sentences make zero sense to me.”

      The first two sentences are linguistic hypotheses. The last one, however, is archaeologically well established fact.

      It is also notable that the package of crops and technologies associated with the South Indian Neolithic is very different from that of the IVC. Many of the crops were domesticated in the African Sahel, which is also the place of origin of many South Indian Neolithic construction techniques, and there are other cultural features like mythology motifs, musical instruments and games found in the Dravidian linguistic area that have strong overlap with the African Sahel area where those crops were domesticated. Usually, the kind of cultural transfer associated with mass migration or conquest is not so selective and does not include so many foreign elements that the source culture did not adopt itself.

      “How do you reconcile the above with the ubiquity of retroflexion – a key phonetic feature of Dravidian – across S Asia and nearby zones on the Iranian plateau? E.g. retroflexion in Pakhto spoken in Kandahar (Arachosia) or Khotan Saka?”

      As you note, this feature is found in an area much wider than Dravidian and includes language isolates that are neither Indo-Aryan nor Dravidian. “Retroflex phonemes are now found throughout the Burushaski, Nuristani, Dravidian and Munda families. . . . They are not reconstructible for either Proto-Indo-European or Proto-Indo-Iranian, and they are also not found in Mitanni–Indo-Aryan loanwords. . . .There is a clear predominance of retroflexion in the Northwest (Nuristani, Dardic, Khotanese Saka, Burushaski), involving affricates, sibilants and even vowels (in Kalasha), compared to other parts of the subcontinent. . . . ” from https://en.wikipedia.org/wiki/Substrata_in_the_Vedic_language The presence of retroflex phonemes in Munda languages which population genetics suggest received little South Indian migration, and the fact that they are less common in the Dravidian linguistic area of India than in Northwest India (and beyond into Central Asia) also argues against a Dravidian source for these phonemes.

      It is centered in the areas where Meluhha, which is what the Harappans called their language, was spoken, suggesting that these phonemes were more widely used in this region. The Harappan civilization was initially predominantly in the Indus River Valley and now extinct Sarasvati River basin. When the Sarasvati River declined, the people who lived there migrated to the Ganges River basin. This is consistent with the Northwestern subcontinent scope of this feature. Harappan civilization didn’t really extend as far south as the region where Dravidian toponyms are found.

      In my view, the most likely possibility is that this was a pervasive areal feature of the hunter-gatherer languages of both West Asia (mostly Iran) an South Asia prior to the arrival of Indo-European (ca. 1900 BCE to 1500 BCE), prior to the likely emergence of Dravidian (ca. 2500 BCE to 2000 BCE), prior to the likely arrival of the Munda languages in South Asia (ca. 2000 BCE), and prior to the arrival of Iranian farmers who founded the IVC culture (ca. 6000 BCE). The Harappans, with more leisure and a stable unified society with few outside adult language learners then could have elaborated on this phonetic feature over the years to a greater extent than hunter-gatherers had.

      From sometime prior to the Toba eruption (ca. 75,000 years BP) until the Neolithic Revolution in the Fertile Crescent (ca. 8000 BCE), India had the largest absolute population of in the world outside of Africa, and was rivaled in population density only by fishing communities in places like Japan, the Pacific Northwest, and the Baltic Sea area.

      Population genetically, in deep prehistory, it looks like there was a very deep divide between the Levant on one hand and Mesopotamia and points east of it on the other (based on ancient DNA data), which faded less abruptly into a population genetic region in South Asia that blended somewhat on either side of the Indus River Valley (based on analysis of modern population genetics). I don’t entirely rule out the possibility that Dravidian borrowed some phonemes or traits from Meluhha, but the absence of words in the earliest Rig Vedic substrate of Sanskrit from Dravidian disfavors the theory that Dravidian and Meluhha had a lot of vocabulary in common.

      You see something similar in the Bantu languages of Madagascar which retain click phonemes from the substrate hunter-gatherer languages of the region that were not part of the Bantu source languages that replaced those languages.

      Similarly, there are phonemes that are areal features found in several distinct language families that are geographically close to each other in the Caucasus mountains region, and tonality is an areal feature found in several different language families in both Southeast Asia and China (especially southern China), and again in parts of Easternish Africa, probably also due to substrate influences from hunter-gatherer linguistic regions that pre-date the expansion of the languages of the first farmers in various places.

      “And if you are going to argue that IVC invaded S India, then retroflexes should be the super-stratum in tamiL and kannaDa. Which they are not.”

      I do not argue this. I argue that post-IVC Indo-Aryans who spoke Sanskrit did this and that their period of occupation in the current core geographic range of the Dravidian language was brief and didn’t have much of a linguistic impact. Retroflexes were surely a pan-South Asian substrate before Indo-Aryan or Munda languages arose and before the most recent common ancestor of the modern Dravidian languages.

      “Dravidian also has lent a substantial substrate lexicon in the earliest Vedic speech spoken in N Punjab. And the transition from Old Vedic to early and middle prkRta-s also shows clear signs of Dravidian language throughout.”

      Michael Witzel has described the linguistic substrate in Rig Vedic Sanskrit evidenced by about 300 words. Dravidian loan words are absent from the oldest Rig Vedic Sanskrit and only emerges in parts of the Rig Veda believed to have been composed later. In the later period, the Indo-Aryans had expanded their influence into Dravidian territory and were able to maintain much longer distance networks of people and travel because they introduced domesticated horses and wheeled vehicles into the transportation mix. From the same link as above: “Kuiper identified 383 specifically Rigvedic words as non-Indo-Aryan — roughly 4% of its vocabulary. Oberlies prefers to consider 344–358 “secure” non-Indo-European words in the Rigveda. Even if all local non-Indo-Aryan names of persons and places are subtracted from Kuiper’s list, that still leaves some 211–250 “foreign” words, around 2% of the total vocabulary of the Rigveda. . . . These loanwords cover local flora and fauna, agriculture and artisanship, terms of toilette, clothing and household. Dancing and music are particularly prominent, and there are some items of religion and beliefs. They only reflect village life, and not the intricate civilization of the Indus cities, befitting a post-Harappan time frame. In particular, Indo-Aryan words for plants stem in large part from other language families, especially from the now-lost substrate languages. . . . There are an estimated thirty to forty Dravidian loanwords in Vedic. However, Witzel finds Dravidian loans only from the middle Rigvedic period, suggesting that linguistic contact between Indo-Aryan and Dravidian speakers only occurred as the Indo-Aryans expanded well into and beyond the Punjab. . . . ”

      This suggests that Meluhha and not Dravidian was the source of most of this substrate, and that Sanskrit’s arrival probably post-dated the collapse of Harappan civilization rather than bringing it about, although this is complicated somewhat by the fact that the Rig Veda has many references to the Sarasvati River basin which had run dry by the time that the Indo-Aryans arrived. Hence, the early Rig Veda probably incorporated oral history composed in Meluhha and preserved by a Harappan diaspora in the Ganges area living in rural villages that was translated into early Sanskrit prior to significant Indo-Aryan contact with Dravidian language speakers.

      Witzel’s initial and probably inaccurate label for this substrate as Para-Mundi probably reflects shared substrate influences in both Meluhha, which is what the Harappans called their language, and the Munda languages.

      1+
    3. I wrote a lengthy answer to this that disappeared, possibly to a spam filter. To summarize it:

      // right up until the demise of the IVC culture ca. 2000 BCE, trade between South India and the IVC was very thin //

      This is an archaeological fact, even if the other two points are linguistic hypotheses.

      “How do you reconcile the above with the ubiquity of retroflexion – a key phonetic feature of Dravidian – across S Asia and nearby zones on the Iranian plateau?”

      First, there is actually more retroflexion in NW India than in Dravidian areas, and this is an areal feature also found in Munda languages, Indo-Aryan languages, and nearby Indo-Iranian languages.

      My best guess is that it was a shared feature of the hunter-gatherer languages of South Asia which developed from pre-75,000 years BP to 6000 BCE in the IVC area to 2500 BCE in South India (with the largest hunter-gatherer population in the world for most of that time period), which the IVC with more leisure and not much war until the very end, had more of an opportunity to elaborate. Also, languages tend to get more complex over time in all respects, including phonetics, the longer they go with minimal simplification arising from large numbers of adult language learners.

      The regional phonetic distinction could alternatively have been an adstratem influence on Dravidian and Munda from Meluhha, and a substratem influence on Indo-Iranian languages in the area, although I think that the adstratem hypothesis is less likely.

      Areal influences that don’t flow from a language’s ancestry are common in phonetics. Mozambique has click consonants in its Bantu dialects from pre-Bantu hunter-gatherers. The several languages of the Caucasian mountains has phonetic elements that are shared and not found elsewhere. Tonal languages are common across the several language families of SE Asia and China on one hand, and among several language families of Easternish Africa on the other other. Shared phonetics aren’t a strong indicator of common linguistic ancestry or descent. This is probably because phonetics persist over time more strongly than lexical features or grammar.

      “And if you are going to argue that IVC invaded S India, then retroflexes should be the super-stratum in tamiL and kannaDa.”

      I don’t argue that. I argue that Indo-Aryans did that after Meluhha was a dead language.

      “Dravidian also has lent a substantial substrate lexicon in the earliest Vedic speech spoken in N Punjab.”

      Not so. According to Witzel and a couple of other leading linguistics there is a non-Indo-European, non-Dravidian substrate of about 300 words mostly associated with rural village lie in the Rig Veda. There are 30-40 Dravidian loan words, none of which appear in the earliest Rig Vedic materials, they appear in the middle Rig Vedic materials at a time when there was probably Indo-Aryan communication across the continent into Dravidian areas driven by improved regional transportation networks made possible with domesticated horses and wheeled vehicles. The absence of 90% of the substrate words in Rig Vedic Sanskrit from Dravidian languages argues against Dravidian being derived from Meluhha, even though it had some areal similarities. It is also possible that some of the Dravidian words in the Rig Vedic substrate are actually Meluhha words borrowed by Dravidian languages which belong to a different language family than Meluhha even though it borrowed a few words from it.

      1+
    4. Haha! I absolutely adore how Slapstik likes his own comments sometimes. Most of the ones he likes himself appear to me too to be falling under his best-of-the-best category. I’m gonna give this very comment of mine a like too, imitating the wonderful Slapstik!

      1+
    5. Haha! I absolutely adore how Slapstik likes his own comments sometimes. Most of the ones he likes himself appear to me too to be falling under his best-of-the-best category. I’m gonna give this very comment of mine a like too, imitating the wonderful Slapstik!

      2+
  12. I have not listened to the podcast, but am in general agreement with everything here; so glad that we do not have a large OIT contingent of commentators. The only wish is to restate the discussion in terms of INPE-AASI admixture and INPE-Steppe_MLBA admixture events. However, that method still does not resolve the broad spread of R1A through India, where as ANI-ASI admixture events can explain better.

    The one comment is that, if it is accepted that the InPE people were the originators of the Dravidian languages, then they have more than 4500 years living in the periphery until the IVC fell. There is sufficient time to work through the proto-dravidan language. I am also sympathetic to the view of Eastern Iranian farmer continuum to INPE (~17K BP), and then into South India after 4000 BP.

    Another part (beyond genetics, history, language) that is ill understood is Indian geology. The last glacial maximum and the vegetation in India (tropical grassland vis-a-vis tropical desert in Iran) would have allowed for population flow into India even before 8500 BP.

    What is the title of Tony Joseph book?

    1+
  13. For those bold enough to wade into geology and climatology, read through an example of the geography of Thar desert between quaternary and holocene period in e.g.,
    “CLIMATE VARIATION IN THE THAR DESERT SINCE THE LAST GLACIAL MAXIMUM AND EVALUATION OF THE INDIAN MONSOON”

    “We evaluated the hydrological conditions with respect to strength (i.e., amount and geographic coverage) of the southwest summer monsoon since the last glacial maximum (LGM). Between LGM and c.15cal. ka BP, the eastern basins hosted saline and hypersaline playa lakes and the western part had an intermittent variable lake. A shift from saline-hypersaline playa lakes to perennial deep lakes occurred in the eastern margin at c.15cal. ka BP as more summer insolation increased sea surface temperature (SST) of the Indian Ocean and strengthened the southwest summer monsoon. During the Pleistocene-Holocene transition, the highest summer insolation as well as warmer SST of the Indian Ocean increased the amount of summer precipitation and expanded the southwest monsoon over the entire desert. However, more winter precipitation and minimal summer rainfall maintained perennial lakes across the desert during the early and middle Holocene(between 11.7 kya to 4.2K BP). Over the middle-late Holocene(8.2K to 4.2K), the regional arid conditions were contemporary to intervals of reduced summer insolation, southerly located Inter-Tropical Convergence Zone and frequent El-Niño Southern Oscillation.”

    The 4.2 K event started a 200 year drought that signaled the end of IVC and start of the Meghalayan period. This age started with a “abrupt and critical mega-drought and cooling”, in a cave in Meghalaya. This “climatic event affected agricultural societies, forcing the collapse of civilisations and migrations and regenerations in Egypt, Greece, Syria, Palestine, Mesopotamia, the Indus Valley, and the Yangtze River Valley”. This is a significant step in the formation of the INPE-AASI admixture.

    A lot of other critical events happen in Kerala, Bengal and Srilanka forming their present green status and large populations after the Meghalayan period begins. My point here is geology and climatology plays a huge role.

    3+
  14. The idea that South Dravidian-I and South Dravidian-II share a slightly different physical ancestor than Central and North Dravidian (or the possibility where all the three subgroups share different geographical ancestors) is quite novel to Dravidian linguistics, it appears. I don’t know how they attempt to eliminate such a possibility but I can perhaps point to some things both in favour of and against the above idea.

    In favour, I think the curious relationship between Central Dravidian and the Telugu-Kui (aka South Dravidian-II but I will talk more about the classification of Telugu-Kui later*) group can be cited. The geographical distribution of today’s Central Dravidian and Telugu-Kui seems to indicate that a Telugu-Kui-speaking population interpenetrated previous Central-Dravidian-speaking territory as there are islands of Central Dravidian within a broader Gondi-Kui territory in central India today. There are also certain close linguistic relationships between Telugu-Kui and Central Dravidian as well; some close lexical affinities exist between Central Dravidian and Gondi-Kui today and some curious grammatical affinities exist between Central Dravidian and Telugu. The most important one is the Modern Telugu numeral for ‘one’ in isolation which is okaTi, just like in Central Dravidian, as opposed to the oblique form oNTi which like the other dominant languages of Tamil and Kannada is still oNNu-based. Does this single feature then point to a possible Central Dravidian substrate in Telugu? Probably not. It might have just as well resulted from some commonplace ancient-contact phenomenon.

    Also, importantly, note that even all the discussion in the above paragraph does not imply that Central Dravidian and South Dravidian have slightly different physical ancestor languages. It might be the case that a previously more widespread Central Dravidian territory was subject to an invasion by the Proto-Telugu-Kui people who might have developed some technological advantage (like iron**) or something after separating from the Central Dravidian people with which they shared a same physical ancestor.

    Against the idea I can’t suggest much except that Bhadriraju Krishnamurti’s textbook seems to present to me a picture that suggests there to have been a very tight relationship between the ancestral dialect groups of today’s North Dravidian, Central Dravidian and South Dravidian. Krishnamurti also suggested in the book that the breaking up of Proto-Dravidian into North Dravidian first might have begun to happen sometime around 1500 BC at the earliest. This may indicate that the currently considered Proto-Dravidian was indeed quite a physical language spoken in a certain geographical territory, wherever it was.

    * Four subgroups within Dravidian, namely, Tamil-Tulu, Telugu-Kui, Kolami-Naiki-Parji and Kurukh-Malto(-Brahui) are well-established. Some authors believe Telugu-Kui and Tamil-Tulu share a recent common ancestor after Proto-Dravidian, called Proto-South-Dravidian and accordingly label Tamil-Tulu as South Dravidian-I and Telugu-Kui as South Dravidian-II. A few others believe Telugu-Kui and Central Dravidian share a common ancestor in Proto-Central-Dravidian. The dominant view at the present is the first one; at least Bh. Krishnamurti believed in that.

    **- Iron may actually be one another possible indication that may point to a slightly different parentage of Central Dravidian and North Dravidian on the one hand and South Dravidian on the other. A Dravidian-origin word for iron is present in South Dravidian-I and Telugu-Kui (several dominant languages within the Gondi-Kui subgroup of Telugu-Kui group retain this word) and absent in Central Dravidian and North Dravidian. But then, many lexical items are not retained currently in North Dravidian and Central Dravidian as well. How reasonable is it to read too much into the fact of absence of things in CDr and NDr? Also, I wonder what Mr Razib Khan thinks about the seemingly early beginnings of iron in south India (1200 BC at Hallur) and the presence of a Dravidian-origin word for ‘iron’ in all the biggy Dravidian languages (also including many non-literary languages of the Tamil-Tulu and Telugu-Kui subgroups), given that the IVC did not seem to have developed iron and they had bronze which the Reddy-Kodava-Vellalar et. al. did not seem to have carried with them to south India. This is one of those other conundrums that many on Quora fight over too. Haha!

    Thanks to all very much!

    2+
    1. My only response is to Iron. The beginning of Iron age, both, in Crete, Aegean, Anatolia, Greece, Europe and South India is 1300 BC or 3000 BP. There could have been no iron in IVC which started falling apart between 4200 BP and 3800 BP, as there was no nation that has gone past bronze. It should also not be suggested as a trademark of Dravidan culture; as ANI admixture started as early as 4200 year BP to 1900 BP, granted there is a shortage of sites in North India. The chances of iron coming from Anatolia, Greece or Steppe are all equally strong. This is similar to spread of crops; a whole lot of dryland crops from Africa and wetland crops from places as far as New Guinea spread into Indiaafter farming was established. Iron says something, but not a feature of Dravidan.

      3+
    2. A small correction:

      The Central Dravidian Kolami seems to have the ‘iron’ cognate (but there is a possibility that it is not a true cognate but some early borrowing from Telugu or something), so it is only the North Dravidian subgroup that does not have any cognates.

      0
    3. Re Iron: It isn’t uncommon for trade goods to carry descriptive nouns for them with them that are unrelated to any roots in the languages that borrow these words. For example, “chai” for tea is found in many different language families but has its origins in Austronesian languages of the place that tea was originally imported from. A similar origin of a word for iron wouldn’t be surprising.

      0
      1. Hello ohwilleke,

        I don’t fully understand what you have suggested above but the Dravidian ‘iron’ word is not a Wanderwort like the ‘tea’ word or ‘rice’ word in that it is not borrowed from any source and has well-explained internal etymology (Tamil irumpu [pronounced irumbu], Telugu inumu, etc. from Proto-Dravidian/Proto-South-Dravidian *cir-umpu, ‘dark thing’, ‘black thing’ connected to the root *cir- meaning ‘dark’/’black’. The common Telugu word for ‘darkness’, cIkaTi also is connected to the root *cir.) that adheres properly to all the known sound laws and word formation strategies of Dravidian. But that said, I do agree that it may have been some kind of descriptive coinage from native Dravidian roots on adoption of the concerned ‘iron’ technology from any external source; it may be the case otherwise too.

        0
        1. “the Dravidian ‘iron’ word is not a Wanderwort like the ‘tea’ word or ‘rice’ word in that it is not borrowed from any source and has well-explained internal etymology”

          Fair enough. Your explanation sounds quite compelling.

          0
    4. And we know that the epicenter of iron ore extraction, even to this day on the Karnataka-Rayalseema-Telangana border corresponds to the spread of neolithic ashmound sites around the lower Tungabhadra basin. Metallurgy would have been so central to this culture that later during the middle ages steel ingots from there were in demand in distant international markets. The culture was markedly agro-pastoral according to some, the added emphasis on pastoral hinting that it may have been a compound population. Were pastoral dravidian colonists in the deccan living symbiotically with advancing agriculturalists? Rainfed agriculture makes the deccan risky territory for settlers and may have needed a risk mitigating practice like sheep rearing in the scrubland of that landscape.

      0
      1. Some answers to this question are given in the book “recent advances in Indo-pacific prehistory”, particlyalry chapter 51, “early settlements and subsistence patterns in Deccan”.
        The late harrappans settled into Jorwe site in about 1800-1600 BC. The first Deccan farmers surprisingly cultivated wheat and Barley, not Sorgum and horsegram. Given this being Deccan, and the lack of iron plough means they were not wildly successful and transitioned later to millet and horsegram. But even here iron and ploughs and diversionary channels arrived as early as 1300 BC. Whule the identification of iron ore belt as the beginning of iron age is important, iron went north and south all the way up Maharashtra and south in Andhra. Peter Bellwood, Dorian Fuller and others have published extensively on the first farmers in IVC and Deccan, and do notice attempts to use iron as a plough, but ploughs and irrigation channels are the first use for iron.

        1+
      2. Yes girmit, there seems to be overall quite a bit of cultural continuity within all the phases of the Southern Neolithic, beginning from the ashmounds in 2800 BC down to the megalithic period and Early Historic period with not a great deal of disruptions but then the Dravidian languages on the other hand seem to date to a much younger date than at least the earlier phases of the Southern Neolithic. Even genetics seems to indicate that the Dravidian languages might have been brought to south India by an actual physical post-urban-collapse Indus population. Perhaps archaeological stuff is not very reliable in assessing the migration thingies because in some cases, invaders may not disrupt existing culture too much. So it’s very interesting to me – when you say advancing agriculturists which I’m inclined to consider are the Kapu-Kamma-Reddy-Okkaliga-Vellalar-etc.-type people (please correct me if I understood you wrongly), what language family do you attach to them originally? Something other than Dravidian, like Indo-Aryan or some such family? If Dravidian, some other qualitatively different Dravidian dialects than that of the surrounding early pastoralists? Thank you very much!

        0
        1. Santosh, I’m just speculating quite randomly on the possibility that the languages of the neolithic farmers and/or IVC were neither IA or proto-Dravidian. It wouldn’t surprise me if an entire language family became extinct e.g. Elamite,Sumerian and Hurrian. Does any modern language group represent the near east farmer expansions? If not why would we expect the IVC to have a living descendant?
          Is it possible that minority languages usurp majority languages in pre-literate cultures when they have a more spiritually seductive bardic tradition? Could the kurubas or gollas, as lingam worshiping proto-shaivas, been the the cultural elite for long enough to see a language shift?

          1+
          1. Currently, there is even no evidence that IVC had a written language, or that language is Dravidian. All of Parpola or Witzel is extended speculation. Thus identification of Elamite with IVC is as good a guess, except that old Elamite records are scarce and the middle Elamite Cueneiforms do not resemble IVC signage (you can look at all of the Happa signage at harappa.com in Facebook).

            There are books on this subject e.g., “The Proto-Elamite Settlement and Its Neighbors: Tepe Yaya Period IVC “; that lead to zero conclusion.

            Today, Omer Khan maintains a facebook page on harappa.com that has discussions on this subject. A video on this site with DrTayyab Singhanvi and DrAshutosh Gautam talk about the Iranian city of Jiroft which is considered early Elamite .

            0
          2. Yes girmit, as I remarked elsewhere once, it would be one of the funniest situations in the world if the Meluhhans (ancestors of the Reddy et.al. batch) lost their native language both in north India and south India. Nothing serious or offensive against anyone; just saying in jest. The situation you talk about is definitely possible given that the biggest parallel of pastoralists causing language shift in (situationally weak) agriculturists exists right in the north of India. But maybe there is something in the nature (cruelty, historical feudal ruthlessness and things like that perhaps) of the currently powerful agriculturalist south Indian non-Brahmin castes (at least the Telugu country is filled to the brim with their political dominance at least) that makes people don’t even consider such a possibility seriously? My currently rather dormant and saturated interest in this entire topic would definitely soar in a renewed manner if any indication surfaces that the situation you suggest might be true.

            On a related note, I’m totally on-board with the idea that pastoralist bardic traditions are spiritually super-seductive. Pastoralists have that something having to do with simplicity or whatever – I don’t know – which is quite magnetic. (No surprise Krishna is the ultimate attractor of humankind.) Their cultures also tended to have a better morality in the ancient world perhaps; we see early pastoralist Hebrew’s-ancestor-Canaanite-dialect-speaking Jewish people trying as hard as they could to reform certain evil social practices of the Near East of their day and there is also a very significant possibility, as bharat comments in this massive thread somewhere above, that the Indo-Aryans helped abolish or reduce at least, some evil practices in India at the time such as human sacrifice.

            0
  15. 1) Who were Aryans – the nameless, advanced civilization?
    2) What was their language before arrival to SA?
    3) Which language was spoken in NE India before Aryans arrival?
    4) Have they brought Sanskrit to India?
    5) Have they brought Rig Veda?
    6) Have they brought toponyms from old homeland? Any?
    7) Have they brought R1a to SA?
    8) Who are the ancestors for 100+millions (R1a) of SA/Indians?
    9) Have Aryans ruled Sri Lanka in one period?
    10) Did they have their state/territory in SA and what was its name?

    0
    1. “1) Who were Aryans – the nameless, advanced civilization?”

      One branch of people descended from steppe herders of the Pontic-Caspian steppes. The Proto-Indo-Europeans as they are known were probably the people of the Sredniy Stog culture (ca. 5000 BCE to 3500 BCE), although some of their close neighbors in time and space are also contenders for that designation.

      The Indo-Iranian branch of Indo-Europeans who are notable for high frequencies of Y-DNA haplotype R1a-Z93 spread out into Central Asia and then split into an Eastern Iranian branch which spoke the Avestan language that eventually evolved into Old Persian and a South Asian branch which spoke a language which was or was very similar to Sanskrit.

      Their rapid expansion was driven by the domestication of horses, the development of the first wheeled vehicles including chariots, and superior metallurgy to their neighbors. Another factor was that they were culturally adapted to harsh arid environments of the type that caused civilizations including the Harappan civilization to collapse, leaving a political vacuum for a people better adapted to surviving in that kind of environment.

      “2) What was their language before arrival to SA?”

      An Indo-Iranian language ancestral to both Sanskrit and Avestian. Sometimes this is called Proto-Indo-Iranian. It was probably similar to the small Indo-European language families of Central Asia today.

      “3) Which language was spoken in NE India before Aryans arrival?”

      A language which was known to its own speakers as “Meluhha”, which is one of the few things we know about it for sure, based upon Sumerian records of an IVC ex-patriot traders colony in southern Mesopotamia. We can’t say that much about that language, however, because there is no written form of it that survives and is understood by people today. It isn’t clear if their symbolic seals were a full fledged language or were more of a set of accounting codes and firm and location logos.

      “4) Have they brought Sanskrit to India?”

      The Indo-Aryan people are the people who brought Sanskrit to India.

      “5) Have they brought Rig Veda?”

      Not really. While the Rig Veda which we know today is composed in archaic Sanskrit, the setting of the Rig Veda is mostly Northwest India and one of the rivers which figures prominently in it, the Sarasvati River, has probably mostly dried up by the time that the Indo-Aryans arrived and is described as healthy and fully flowing. So, so parts of the Rig Veda incorporate oral histories of the Harappans prior to the arrival of the Indo-Aryans. On the other hand, there are elements of the Rig Veda, like the presence of chariot riding warriors and clear parallels to Indo-European dieties, that suggest an Indo-Aryan influence. So, the Rig Veda was probably a synthesis of Harappan and Indo-Aryan source material.

      “6) Have they brought toponyms from old homeland? Any?”

      Probably, but I don’t have the ability to identify any of them easily.

      “7) Have they brought R1a to SA?”

      Yes. Specifically, predominantly the haplotype of Y-DNA called R1a-Z93 which is a clade of Y-DNA particular to people with eastern Indo-European patriline ancestry. European Indo-Europeans have Y-DNA R1a from haplogroups which are sister clades of R1a-Z93.

      “8) Who are the ancestors for 100+millions (R1a) of SA/Indians?”

      Their patriline ancestors are Indo-Aryans. But, everyone has lots and lots of lots of ancestors in addition to their father’s father’s father’s . . . . ancestor. The Indo-Aryan migration into South Asia was disproportionately male, so lots of of the 100+ million South Asians with Y-DNA R1a have mtDNA inherited from your mother’s mother’s mother . . . which is found predominantly in South Asia and in all likelihood in native to South Asia and emerged there.

      All surviving people in South Asia (with the possible exception of some people in Northeast India with a mix of Tibetan-Burman-Southeast Asian ancestry and South Indian native hunter-gatherer ancestry) have a mix of Iranian farmer ancestors and South Indian native hunter-gatherer ancestors, although the proportions vary by geography and caste. The lion’s share of South Asians with Iranian farmer ancestry also have Indo-Aryan ancestry. Generally speaking, Indo-Aryan ancestry is more common in higher caste individual and the closer you get to Northwest India. The Kalash people are probably the closest genetic match to the early Indo-Aryans.

      “9) Have Aryans ruled Sri Lanka in one period?”

      We don’t know for sure, because if they did, this would have happened before writing existed in Sri Lanka or nearby places that had regular dealings with Sri Lanka. We do know that Indo-Aryan influenced religious concepts (the Hindu religion like the Rig Vedas was probably a synthesis of Indo-Aryan and Harappan influences and possibly other influences as well) reached Sri Lanka, and we known that there are lots of people in Sri Lanka who have significant Indo-Aryan ancestry. But, we don’t know if that arrives during a period when Sri Lanka was ruled by Indo-Aryans, by Indo-Aryan missionaries, or by other migrants to Sri Lanka who had Indo-Aryan ancestry and had cultural practices influenced by Indo-Aryans, but who may have spoken a Dravidian language rather than a language derived from Sanskrit.

      I personally think it is more likely that there was some time period of at least a century or so during which Sri Lanka was ruled by a Sanskrit speaking elite with lots of steppe ancestry. But, it is impossible with the information that we have now to prove that definitively.

      “10) Did they have their state/territory in SA and what was its name?”

      The written history of India is about 2500 years old, with fragments a few centuries older. This is at least a thousand years after the Indo-Aryans arrived in South Asia. The names of their states, territories or tribes are speculative at best and a matter of legendary history. It isn’t even clear if they were organized into territorial states as opposed to genealogically organized tribes.

      It is likely that at least some of the names of the tribes and territories described in the Rig Veda belonged to historic Harappans and historic Indo-Aryans.

      But, where history ends and legend begins is very hard to know. The Rig Veda is recounting events that purportedly took place more or less around the same time as the earliest portions of the Avesta, the Hebrew Bible (a.k.a. the Torah), the First Dynasty of Egypt, the Akkadian Empire of Mesopotamia, and the arrival of the Bell Beaker people in Britain. The Rig Veda is set in a time period long before the Illiad and the Trojan War, or King Arthur’s Court. I call this era the era of “legendary history”. Legendary historic accounts often have seeds of truth, but also often contain anachronisms associated with the periods of time many centuries later when these legendary histories were compiled in written form such as attempts to associate modern place names and modern political body names with locations and political bodies that were present long before those names were invented.

      2+
      1. Thanks for your effort, I really appreciate this.
        However, the Aryan people still remain nameless. How come that such high civilization for that time, with steel weaponry, developed language (Indo-Iranian language at Sredniy Stog???). What does it mean Sredniy Stog and what is this language? People who made at least one part of Rig Veda (Karan said – the origin of RV is foreign (to India)), who brought Manu’s law still do not have their name. Could nomads do all these things or people from very developed culture, literacy and metallurgy? Sri Lanka – what was the first name of Sri Lanka, how it was mentioned in Sanskrit? Ancestors – I was thinking about pra-ancestors, the oldest R1a.

        And toponyms! One of rare taboo topics on this blog. I provided thousands so far without any comment. The most of them can be seen on colonial maps, in writings of ancient historians, some are unchanged from ancient times, some localized. And almost all of them already exist in Europe. Last week I provided a hundred of toponyms in Tibet, I have another thousand. Earlier, I provided hundreds of Aryan toponyms in India and the names of their tribes.
        Before I leave, as I promised, I will provide later on Open Thread STAN’s (Indo-Iranian?) toponyms for future researchers.
        Thanks again for your answers. Cheers.

        0
        1. “How come that such high civilization for that time, with steel weaponry, developed language (Indo-Iranian language at Sredniy Stog???).”

          The Indo-Aryans were a Bronze Age civilization. They didn’t have steel.

          Their language wasn’t especially developed or special, it just rode the wave of a culture that was expanding rapidly .

          Sredniy Stog was an archaeological culture in Eastern Europe (more or less in modern Ukraine’s territory) where proto-Indo-European was probably spoken. Google it.

          Indo-Iranian is one branch of the Indo-European languages that expands to the Southeast of Eastern Europe. Other branches of the Indo-European languages include Tocharian (to the far east), Anatolian (to the South, most notably the Hittite language), Greek (to the Southwest), the Italic languages (most notably Latin and its daughter Romance languages) and the Celtic languages (to the West), the Germanic languages (to the Northwest) and the Balto-Slavic languages (to the North and closer Northwest).

          “Could nomads do all these things or people from very developed culture, literacy and metallurgy?”

          They were herders and weren’t entirely nomadic. Better analogies might be herders in the hills of Greece or Scotland or the Appalachians even though they lived more on plains than in hills. Eventually they transitioned from an economy based mostly on herding to one based upon ruling farmers and collecting what they produced as a substantial part of their means of support.

          You are also unduly emphasizing a “very developed culture.” They were illiterate. They didn’t even have seals that carried meaning, let alone a full fledged written language. The Rig Veda was committed to writing roughly a thousand years after it was first composed orally in Sanskrit. Their metallurgy was better than many of their neighbors, but we are talking bronze and crude iron weapons, not steel, and sophistication in metallurgy that was, on average, less than that of a medieval village blacksmith. There was nothing intrinsically special about their language that gave them any kind of advantage. They didn’t have Arabic numerals or an ability to do math more advanced than basic arithmetic. They weren’t known for any sophisticated architecture or temples, although they had some elaborate graves. They didn’t have the mastery of surveying that the Egyptians did. They surely had religious ideas, but probably didn’t have anything comparable to the philosophical ideas of later Western and Eastern philosophers. Their highest level of political organization was probably not more than a chiefdomship at first, although over time their descendants would develop full fledged states and preside over cities (something the Harappans had already invented). The advanced urban civilizations they encountered would have considered them to be barbarians. Their main technological advantage was that they had domesticated horses and knew how to use those horses in war, while their opponents did not.

          “Ancestors – I was thinking about pra-ancestors, the oldest R1a.”

          The oldest bearers of Y-DNA R1a that is ancestral to that found in India lived on the northern part of the European steppe. The oldest bearers of Y-DNA R probably came from somewhere around the Altai Mountains to the south of Siberia.

          “I provided thousands so far without any comment.”

          Great. They have gone without comment, no doubt, because this is an area where you know more than most of the rest of us who have little to say about the subject because we are less informed. This does not make your knowledge less valuable, it makes it more singular and important.

          As for not having names, we simply can’t know what names people called themselves five to eight thousand years ago when they were all illiterate. We can and do give them names, but we can’t know which ones they used themselves.

          “People who made at least one part of Rig Veda (Karan said – the origin of RV is foreign (to India))”

          I would disagree. There were elements of the Rig Veda that drew upon foreign sources, but the RV itself was almost surely composed in India. It is in a language spoken only in India. It refers to geographic landmarks particular to India. Parts of its describe things that happened before the Indo-Aryans arrived in India. Other parts describe things that probably happened right around the time of the arrival of Indo-Aryans in India like the transition from inhumation as the sole means of disposing of dead bodies to cremation. Other parts describe things that could only have happened after the arrival of Indo-Aryans (e.g. chariot based warfare). And, all of these elements are jumbled together in an anachronistic fashion.

          By analogy, in the Hebrew Bible, the story of Noah’s flood, of the infancy of Moses, and part of the story of the garden of Eden, are all adaptations of Sumerian myths composed before the Hebrew language existed. And, these stories may have actually been fairly late additions to the Torah not added until the Hebrews had spent time in exile in Babylon near where the earlier versions of these stories were first told, and not too long before the Hebrew Bible was assembled into a coherent canonical collection of stories. But, I would not say that the origin of the Hebrew Bible books of Genesis and Exodus, in which those adapted stories appear with their own adaptations to Hebrew lore and culture, had an origin which was “foreign” to the Jewish people.

          In the same way, the Indo-Aryan people synthesized local material and their own culture that they brought with them into one mashup that was made into epic verses that were recited orally for centuries before being written down.

          1+
      2. ohwilleke says
        and we known that there are lots of people in Sri Lanka who have significant Indo-Aryan ancestry. But, we don’t know if that arrives during a period when Sri Lanka was ruled by Indo-Aryans, by Indo-Aryan

        Really, “Sri Lanka who have significant Indo-Aryan ancestry”. Any research article to prove based on genetics?

        Never been to Sri Lanka I guess, we are some of the darkest people in Asia.

        0
        1. Ancestry != skin color which is due to a dozen or two genes which are subject to selective fitness selection, with latitude, rather than being neutral and ancestry informative.

          In the Sinhalese of Sri Lanka, 23% of the men had Y-DNA R1a1a. Kivisild, Toomas; et al. “The Genetics of Language and Farming Spread in India”(2003)..

          The Kurumba of Kerala State, the closest sample geographically to Sri Lanka in the study in question, had 43% ANI ancestry, and while ANI isn’t pure steppe, it has a substantial steppe component. See David Reich, et al., “Reconstructing Indian Population History”, Nature (2009).

          The linguistically Indo-European Vedda people of Sri Lanka were 36%-41% ANI. No population in India is less than 16% ANI, and high caste Dravidians are about 50% ANI. Priya Moorjani, et al., “Genetic Evidence for Recent Population Mixture in India” AJHG (2013) (Table S4).

          “The demographic impact of steppe related populations on South Asia was substantial, as the Mala, a south Indian population with minimal ANI along the ‘Indian Cline’ of such ancestry is inferred to have ~18% steppe-related ancestry, while the Kalash of Pakistan are inferred to have ~50%, similar to present-day northern Europeans.” Iosif Lazaridis, et al., “Genomic insights into the origin of farming in the ancient Near East” Nature (2016). See also Extended Data Figure 5, showing that roughly half of ANI is Steppe ancestry in Southern India.

          Assuming that Sri Lanka would consider 18% autosomal steppe ancestry and on the order of 23% steppe Y-DNA, which would be, for example, roughly twice the amount of Turkish ancestry found in modern citizens of Turkey. Someone with that much Native American ancestry in the U.S. would probably be eligible to be a member of an Indian Tribe (the actual test is genealogical rather than genetic but similar in magnitude of ancestry). I’d say that this is significant Indo-Aryan ancestry.

          0
          1. No population in India is less than 16% ANI, and high caste Dravidians are about 50%
            The Kurumba of Kerala State, the closest sample geographically to Sri Lanka in the study in question, had 43% ANI ancestry, and while ANI isn’t pure steppe, it has a substantial steppe component.

            All above seems to imply ALL of India/South Asia (Including tribals) have significant ANI (above 30%).

            I guess I am biased toward the non Aryan part so the 50% and above ASI is more important to me.

            These are the Sri Lankan subset from HarappaDNA. Not my analysis just extracted and presented the subset.
            http://sbarrkum.blogspot.com/2013/04/sinhalese-and-tamil-dna-admixture.html

            http://sbarrkum.blogspot.com/2015/06/mtdna-of-sri-lankans.html
            It is quite astonishing to see such a lower frequency of M haplogroup in the Vedda population when compared with southern Indian tribal groups (70–80%) as well as southern Indian caste populations (65%).
            The Vedda mtDNA, Specially Vedda-Rathugala (VA-Rat) and Vedda-Pollebadda (VA-Pol) seems a genetic isolate. See Figure 5 in Ranaweera et al.

            Vedda Language (note only about 300 people left)
            The parent Vedda language(s) is of unknown linguistic origins, while Sinhalese is part of the Indo-Aryan branch of Indo-European language family. Phonologically, Vedda is distinguished from Sinhalese by the higher frequency of palatal sounds [c] and [ɟ]. The effect is also heightened by the addition of inanimate suffixes.

            https://en.wikipedia.org/wiki/Vedda_language

            0
        2. To sBarrkum, especially the second comment:

          I believe the science has moved a little bit beyond ASI/ANI as the issues with ASI/ANI became quickly apparent as early as 2010/11.

          Vagheesh Narasimhan et al. (supplement pages 168 onwards) indicates that both ANI and ASI are rich in Iranian farmer ancestry (max 38.9%+/- 4.1% in ASI and max 70% in ANI). Now we view Indians and Srilankans as mixtures of 3 populations (OK, 4, SET MUNDA aside) : AASI ( that is enriched in ST of India, and correlated to Onge), Indus periphery farmer (that is present in most Indians) and the Steppe_MLBA (gradient from north and west to south and east). Thus Srilankas have no worse than 13% Steppe-MLBA, and may be up to 53% Iranian farmer, and ~ 34% AASI.

          In a similar manner, Razib broke down genes of skin color by population, and I spent hours looking at dependence of occurrence of skin color gene(s) on Steppe_MLBA or IP, and again, with distance from Northwest India. Unfortunately, it will need a machine language/big data type of approach, like the Gausian approach for sparse ML of Mrs.Zavide. The data can be convoluted in terms of AASI, InPe, and Steepe, but needs work.

          0
        3. Isn’t Sinhala considered an Indo-Aryan language? Are you implying that this is a case of language transplant over native population?

          0
  16. It seems you are latecomer to this site and missed many writings in previous months. Apparently you do not know anything about European aborigines, about Vinca and Lepenski Vir, about the oldest R1a (12000 years) in Europe (which later came to India), the oldest alphabet in the world from Vinca, the oldest calendar in the world, the swastika, the language which still has several thousands of words identical with Sanskrit, about the first Aryan leader who founded Babylon (and Baghdad got the name after him). Indo-Iranian language was not spoken deep in today’s Russia and thousands of years before they came to India that had this name. It is so illogical. Indo-European languages do not exist. This term is idiotic (it was called Indo-Germanishe for 100 years before this). Only one language was spoken in Europe at the time of first Aryan invasion. The language which is older than Hebrew, Greek and Roman. Which is (according to, for e.g. linguist Tagore’s, granddaughter) now 30% identical with Sanskrit…etc, etc. If you are interested you may search comments in previous months. Thanks.

    0
    1. “about the oldest R1a (12000 years) in Europe (which later came to India), ”

      The oldest R1a in Europe (apart from the European steppe near the eastern boundary between Europe and Asia at the Urals) is not the source of R1a in India. The oldest R is Ma’alta ca. 20,000 years ago near the Altai. Indian and almost all modern European R1a is derived from R1a on the northern European steppe from ca. 5000 BCE and later.

      “the oldest alphabet in the world from Vinca,”

      Like the Harappan seals, the Vinca script is not an alphabet, its basically an accounting shorthand for a limited number of topics in the form of logograms. You couldn’t write a poem or a fable or a history with it.

      The oldest written full fledged languages in the world were Sumerian Cuneiform and Coptic Hieroglyphics.

      “the oldest calendar in the world, the swastika,”

      Nope. There are older calendars in Anatolia (ca. 12000 years BP) that pre-date the Neolithic revolution.

      “about the first Aryan leader who founded Babylon (and Baghdad got the name after him).”

      False.

      “Indo-European languages do not exist. This term is idiotic (it was called Indo-Germanishe for 100 years before this).”

      Nope.

      “Only one language was spoken in Europe at the time of first Aryan invasion. The language which is older than Hebrew, Greek and Roman. Which is (according to, for e.g. linguist Tagore’s, granddaughter) now 30% identical with Sanskrit…etc, etc.”

      False.

      0
      1. I am not here but coincidentally I saw this reply. Initially, I was thinking that you are genuine but, I confess, you tricked me. Maybe there is a lack of knowledge but it is compensated by high degree of arrogance. And this quasi knowledge is also based on Wiki. It is illusive to comment on anything in your reply because there is nothing there. I see that you are expert for Vinca’s alphabet too, which allegedly is not alphabet, although modern Serbian alphabet, which is probably the simplest in the world, has 27 out of 30 letters from this 7000 years alphabet. And Raseni (Etruscans) who founded the city of Rome (Ruma) used these letters which were also a base for Roman alphabet (and their language, too). But, off course, Wiki does not say anything about this.

        I could expect similr things after such dumb construction that some (?) noname people who lived in Ukraine and spoke Proto-somelanguage, somehow changed this to new one, named as Indo-Iranian (?) probably couple thousands years before they actually came (if they came) to SA and became Indo-Aryans, Afgano-Aryans, Tibeto-Aryans, etc…And from this I-I language originated Sanskrit and Avestan? And, they came with new language to SA but the locals made all those epics on their language! Wow, what imagination! Just missing, how Sanskrit (or I-I?) came to Europe. It is similar If I say that I am writing in Proto- Martian because in 1000 years some people probably will be living on Mars.

        After such verbal diarrhea we still don’t know who Aryans were and which language spoke but we know about anonymous ‘steppe’, ‘farmers’, cowboys who spoke Indo-Europeans languages, Balto-Slavic (3-4000BC???), etc.

        If you think that you are not hopeless (I think you are not), you may take this comment as a referral to Dr Jaggu@Stan and ask him to prescribe you the ‘barki’ medication. It should not only cure the diarrhea, it would tug you form Wiki back to reality.

        0
  17. @Falana

    // Per “Substrate Languages in Old Indo-Aryan”, M.Witzel, there is no Dravidian substrate in the oldest parts of RgVeda. Dravidian influences only begin in middle and late part of the corpus. //

    What you are citing is pretty old work (maybe from 20 years ago). You really need to cite modern linguistics research. (Besides Witzel has his pet theory about para-Munda being the language of IVC which few linguists believe in)

    Retroflexion was present in the earliest parts of the Rg Veda, even *before* the consonant clusters (with voiced sibilants) evolved, e.g. Proto-IA *nIzDa > Skt nIDa, and with ample evidence from vyaJjana saMdhi rules, e.g. prch + ti = prSTi in Vedic. The fact that the oldest retroflexion in Sanskrit is dental shows the early effect of Dravidian (dental retroflexion is a core feature of that family), so does syntactical evolution of Vedic dialects all the way into post-Panini prkRta phase.

    Retroflexion is indeed present in isolates like Burusho, but modern research (see Sadaf Munshi’s thesis) suggests that this development may have been under the influence of Dravidian.

    3+
    1. Apologies for quoting another citation from 20 years ago. But facts haven’t changed so drastically in linguistics like they have in the field of genetics 🙂

      From “Language History, Language Change and Language Relationship”, H.H.Hock (1996) pages 409-410

      The early Dravidian has a triple contrast between dental/alveolar and retroflex. This is divergent from early Indo-Aryan (RgVeda) which has a dual contrast (delta and retroflex). In addition, early Indo-Aryan had extensive retroflexion on sibilants. Totally absent in early Dravidian.

      It’s only in the modern Indo-Aryan that we see a triple contrast. And it’s only in modern Dravidian where we see retroflexion on sibilant (Malayalam). Both of these are a result of convergence and language contact.

      The evolution of IA from 2 way contrast to 3 way indicates that retroflexion was an areal feature, not a substrate borrowing from Dravidian.

      > The fact that the oldest retroflexion in Sanskrit is dental

      Nope. OIA has retroflexion on dental stops as well as sibilants (unlike old Dravidian). The vedic ritualists who laid the foundation of Brahminism in the subcontinent were clearly in language contact with a retroflexion-heavy language family and probably even bilingual. But sorry, it wasn’t Dravidian.

      2+
  18. Slapstik,

    How do you reconcile the above with the ubiquity of retroflexion – a key phonetic feature of Dravidian – across S Asia and nearby zones on the Iranian plateau? E.g. retroflexion in Pakhto spoken in Kandahar (Arachosia) or Khotan Saka?

    Retroflexion is certainly not a super-strate in Dravidian but a deep and defining feature of this language group. And it is a substrate in both IA and E Iranic and certainly not present in PIE.

    You do know, don’t you that even the likes of Michael Witzel, an avowed invasionist, does not support Dravidian as the language of the IVC ?

    Most of the western academia has almost given up on this Dravidian = IVC theory because it is simply unsubstantiated. Care to show me which major linguist supports it ?

    Retroflexion is an areal feature of the entire subcontinent barring perhaps the North East. It even extends deep into the North, perhaps as far as the Pamirs. Are you going to argue that all of this vast sprachbund acquired its peculiar linguistic characteristics from Dravidian ? Don’t you think that it is a bit of a tall claim ?

    And you also do know that PIE is little more than educated guesswork, don’t you ? There is really no solid evidence that PIE reconstruction as we have it today is 100 % a reflection of some actual PIE 6500 years ago.

    So to assert that retroflection was not a feature of PIE and therefore Indo-Aryans acquired it when they came to South Asia is also guesswork.

    The reconstruction of PIE is heavily tilted in favour of the IE languages outside of India, since they have been broken up into several branches while the Indo-Aryan is treated as one branch even though the Indologists quite clearly know that there are several features in ancient Prakrits (Middle Indo-Aryan) and NIA like Pahari languages and even those like Oriya & Assamese, which are not inherited from Rigvedic Sanskrit but from the earlier Indo-European stage. There has not even been a proper classification of Indo-Aryan languages in India and there has been no attempt to answer what the archaic Indo-European features, lacking in Old Indo-Aryan (Sanskrit), but present in MIA & NIA mean. Most linguists/Indologists like Witzel, K R Norman, Thomas Oberlies, Claus Peter Zoller try to explain this by multiple waves of Indo-Aryan migration into South Asia with the Rigvedic Indo-Aryans being the last.

    But the question is – on what basis do you classify all these disparate migrating groups into South Asia under a single branch of Indo-Aryan ? The linguistic situation in South Asia is clearly quite complex, and the linguists are having a hard time grappling and coming to terms with it. They come up with ridiculous arguments like Para-Munda hypothesis for the IVC but now we know that the Mundas are recent immigrants into South Asia from the East.

    Retroflexion is an inherent Dravidian feature because all of the Dravidian languages are spoken with the South Asian sprachbund where retroflexion is ubquitous while a large percentage of IE languages are spoken outside of South Asia where retroflexion is absent.

    Languages migrating out from South Asia can also lose retroflexion just as languages migrating into South Asia come to acquire it. The Romani language, which definitely migrated from NW India, has lost retroflexion.


    And if you are going to argue that IVC invaded S India, then retroflexes should be the super-stratum in tamiL and kannaDa. Which they are not.

    Retroflexion is a pan-South Asian phenomenon, which means it was already present in South India and was not brought about from the North.


    Dravidian also has lent a substantial substrate lexicon in the earliest Vedic speech spoken in N Punjab. And the transition from Old Vedic to early and middle prkRta-s also shows clear signs of Dravidian language throughout. E.g. endemic retroflexion, diphthongs > long vowels, simplification of consonant clusters esp with voiced sibilants etc. How did that happen so far up in the North, so early on?

    Can you substantiate these claims with references ? Thanks.

    For your general reference, I am sending a link of the homepage of an upcoming Russian linguist

    https://independent.academia.edu/IgorTonoyanBelyayev

    Do read some of his papers. Maybe you would find them interesting.

    1+
    1. There has not even been a proper classification of Indo-Aryan languages in India and there has been no attempt to answer what the archaic Indo-European features, lacking in Old Indo-Aryan (Sanskrit), but present in MIA & NIA mean.

      The minutiae of linguistics are quite arcane to me, so I can’t really judge any of the competing claims, but I recall reading that Sanskrit is not considered an ancestor of Prakrit and other Indo-Aryan langauges, but rather a sister of Old Indo-Aryan. It’s possible that Sanksrit was explicitly spun off from the colloquial language(s) for the purpose of liturgy (and really, to make it easier for memorizing and oral communication through the generations.) What do you think of that theory?

      The Romani language, which definitely migrated from NW India, has lost retroflexion.

      That seems to be a strong point in your favor, but again I can’t judge. But I guess the one classification that is completely incontrovertible is that the Indo-Iranian language (family) existed. So can a comparative analysis of Indo-Aryan and Iranian languages help in this debate?

      A more general question: you seem to be trying to counter genetics-based theories with linguistics-based ones. And it looks like you may have a reasonable linguistic model for OIT (though I can’t judge.) Shouldn’t you be taking the evidence from both genetics and languages at face value and trying to come up with a model that fits both? Would that still give you a model that is stronger on OIT than AIT?

      To me, when I look at the genetic theories, they seem to be simply confirming what I would think when I look around at the different people of the Indian subcontinent. That we are extremely diverse (skin color being the most obvious marker.) And if the diverse mix that are Indians migrated a few thousands of years to Europe, to the extent that they replaced existing populations (as seems to be the case in places like Ireland), it’s really hard to imagine how they all turned out be lily-white (pardon my bluntness.) The Romanies, after a millenium in Europe, still don’t seem to have lost any of that diversity, for example.

      2+
      1. I would be careful about attributing lack of dark skin in modern Europeans to some genetic uniformity in them and likewise attributing complexion diversity in subcontinent to multiple genetic lines. Dark complexion was there in Europe prior to farming, and likely got wiped out very fast following agriculture and Indo European expansion. It looks like they had heavy selection against dark complexions in the early history phase, something I doubt ever happened in India.

        Truth be told this might actually be happening today. I can guarantee you that Indians as a whole are getting lighter. Look up stock photo of Kashmiri Pandits on wiki. I think people in urban India are breaking down caste as far as marriage goes, and basically light skin is coaggulating whereas dark skinned individuals are pairing up with light partners. Also I don’t think modern Indian dark and modern African dark are the same biologically at all (this notion comes from frankly racist notions about black being “original” color). It may appear so in adulthood but the base color of darkest Indians is lighter. I am yet to meet a single person from the recently upwardly mobile urban middle class that has made its way out of poverty (generally OBC and SC) that are as dark as their parents. You combine these things, and Indians are getting light fast. Maybe within a span of last century this process is unfolding. I even believe in various pockets, the dark skin of a particular sub caste may have been a symbol of elitism like some Rajput clans in Pak Punjab, or some Konda in Andhra (clearly Gond association…adivasi–>kshatriya group similar to Maratha ascendancy) but they are hurriedly dropping the dark elitism in favor of light elitism now and leveraging their elite status to acquire lighter skin.

        0
    2. @Jaydeepsinh

      I do not have the time to respond to your long comment inline. However, I will respond to the general issues raised by you and others (most of which are misconceptions or simply a lack of understanding of linguistics) in a separate blog post.

      However, just few short points will suffice for now:

      – Major linguist that supports Dravidian IVC = Asko Parpola, probably the world’s foremost authority on IVC

      – Your point about Romany losing retroflexion is trivial. Heck! why go to Romani when Indians living in US/Britain approximate dental retroflexes as alveolar stops within a lifetime. Retroflex dental — Alveloar stop is a continuum of articulation and switching between these modes is common; especially when the larger linguistic community one is immersed in is at one end of the spectrum. Humans mimic what they hear.

      – The above point also indicates how dental retroflexes entered early IA (as its speakers were surrounded by the biggest Indian language group – Dravidian – with these features)

      – As far as backing claims about saMdhi and phonetics goes (which I think most IE linguists will agree with), but anyway:

      http://www.academia.edu/472464/The_influence_of_Dravidian_on_Indo-Aryan_phonetics

      – Finally “PIE is little more than educated guesswork” and so are Newton’s laws of Physics. Or all of Physics for that matter.

      2+
      1. “Major linguist that supports Dravidian IVC = Asko Parpola, probably the world’s foremost authority on IVC” I would beg to differ. While he has certainly studied IVC in depth, he is very much old school with conclusions that are dated, and is not considered a leading current scholar in the linguistics world. Witzel would be a pretty clear contender for the current foremost authority in IVC linguistics.

        0
        1. And despite that, Witzel’s theory is bunk. There’s no way you can square IVC proto-Munda with the current genetic evidence.

          3+
          1. Witzel’s choice of name “Para-Mundi” is wrong and was always speculative.

            But, Witzel’s underlying data about what words are derived from a substrate to Indo-Aryan in the Rig Veda is strongly in consensus with independent scholarship by at least two others in the field, as is his conclusion that there are only 30-40 Dravidian loan words, and as is his conclusion that none of the Dravidian loan words appear until the middle to later Rig Vedic texts and is absent from the earliest ones. This evidence strongly disfavors the hypothesis that Meluhha was a Dravidian language.

            1+
        2. @ohwilleke

          I give very little credence to the views on linguistics (or linguists) of a person who thinks Old Persian and “a language very similar to Sanskrit” evolved from Avestan!

          “the Avestan language that eventually evolved into Old Persian and a South Asian branch which spoke a language which was or was very similar to Sanskrit”

          2+
          1. What I mean to say is that Avestan and Sanskrit evolved from Proto-Indo-Iranian, and that Old Persian evolved from Avestan. Sorry if the sentence didn’t parse right.

            0
      2. Just read this whole long winded thread about retroflexes, and frankly this whole ratholing about retroflexes in sanskrit appears making mountain out of molehills.

        Firstly, I don’t even understand what the so called “dental retroflexes” are. There are dental consonants and there are retroflex consonants. Never heard of anything called “dental retroflexes”. Is it possible that you are confusing between alveolar consonants (English t/d -https://en.wikipedia.org/wiki/Alveolar_consonant) and retroflex stops ( Sanskrit T/D – https://en.wikipedia.org/wiki/Voiced_retroflex_stop) ?

        Similar phonemes can arise in unrelated languages independently. You don’t have tie yourself in knots to seek a deep connection between them.

        Secondly, to me the most outstanding feature of Sanskrit is its abundance of aspirated consonants. Dravidian languages scarcely have them, and the most Dravidian of the Dravidian languages, Tamil completely lacks them.

        And finally, any one who quotes Witzel must be reminded that this guy is a complete crank. He once came up with a crank theory that IV symbols do not even constitute a script. To seek publicity for his theory he even offered 10 thousand dollars as reward to anyone who would disprove his theory! Have you ever heard of any serious academic offering rewards to disprove their work? Doesn’t it sound like some Bangali Baba proudly claiming- Jhoota saabit karne wale ko dus hazaar rupaiyya inaam! Seriously, Witzel should be kept out of any serious debate.

        1+
  19. Dr. Khan’s article takes a conciliatory tone but glosses over many burning questions posed by Milan that simply cannot be set aside.

    “One can push the date a few centuries earlier (there is a possible Indo-Aryan text in Syria dated to 1750 BC), but not too much earlier.”

    No sir. They surely can be pushed back by 3000 more years. Check out this latest paper by Andrew Garrett

    https://www.amphilsoc.org/sites/default/files/2018-08/attachments/Garrett.pdf

    If truly scientific biology based phylogenetic approaches are applied to linguistic the dates get pushed back by the 3 or 4 millennia. The linguists who have been busy reconstructing languages for 150 years basically ASSUME that Old Irish is the mother of modern Irish languages, Vedic is the mother of Indo Aryan and so forth. But in fact they are aunts. The problem becomes obvious after comparing Figure 3 and with figures 4& 5 in the paper. In Garret’s own words:

    ” In linguistics, it is conventional to assume that Latin is the ancestor of the
    Romance languages, and that Old Irish is the ancestor of Modern Irish
    and Scots Gaelic. But in the inferred tree in Figure 3, produced by
    phylogenetic methods, Old Irish is actually an aunt or uncle of Modern
    Irish and Scots Gaelic, and Latin is an aunt or uncle of the Romance
    languages”

    The same relationship would hold true for Vedic Sanskrit and Indo Aryan.

    Garrett concludes:
    “Because the Anatolian model had the virtue of being tied to a well-understood mechanism of language spread—the diffusion of agriculture—the emerging consensus for Indo-European means that models of global language spread must explain more clearly how pastoralists expand their territories.”

    “But the presence of people with a “steppe” genetic profile, which is ubiquitous in much of Northern India today, seems to date to the period after 2000 BC. ”

    Dr. Khan is a geneticist himself and I believe he is well aware of the problems with the Narasimhan (2018) study as explained in great technical detail by Premendra Priyadarshi.

    https://aryaninvasionmyth.wordpress.com/

    Leaving archeology and genetics aside the “main stream” steppe theory cannot stand LINGUISTIC scrutiny also.

    2+
    1. You clearly didn’t read or understand the article.

      “The steppe model is favored by a consensus now emerging from linguistics, archaeology, and genetics (see David Reich’s contribution to this issue).”

      Garrett is 100% behind the steppe hypothesis. Garrett refutes the idea that, for phylogenetic purposes, we should consider Old Irish or Latin as uncles instead of ancestors to modern Gaelic or Romance languages. When you run the same analysis as the New Zealand team, but explicitly code archaic languages as ancestors to modern languages, you get a time depth of 4000BC. When you abandon ancient languages altogether and only use modern IE languages, you get a time depth of 4500BC. I.E. both within the range expected of the steppe hypothesis. Including ancient languages and NOT explicitly coding them as ancestral leads to a distortion of the estimated time depth. That’s why the original analysis by the NZ team put the origin of PIE so much earlier than traditional linguistics.

      When Garrett says that we need to explain how pastoralists spread their territory, he means that, having established beyond reasonable doubt that PIE originated on the steppe, we now need to explain how it spread from the steppe, because the mechanism for that spread supposedly isn’t as clear as the mechanism for the spread of language by agriculture.

      Personally, I don’t see what we need to explain. There are historically attested examples of steppe languages spreading far and wide across the world, including into areas that are not part of the steppe. And it makes sense that pastoralists would have an even easier time spreading their languages in an era like the Early Bronze Age, when settled societies tended to be less complex.

      1+
      1. it’s not implausible (and an accepted model) that PIE started in anatolia and a major offshoot went to steppe via maykop and then steppe expanded. i assume u know this?

        0
      2. btw, dod uyou bother to read the citations you put here? conclusion:
        “Despite support for an Anatolian Indo-European origin, we think it unlikely that agriculture serves as the sole driver of language expansion on the continent. The five major Indo-European subfamilies – Celtic, Germanic, Italic, Balto-Slavic and Indo-Iranian – all emerged as distinct lineages between 4000 and 6000 years BP (Fig. 2 and Fig. S1), contemporaneous with a number of later cultural expansions evident in the archaeological record, including the Kurgan expansion (5–7). Our inferred tree also shows that within each subfamily, the languages we sampled began to diversify between 2000 and 4500 years BP, well after the agricultural expansion had run its course. “

        1+
      3. That’s the New Zealand team that Garrett was talking about. He and Chang released a follow-up that refined the methodology and confirmed the timing of the steppe hypothesis. The problems with the New Zealand team’s work, and how Garrett and Chang improved on it, are found in the first article you linked.

        1+
      4. Quentin Atkinson’s work has consistently been criticized as bad linguistics by linguists (rightly so), and a few years later he wrote another paper acknowledging the mistakes in this one and arguing for a much shorter chronology.

        0
  20. Razib, would I be correct in inferring from your post that Joseph’s book implies that the South Indian Neolithic is an offshoot of IVC agriculture?

    2+
  21. Kudos to the podcast. There is little to no information regarding Niraj Rai’s poster at ASHG 2018 with ancient DNA, but the podcast did offer valuable clues on the composition of IVC culture as AASI + Early Iranian Farmer (Zagrosian) with a Proto-Dravidian speaking population.

    I am curious to understand the frontiers of IVC expansion in the Indian Subcontinent before the arrival of Yamnaya Pastoralists between 2000-1000 BC.

    We had this 2013 paper from Max Planck detailing Halocene gene flow from India to Australia 4230 years ago (before 2000 BC) via boats, ships (as they ruled out land route via South East Asia) that also brought Dingo to the Australian continent.

    http://www.pnas.org/content/pnas/110/5/1803.full.pdf

    Can the Dingo offer clues on the frontiers of IVC in 2000 BC whether it was Sindh, Gujarat, Konkan coast or were they deep inside Peninsular India ?

    0
    1. We had this 2013 paper from Max Planck detailing Halocene gene flow from India to Australia 4230 years ago (before 2000 BC) via boats, ships (as they ruled out land route via South East Asia) that also brought Dingo to the Australian continent.

      no one in the field believes that paper. has not been confirmed by the follow-up. statistics were not good.

      1+
  22. In the podcast, Razib did mention that Dalits in UP and Tamil Nadu are genetically similar despite geographically separated by thousands of miles. It would be great to cover this phenomenon in detail in subsequent podcasts when other such disparate groups are found.

    Is there any Iranian cline (similar to the Steppe cline) amongst ASI shifted groups in India ? It is mentioned in the podcast that Early Iranian Farmers have been in Mehrgarh which preceded IVC for a long time.

    How will population geneticists model peopling in the subcontinent before 2000 BC as I am under the impression that ANI, ASI model requires Eurasian gene flow that only occurs in the Bronze Age ?

    Another noteworthy information from the podcast was that rice cultivation in the sub continent preceded Munda migration and they could have brought just a specific type of Rice. What are the prevailing theories on the origins of rice cultivation in the Indian Subcontinent ?

    0
    1. Is there any Iranian cline (similar to the Steppe cline) amongst ASI shifted groups in India ? It is mentioned in the podcast that Early Iranian Farmers have been in Mehrgarh which preceded IVC for a long time.

      the iran cline is sindhi to groups like reddys to dailts. these are groups skewed for iranian as opposed to indo-aryan

      4+
    1. It not the 100 comments that is important, but no more than 3-4 comments trashing the model and trumpeting the OIT! BP commenters must be an exclusive club and no way representative of Indians. If this has been published in an Indian paper, there will be a 100 comments staring with “now disproved AIT/ max mueller/Caldwell—–“.

      3+
      1. I feel that i am missing out on something. Like are people really that into “we came here 2000BC/ No we came here 1500 BC/ No we went out 3000 BC ” stuff ? Its not even funny anymore

        1+
        1. An average Indian does not care. However, over the last 30 years, a huge contingent(numerically, not percentage-wise) has launched themselves to establish:
          1. Everyone/Everything is Indian, meaning Hindu (people or concepts)
          2. There was no invasion of anykind, and there is little to differentiate between all the people of India
          The original contingent used internal references from vedas and epics to prove anything, the newer version uses linguistics and similar sounding words in other languages to build the OIT.

          To be completely honest, every group of Indians indulge in this; for the Tamils, they arrived as fully formed Dravidans from Lemuria or Kumarikandam; all languages are evolved from tamil. The Aryans as a separate people did them in using religion, before the arrival of whom, the Tamil society was the egalitarian rural heaven that we still need to attempt. They use literature selectively, parpola is fine as long as he says IVC used Tamil words; Reich will be used as long as they can interpret ANI=R1A via Khyber. However, subsequent interpretation of Dravidian as invaders from Iran and proto-Dravidian is not Tamil, will be ignored. For selective quoting, Tamils are no less than Hindu contingent, with fanciful history of Jain and Buddhist tamilian societies that were ellysium disrupted by Brahminical Hindus from Khyber pass. The use of scientific papers of a particular time (tamils populated Australia, Cambodia, SriLanka, Antartica) is a trademark of this contingent

          The need to dump all of these arguments, and move on to the new paradigm is essential to India. The good news is people do not seem to care, specially now, compared to say, 1990s. At that time, the congress government had a ban on blood screening and DNA evidence. I am thankful for small steps.

          3+
  23. “To be completely honest, every group of Indians indulge in this; for the Tamils, they arrived as fully formed Dravidans from Lemuria or Kumarikandam; all languages are evolved from tamil. The Aryans as a separate people did them in using religion, before the arrival of whom, the Tamil society was the egalitarian rural heaven that we still need to attempt. They use literature selectively, parpola is fine as long as he says IVC used Tamil words; Reich will be used as long as they can interpret ANI=R1A via Khyber. However, subsequent interpretation of Dravidian as invaders from Iran and proto-Dravidian is not Tamil, will be ignored. For selective quoting, Tamils are no less than Hindu contingent, with fanciful history of Jain and Buddhist tamilian societies that were ellysium disrupted by Brahminical Hindus from Khyber pass. The use of scientific papers of a particular time (tamils populated Australia, Cambodia, SriLanka, Antartica) is a trademark of this contingent”

    This contingent though seems localized to tamil nadu, is this view rife among the diaspora tamil populations? I see this as no real danger as these types are regularly ridiculed in even tamil nadu itself Unfortunately the OIT and their supporters have a larger demograhpic prescence, with representation pan-india wise, in all parts of the world among the diaspora, with scientists in high levels of indian or western academic instituions, and even european origin converts/sympathizers to hinduism trying to “scientifically” support this view. Even tamils in malaysia support OIT (I’m not sure if it’s because the use of a “Hindu” identity to combat discrimination by the bhumiputra malays).
    Though I do share your worry for tamil nadu and would not wish this type of quackery to infect it as well as any other.

    2+
    1. I cannot comment on what Malaysian and SriLankan Tamils think about this; but I am willing to let people indulge in past glory to the extent that it does not affect others.

      Southern whites seem to view the confederacy as glory days, but cannot understand why the former slaves should be offended by displays of confederacy and white power. Muslims in several nations consider direct descendance from the prophet important, the regime of the four guided Caliphs to be the glory that they would like their present governments need to return to, ad PreArabic cultures to be barbaric. It is the part of every culture to consider one small piece of their prehistory (imagined or otherwise) as the golden mean to attain and freeze it forever. The last election may be viewed as an attempt by a disappearing majority to reach a glory that never existed or imagined. Who is to tell tell them that the past glory never existed, and never can be reclaimed?

      0
  24. @snake charmer

    Just to clarify the “dental retroflex” is a correspondence, not a thing.

    I clarified it via the saMdhi rule: prch + ti = prSTi, i.e. Vedic/Sanskrit speech does not tolerate the primary dental (dantya) consonants in some cases and re-maps them to retroflexes. And that this correspondence holds from the earliest part of Vedic lit.

    I agree given the place of articulation, it is simply a retroflex stop. This special correspondence with dental is entirely an Indic thing (evidenced from the time of the Vedas).

    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.