Harappan site of Rakhigarhi: DNA study finds no Central Asian trace, junks Aryan invasion theory:
The much-awaited DNA study of the skeletal remains found at the Harappan site of Rakhigarhi, Haryana, shows no Central Asian trace, indicating the Aryan invasion theory was flawed and Vedic evolution was through indigenous people.
…
“The Rakhigarhi human DNA clearly shows a predominant local element — the mitochondrial DNA is very strong in it. There is some minor foreign element which shows some mixing up with a foreign population, but the DNA is clearly local,” Shinde told ET. He went on to add: “This indicates quite clearly, through archeological data, that the Vedic era that followed was a fully indigenous period with some external contact.”
I haven’t heard anything definitive, but this is what I have heard: that the genetics they could analyze indicates continuity, but none of the steppe element ubiquitous in modern North India (and that there was contamination in the Korean lab). The Rakhigarhi samples date to 2500 to 2250 BC last I checked. That means they shouldn’t have any steppe ancestry if the model of the relatively late demographic impact of Indo-Aryans after 2000 BC is correct.
Basically, the whole article is kind of a non sequitur. I do understand that many archaeologists think there was continuity culturally. And there could have been. But taking into account the genetics of the modern region of India where Rakhigarhi is located, there was a major demographic perturbation after 2250 BC.
Am a bit puzzled about the appearance of the story because AFAIK that journalist doesn’t have any Hindutva sympathy, nor does the TOI group. Most likely a researcher fed her technically correct but carefully cherrypicked info (say, suggesting that there is more cultural continuity than AIT guys think, that quote “He also discounted the notion of any violent conflict.” etc.), and let her extrapolate.
Not that I am complaining 🙂
Due to the single sample and long wait time, this is getting to be the most overhyped ancient DNA paper ever.
Not expecting to change your priors on Indian media, but for the record here is the author’s clarification:
https://twitter.com/anubhutivishnoi/status/1006856699435626497
This news report is again quite puzzling. Unless the paper is out there is no use speculating on such news stories. Perhaps, with this new article, it maybe that the Rakhigarhi paper is indeed close to release.
I should also submit that the Narasimhan et al paper make no conclusive case for steppe intrusion into South Asia in the 2nd milenium BC.
Just look at the p value table with 2 right sets of outgroups for modelling of Mala, Punjabi, SPGT & Swat – H. (Page 176 – Supplementary Text).
The best p values for Mala are with Okunevo & Tepe_Hissar while for Punjabis it is Khwalynsk_EN + BMAC (1st right set) or Khwalynsk_EN + Indus_P (2nd right set). Khwalynsk elsewhere shows itself to have one of highest levels of allele sharing with WSHG (fig S3.22), higher than even Okunevo. So we cannot argue that the presence of Khwalynsk as the steppe source for Punjabis indicates EHG intrusion into South Asia and not something related to WSHG or ANE as is already found in Indus_P and Chalcolithic Central Asia.
Even for SPGT, which by the way is an outlier population on Indian clone, it is steppe_mlba_West rather than steppe_mlba_east which makes the best fit. Geographically this makes no sense whatsoever since there is no archaeological link that connects the western steppe groups to populations in South Asia.
If this was not enough, they argue that the steppe did not impact BMAC and the BMAC did not impact South Asia.
The authors appear to have taken Kuzmina too seriously on this matter. They should read the archaeological supplement that has come out with the Daamgard et al paper. The archaeological support for a 2nd millennium BC movement into South Asia is non-existent. The supplement by the way has no Indian archaeologist and it is their word not mine.
they’ve been talking about posting the preprint to journalists for a few months. the contamination in the korean lab seems to have slowed them down. basically it was east asian ancestry and i think they assume it’s just korean.
the general results have been circulated widely already.
Out of 148 skeletons, only 2 yielded any relevant DNA. Even that DNA was contaminated with presumably some Korean DNA. And via this pathbreaking research, contradictory conclusions are already being drawn (“junks AIT” vs reinforces AIT).
Genetic articles are no longer academic. Now they are pure entertainment.
snake charmer, you’re stupid. don’t comment here again.
I can only imagine how the headlines in morning newspapers in Seoul would have looked like had they not discovered the contamination. “Geneticists discover ancient Koreans were the builders of Harappan civilization”.
Please post this last one 🙂 It is so funny. I promise I won’t troll you again 🙂
.
[yes, that was funny enuf. the contamination would have been obvious -razib]
Rakhigarhi people who lived around ~2600 BCE did not have steppe ancestry, but the Rakhigarhi villagers who are alive in 2018 CE have proven steppe ancestry — so, those steppe genes (politically correct term for “Aryan”) must have got in in those 4600 years.
I didn’t realize it yesterday, but this seems to be the thought process which made some interpret the results as undermining AIT:
In other words, the assertion is: “Aryan is a cultural word, the culture comes from the Indus valley civilization, the contribution of Steppe pastoralists being nominal, so whatever migration or invasion happened it was not Aryan”.
“Rakhigarhi people who lived around ~2600 BCE did not have steppe ancestry, but the Rakhigarhi villagers who are alive in 2018 CE have proven steppe ancestry”
I wonder why the Jatts from Haryana have so high steppe ancestry. And also Brahmins being strongly steppe+ASI and non-Brahmins being mostly west asian_farmer+ASI looks weird.Phenotype wise they dont look much different, looking at Brahmins and non Brahmins(excluding Dalits) from cow belt region, Bengal, South India ect. They look pretty similar, the only people that look different are Dalits.
Do you think there could be any major findings moving forward which could lead to any conclusion to the whole thing. Is there something (like DNA sample after 2250 BC) which would decisively settle this issue one way or the other?
Razib,
Since the sample size is so small (n=1?2?), do you know if there might be any additional, uh, “caches” of skeletons in that area to mine off of besides ~150 they used?
Looks like my comment yesterday had gone to spam. That is okay, don’t bother, it was just a casual remark.
@froginthewell That journalist may be writing her most life endangering story in her life even though it is almost a throwaway article based on a press release. I wouldn’t want to be in her shoes.
@razib
Honestly, what do you think could be the language of Indus?
What haplotype/haplogroup they could have possessed?
Piece by piece the classic interpretation of Indoeuropeanisation was proven to be right whereever Indoeuropeans were or still be found. The only open question to remain is Anatolian.
As for India, some of the commenters seem to be unteachable. It has nothing to do with science any more, especially if they argue the way they did recently. How can anybody who calls themselve a scientist still argue against AIT? Its impossible. Its impossible to explain the modern Indians any other way and thats what matters. Or do they argue that the modern distribution of the steppe component and R1a in particular is the result of Indoscythians or whatever? Thats just clutching for straws with no other reason than the unwillingness to accept the obvious for ideological reasons.
The first fun aspect is that neither the Indus civilisation nor the Indoaryans can be explained by an independent Indian development. For both the decisive factor came from outside, from Iran for the IVC, and from Central Asia for the IA. I’m pretty sure the Dravidian language came with the Iranian farmers.
So the fight against AIT is just the Anti-European, anti-colonial and religously motivated sentiment.
Considering this, the 2nd fun aspect is that the situation is pretty much the same for Western Europe – and England in particular! Modern Europeans are mostly the descendents of Near Eastern Neolithic settlers and Indoeuropean steppe people. What was before, the hunter gatherers, left a mark but are a small minority genetically and gone culturally. The parallels are there and modern Europeans are, just like modern Indians (majority), the result of a parallel development from early Neolithisation to Indoeuropeanisation.
As for the (decisive) Indoaryan migration to India, the most common date in classical literature was about 1500 BC. But never earlier than 2000 BC. That doesnt mean that there were no earlier incursions, but the movements which changed the subcontinent and left the huge impact we see today happened most likely after 2000 BC. Again, the classic interpretation seems to be vindicated.
Even academicians post links to newspaper articles!
It is a simple enough explanation. Dr. Shinde is pleasing the political party in power by obfuscation. The report shows that the so-called Aryans were not part of Harrapan civilization; they came later, whether by migration or invasion. It blows up the OIT theory that Aryans were native to India and spread out from here which is a favourite theme of the Hindu right. No one ever doubted that the Harrapans were native to India, which the study confirms, as it confirms that the steppe genes came to India post the high noon of Harrapa.
@Obs
“I’m pretty sure the Dravidian language came with the Iranian farmers.”
Not really. Dravidians are genetically always identified as native to peninsular India (sengupta et al.2006).
Linguistically too there is no established link to Iran.
(Elam-Dravidian hypothesis is refuted by George Starostin)
I don’t know by which basis you claim that they arrived from Iran. Is there proof for your assumption?
Not really. Dravidians are genetically always identified as native to peninsular India (sengupta et al.2006).
you are out of date citing stuff from 2006. i will let Obs talk about the preprint….
The development of the IVC was happening in India indeed, but on the basis of the settlements from farmers which in turn came from Iran. These Neolithic settlers seem to have mixed with local hunter gatherers (ASI proper) and were very successful in developing a growing farmers culture which eventually developed the flourishing IVC. But all of that happened on the basis of the Iranian farmers and their culture – with local adaptations and Eastern influences of course. So we can’t be absolutely sure, the South Asian hunter gatherers could have been the language givers, but I seriously doubt it. Because there is a continuous development from the Iranian farmers, their settlemetns and villages to the IVC. Also, the typical modern Dravidian has a lower steppe, but a high Iranian farmer component. So the typical Indian Dravidian is still more West Eurasian than most tribals. Recent genetic studies, Razib already commented on those and made some valuable analyses, point to the main difference between Indoaryan and Dravidian: It is the ratio of steppe vs the Iranian farmer component if looking at the West Eurasian ancestry of Indians.
Both Dravidian and Indoaryan are Indian transformations of major West Eurasian impulses, one from Iran, one from the Eastern European/Central Asian steppe. Because almost no place in India is free from at least one of this two impulses influence, you don’t find any “pure” ASI in India any more. The true bottom line can only be drawn with more ancient DNA from India. I wouldn’t wonder if there were much more than 3 major migrations to South Asia before the Indoaryan consolidation.
@Obs
Forgive me if I’m being rude here but is there any need to stress the West part of the Eurasians who historically may really have given all their “impulses” to India? (This is categorically not meant to be a new Indigenous Dravidian theory or something but Dravidian languages are really not as decidedly shown to be of Iran origin as the origins of Indo-European languages outside India (not that it is bad and I’m not even being politically correct here- 4000 years is a very long time and all directions of Eurasians get nativised if they decide to make a place their homeland for such a long period); the fact that Dravidian is completely localised to India currently does not permit us to locate Dravidian homeland in Iran or such place with such ease in linguistics. There is lots and lots of new evidence required to conclusively prove such things at least as of now. A PDr. homeland in Gujarat or Maharashtra while descended from one of the AASI speeches of those areas is as possible as one in (eastern) Iran or in northwest India with ultimate origins lying with Iran_N.) Did not the similar subgroups of genetic West Eurasians as above give similar impulses to not East but other West Eurasians of pre-neolithic Europe? Is it a seemingly (to me) twisted way to come back at one of the traditional beliefs of the Indian civilisation that viewed the lands to the west beyond the northwest as “barbarian”? Or were the East Eurasian AASI somehow truly inferior to the West Eurasians?
Did not the similar subgroups of genetic West Eurasians as above give similar impulses to not East but other West Eurasians of pre-neolithic Europe?
what are you talking about?
@Razib Khan
“what are you talking about?”
I thought I had deleted that comment of mine- did I not do it? Anyway, no problem.
I was having in mind the WHG when I said that. Weren’t they another West Eurasian population like the neolithics of Anatolia, Levant and Iran? Did the latter not have similar effect on WHG to the one Iran_N had on AASI (culturally, etc.), as the commentator Obs himself/herself seemed to have written in an earlier comment? I may be even wrong in believing that the fates of WHG were similar to that of AASI, but I thought that that was the case at the time of writing that comment. And I felt my comment was a bit of an unwarranted overly emotional reaction as perhaps there was no much “stress” put on “the West part of the Eurasians who historically may really have given all their “impulses” to India” (quoting myself), on the part of Obs to begin with, and that it may have been entirely my own unnecessarily convoluted reading of the comment, so I thought it was best to delete it.
@Obs You don’t quite come out and say it, but you strongly imply that IVC was a linguistically Darvidian culture. But, there is very little to support this hypothesis. It is not a fit with the linguistic substrate of early Rig Vedic Sanskrit. It is not a fit with the reality that trade ties between the IVC and the people of South India were very thin. It is not a fit with the lack of Dravidian toponyms in the region where the IVC flourished.
The IVC language very likely did have deep roots with the pre-IE language of early Neolithic Iranian farmers. But, since the IVC language was very likely not Dravidian, this provides little or no insight into the source of the Dravidian language.
The other difficulty is that given the very modest amount of variance between the Dravidian languages, there are really only two viable possibilities. One is that the Dravidian languages are quite young (too young to be an IVC language since that civilization predates the South Indian Neolithic which is likely associated with the expansion of the Dravidian languages by thousands of years), and the other is that the Dravidian languages experienced an extreme bottleneck that extinguished all but one of the languages in the language family (possibly due to Indo-Aryan expansion), but that this one remaining Dravidian language then re-expanded into a larger part of the pre-bottleneck range of the Dravidian language (in which case our ability to learn about the deeper prehistory of the Dravidian language family was probably forever lost in the bottleneck event).
Many First Farmer Neolithic Revolution languages that expanded are now entirely lost and extinct so it is hard to know where they come from, even though we have a much better idea about where post-First Farmer languages come from.
It seems as if a transition from hunter-gatherer to herder doesn’t necessarily involve language loss and is in general less traumatic in terms of cultural change, while a transition from hunter-gatherer to farmer often just doesn’t happen at all and involves profound cultural changes when it does that often involve language shift to the language of the people who are the source of the farming technology. One would assume that first farmer languages arise from the rising dominance of whomever ends up developing farming and expanding first, such that, for example, the language of the LBK culture may well have been derivative of some exceptionally lucky hunter-gather tribe in Anatolia that was the first to master farming.
OTOH, even first wave Anatolian farmers in Europe show a diversity in mtDNA that suggests that this group might have been a blend of several distinct hunter-gatherer populations which may have each had different languages from communities that perfected different parts of the European Neolithic package that eventually fused into a single technological package, in which case it wouldn’t be too surprising if a typical first farmer language was some sort of creole of the languages of contributing hunter-gatherer populations.
I thought I had deleted that comment of mine- did I not do it? Anyway, no problem.
holy shit! you did delete it. i thought it was inadvertently placed in there by me or the software.
@Santosh: The main difference between WHG and AASI in the context of Neolithisation was that the AASI had huge advantages in comparison, especially numbers and immunology, disease resistance. The Indian AASI hunter gatherers might have been one of the largest HG populations in the world.
Through mixture with the AASI population the Iranian farmer’s population gained local adaptations which helped them to survive and spread on the subcontinent fast and successful. While in the process the surviving AASI element was much larger percentage wise than WHG in Western Europe.
As for the Dravidian and IVC issue and Iranian farmers: Obviously IVC and Dravidian are two different problems. IVC could have been speaking the Iranian farmers language, while Dravidians were an independent development with another langauge coming from the AASI HGs.
Some problems with that idea:
– Usually the Neolithics made the cultural impression and gave on their language. We have a lot of cases of warlike pastoralists giving on their language, not just Indoeuropeans, but a very limited number of actual hunter gatherers having an impact on fully established Neolithic cultures.
– The Iranian element spread in the IVC and to modern Dravidians. From the genetic perspective, please correct me if I’m wrong, it seems that the very same impulse found in IVC people (so far) can be recognised in Dravidians. Actually the high Iranian and lower Indoaryan/steppe component is the most typical feature of Dravidians.
– What even more ASI like populations speak today is not important, because it has the same reasons as Bhils speaking Indoaryan: Total cultural dominance of the surrounding cultural element. What Bhils and Gond speak today is the result of the Indoaryan and Dravidian spread on the subcontinent.
– Also, there are attested Dravidian influences on Indoaryan, even in its early stage. That the IVC had a big impact on IA is obvious, so if Dravidian was “somewhere in the jungle” or the South in general, how could it have such an influence on IA centers of the North?
Dravidians must have been wider spread and it seems they spread with the Iranian farmers. I see only arguments in favour of that explanation and little against. Alternative explanations for the current distribution of genetic signatures and the impact of old Dravidian on the earlist Indoaryans are weak.
@Obs
I tend to agree with many things regarding your reply about the Dravidian languages. But there are still many problems in my view. First is that the Indian Iran_N being noticeable from at least 6500 BC or so (if the usual thinking that Mehrgarh Neolithic is ultimately derived from the neolithic cultures to its west in (eastern) Iran is correct), and being such an ancient population, if not already speaking several divergent languages, could have at least given rise to more divergent branches and perhaps even distinct language families than the one Dravidian, in the subcontinent. Coupled to this is the fact that Proto-Dravidian is a very young language, as ohwilleke also notes above: all the extant Dravidian languages are very closely related; they are perhaps only slightly more divergent than the dialect continuum-like situation seen in Indo-Aryan-speaking north India today. In fact, the major anchor used in the dating of Dravidian languages is the reference to a group called “Andhra” in the 7th century BC Sanskrit text Aitareya Brahmana and subsequent references to both “Andhra” and “Dramila” in the 4th century BC text Natyashastra, construed as referring to Pre-Telugu of Telugu-Kui subgroup and Pre-Tamil of Tamil-Tulu subgroup respectively, which when also coupled with internal linguistic evidence and estimates of divergence, puts the Tamil-Tulu and Telugu-Kui split at 1100 or 1000 BC. Now the more divergent older branches of North Dravidian and Central Dravidian are also not dated much older than this split- Bh. Krishnamurti in his “The Dravidian Languages” dated their first split sometime around 1500-1300 BC. This means that undivided Proto-Dravidian continued to be spoken till around that late period of perhaps 2000 BC to 1500 BC and did not diverge earlier, and definitely not in any much more older time frames which we typically consider associated with the spread of Indus neolithic into the northwest. One possibility that I have in mind which may explain such closeness of Dravidian languages while still first spread during the Indus neolithic and associated with the Indus civilisation, is to imagine a linguistic homogenisation process in the Indus civilisation achieved by the selection of a particular Proto-Dravidian dialect (the MRCA of Tamil-Tulu, Telugu-Kui, Kolami-Parji, Kurukh-Malto-Brahui) out of several, with all others on the process of extinction already during the Indus civilisation itself. The other possibility of recapturing captured Indo-Aryan territory by a surviving bottlenecked Dravidian dialect as imagined by ohwilleke is very interesting though as far as my very amateur linguistic knowledge goes, there seems to be no discussion of any Indo-Aryan substrate or of discussion of language shift of Indo-Aryan speakers to Dravidian in this scenario, in Dravidian linguistics.
Secondly, there is also the phenomenon of the presence of non-Indo-European words in Rig Veda which do not seem to be derivable from Dravidian, Munda and Burushaski sources. There is also the Language X of the Gangetic plains that contributed agricultural vocabulary to Hindi and this Language X may not be have been that confined to the Gangetic plains if Franklin Southworth’s view that many of these words from Language X are also present in other IA languages like Marathi, etc. and thus may be taken as the evidence for the consideration that Language X may have been more widespread in the Indo-Gangetic plain is correct.
Thirdly, the old Dravidian ethos as seen from linguistics and philology seems very different compared to the Indus ethos (to me only perhaps though). Proto-Dravidian seems like a language spoken by pastoralists and people practicing subsistence agriculture, rather than like a language of surplus-producing agriculturalists. And Dravidian definitely did not seem to retain any direct memory of big cities and such (though, as per Southworth, there seems to be evidence for contact with an advanced urban civilisation)- the borrowing of the word for ‘brick’ from Indo-Aryan is perhaps one of the most interesting things in this regard. But I agree that this may be a very weak argument. In this context, it is interesting to note that the agriculturalist castes of south India, like Panta Kapu, Reddy, etc. are being modeled in genetics forums as very close (more than 80%) to the Shahr-i Sokhta BA3 individual who had 60% Iranian agriculturalist ancestry and 40% AASI. It is very tempting to associate these very castes as responsible for bringing Dravidian languages to south India as they sit just below Brahmins in the social ladder of south India and their Iranian agriculturalist ancestry being 20% higher than their AASI ancestry seems to support large scale migrations from more northwestern locales to south India rather than an Indo-Aryan-like spread scenario with elite dominance and all that (Swat Valley samples not showing large amounts of steppe ancestry). While better population sources for south Indian middle castes like above may be found in the Deccan as time passes and SISBA3 may become outdated as their source, I don’t know to what extent, being not knowledgeable in population genetics to any significant degree, the new sources would differ from SISBA3, especially in the Iran_N-AASI proportions.
And as you noted, the strongest argument for Dravidian being widespread in the Indus civilisation (perhaps while also being present in the south of India already; perhaps also not) at the time of arrival of Indo-Aryan languages, seems to be the Dravidian-substratum-in-Sanskrit argument. I observed that a dominant view among many Dravidian linguists is that many non-Indo-European grammatical features of Sanskrit like retroflexion, the use of gerund, placement of quotative marker after quoted speech, etc. are best explained as arisen as a result of large numbers of Dravidian speakers shifting native languages to Indo-Aryan. Again, as you noted, there are alternatives to this substratum hypothesis too, but they don’t seem to enjoy a lot of support, at least currently.
@ohwilleke
Though you did not address your comment to me, I would like to mention that I tend to agree with several points you make, except for some one of which is the association of Dravidian with the Southern Neolithic. As it may perhaps be noticeable from my above paragraphs too, I’m of the view that Proto-Dravidian is younger than Southern Neolithic as well and that it perhaps did not begin to diverge at the beginning of Southern Neolithic which is at around 2800 BC. I’m aware of the views of Dorian Fuller and Franklin Southworth but I personally believe Krishnamurti to be correct in dating the split of North Dravidian from Proto-Dravidian at around 1500 BC and not around 2500 BC or earlier. So I consider the late divergence of Dravidian as evidence against associating it with at least the very earliest phases of Southern Neolithic, which is characterised by pastoralism of zebu cattle likely ultimately from the Indus area via the western savanna in Gujarat as Fuller thinks and the ashmounds,
in addition to its association with at least the Indus neolithic and Early Harappan phases and the Indus civilisation, if we go by the assumption that divergence of the language inevitably results and continuously increases with the first spread of the culture of the neolithics, be they of early neolithic Indus valley in 6th millennium BC or of Southern Neolithic in early 3rd millennium BC.
Then, I also respectfully request both you and Obs to comment more on these various possibilities (the newest being the idea that I mentioned above- a scenario of a deliberate selection and spread of a particular Dravidian dialect within IVC and the subsequent extinction of the other divergent contemporary Dravidian dialects) to explain such close relatedness of Dravidian languages, if possible.
@Santosh:
“The other possibility of recapturing captured Indo-Aryan territory by a surviving bottlenecked Dravidian dialect as imagined by ohwilleke is very interesting though as far as my very amateur linguistic knowledge goes, there seems to be no discussion of any Indo-Aryan substrate or of discussion of language shift of Indo-Aryan speakers to Dravidian in this scenario, in Dravidian linguistics.”
From my point of view the most likely scenario is that IA overrun or at least assimilated most Dravidian unities, until they reached the tropical part of India and met fierce resistance by a Dravidian stronghold – with which they made an arrangement ultimately, which would have been not that big of a deal, because they made arrangement with more Northern Dravidian elites too, namely those which were assimilited but not put on the low end of the social and later caste spectrum.
It is definitely no coincidence that Dravidian is mostly South of the tropical border, while Indoaryan controls the dry and subtropical parts of the subcontinent in a more mixed form, while the most steppe like people live in the mountainous areas. These were natural borders which slowed the expansion down and allowed local people to prepare or make a deal. With mixed ones having an advantage for climate and disease resistance even. This was not so much the case in the far North and the mountainous areas, where both the steppe people and the Iranian farmers largely replaced earlier groups.
Because of the bottleneck, Dravidian lost most of its diversity. Again it seems improbable that the timing is a mere coincidence. The major Indoaryan push seems to date to about 1500 BC and the bottleneck of Dravidian being dated to the same time window!
As for the complexity of the Proto-Dravidian language: It seems to have all the vocabularly for a surplus producing, complex society. There are terms for taxes, laws, kings or chiefs and so on. Thats what I read at least, and it seems plausible to me. I’m no expert on that subject as you might have noticed, but some experts say so.
Indoaryans could have been influenced by BMAC or another farmer’s culture, we don’t know. Even if the IVC or part of the IVC people spoke a different language, Dravidian would have been still most likely coming with the earlier Iranian farmers. This would just mean that in the Indus area Dravidian was replaced by something unknown in later times. But that would still relate Dravidian te the earlier Iranian farmers.
We would need to decode IVC texts, if there are real texts to begin with. But if that’s ever possible, I don’t know.
Dravidian must have been older and more diversified, related to other languages from the wider regions, that’s for sure. So the rather limited diversity can best be explained by a bottleneck scenario caused by the Indoaryan conquest. The remaining core group re-expanded only into some IA territories, but was most successful in the tropical part of the subcontinent where it spread rapidly. This fits the low PD age perfectly imho.
@Obs
I agree that a real bottleneck scenario that may have decreased the diversity of Dravidian languages on the Indian subcontinent level is quite possible. But what I’m more reluctant to consider is that even Dravidian in south India was the result of a recapture of south Indian territory captured by an Indo-Aryan advance. There seems to be no evidence for Indo-Aryan substrate influences or such things in the extant Dravidian languages- at the earliest levels of contact, which may have taken place in the Proto-Dravidian period itself as far as the extant languages are concerned, the IA influence seems to be adstratal and later full-on superstratal with some forms of IA like Sanskrit and High Prakrits and Pali and adstratal with spoken Prakrits.
Thus, if we are to reject this type of a possibility, we are left with the problem of low diversity in the Dravidian languages of south India of the time, which we expect to have been considerably higher if the south Indian (and currently the only extant) Dravidian was established there quite early, like in the 3rd millennium BC. A stronger possibility seems to be that an appropriately bottlenecked Dravidian itself arrived in south India first at the same time as north India was getting Indo-Aryanised (sometime between 2000-1500 BC); from somewhere in Gujarat and Maharashtra perhaps which were probably not under too much Indo-Aryan influence at that point.
Regarding the reconstruction of elements like ‘tax’, etc. to Proto-Dravidian that you mentioned, I have noticed a dilemma in Dravidian linguistics- whether it is possible to reconstruct these words both with their forms and meanings, to Proto-Dravidian. The problem arises because of several reasons- in many cases, these types of words are found only in the South Dravidian languages or even just Tamil-Tulu (South Dravidian-I) languages or even just the four literary languages with the exclusion of all the other Dravidian languages including Tulu (an example is a word for ‘law’: caTTam at Dravidian Etymological Dictionary 2304 has ‘law’-related meaning only in Tamil, Malayalam, Kannada and Telugu along with the more physical ‘wooden frame’, ‘bed’ type meanings more widespread and being the only variant present in Tulu; but apparently Kota has a word meaning ‘tribal custom’ though; I don’t know if it is considered a true cognate), who have been historically the most assertive and successful lot among the speakers of various Dravidian languages (in some instances, many non-literary languages of Tamil-Tulu subgroup carry cognates along with literary ones like Tamil and Kannada and which are seen in just the one Telugu of the Telugu-Kui subgroup, indicating that the Telugu words may not be true cognates but prehistoric borrowings). This above Tamil-Tulu subgroup people also have a history of engaging in large scale sea trade and forest product economy, relative to the Telugu-Kui (currently classified as South Dravidian-II) and other subgroups. In many other cases, North Dravidian languages like Kurukh do have formal cognates but don’t have the advanced meanings seen in South Dravidian or South Dravidian-I. An implicit assumption of those who still go ahead and reconstruct the advanced meaning to Proto-Dravidian in such cases seems to be that advanced meanings fell out of use for all such words in North Dravidian as time passed. I don’t know what theoretical basis there is for such a consideration and how such situations are identified in historical linguistics. Due to problems like these, we have two types of reconstruction: a Krishnamurti type which reconstructs to Proto-Dravidian forts, moats, kings, taxes, and other stuff including iron, and a Southworth type that does reconstruct much (but not all) of the above stuff but to Proto-South-Dravidian and not to Proto-Dravidian. (Very cautious considerations like above perhaps lead Southworth to consider the seemingly advanced vocabulary of his South Dravidian as an indication for contact with an urban civilisation which he identifies as Indus civilisation and thus puts a distinct South Dravidian already at such an older period of 2300-2000 BC and in the expanding territory of the Southern Neolithic; while I personally find this particular aspect of Southworth’s dating very attractive, overall linguistic features like development in verbs, etc. seem to indicate to me (on an individual level intuitively) (and perhaps also to Krishnamurti who just stated his opinion without much justification (or so it was apparent to me) that North Dravidian and Central Dravidian may have split not too long before the Tamil-Tulu and Telugu-Kui split of 1000 BC) that North Dravidian did not really already become distinct from Proto-Dravidian at such an early age of 2300 BC; and it is agreed by all scholars currently that North Dravidian split off at around the same time as South Dravidian and Central Dravidian split off and began to develop) The reality maybe somewhere in between these two types and we may not be able to arrive at it properly anytime soon, unless there is more research into North Dravidian languages and their Dravidian lexis collected again in the hope that the situation may improve.
And yes, I agree that Dravidian need not be the Indus language for it to be connected with probably an earlier batch of neolithics from Iran with the actual Indus civilisation people being a fresh later batch of neolithics from Iran- perhaps even a type of Sumerians (lol). But again, with pure linguistics much cannot be known it appears- more knowledge has to be tried to be gathered in such a case from genetics and other anthropological considerations that you mentioned, like pastoralists and farmers more likely to singlehandedly succeed in spreading language, and any type of Iran_N being more likely to be the early pastoralists and farmers in the case of Indian regions like Gujarat, Sindh (?), etc. which may be the relevant ones for Dravidian languages.
@Santosh:
“A stronger possibility seems to be that an appropriately bottlenecked Dravidian itself arrived in south India first at the same time as north India was getting Indo-Aryanised (sometime between 2000-1500 BC); from somewhere in Gujarat and Maharashtra perhaps which were probably not under too much Indo-Aryan influence at that point.”
That’s quite likely. We see something similar with the Thai conquests in South East Asia, resulting from Chinese pressure.
“An implicit assumption of those who still go ahead and reconstruct the advanced meaning to Proto-Dravidian in such cases seems to be that advanced meanings fell out of use for all such words in North Dravidian as time passed. I don’t know what theoretical basis there is for such a consideration and how such situations are identified in historical linguistics.”
That’s a possibility I had in mind and seems just logical, because language, especially if being spoken only, loses terms and meanings meant for things which are no longer in use. A good example for that are terms for parternal and maternal relatives from an agnatic time in various Indoeuropean languages, being given up when a cognatic kin systems and laws came up. The terms just got lost in the spoken language on the dialect level and only survived, if at all, in the written language.
Good post and I can only agree on what you have written.