search this blog

Tuesday, January 30, 2018

Indian smoke and mirrors


On January 4 this year Hindi newspaper Dainik Jagran published a so called special feature on Indo-European languages. In fact, the article claimed to be giving its readers a sneak peak at the results from the upcoming and much anticipated archaeogenetics paper on the northern Indian Harappan site of Rakhigarhi. [LINK]

I knew about this article when it first came out, because it was mentioned in a few off topic comments on this blog, like this one by commentator Sanuj.

Latest news on the Rakhigarhi results, published in a prominent Hindi paper, also quoting Niraj Rai, the lead geneticist working on it. It is essentially saying that researchers, both foreign and Indian have established that India is home of the Indo-European family and that the aDNA from Rakhigarhi is a close match to North Indian Brahmins. The results are to be published in a leading journal soon.

I deleted these comments soon after I saw them, not only because they were off topic, but also because they made absolutely no sense whatsoever.

Why? For one, because over the past year or so I've managed to gather a little bit of intel on the Rakhigarhi paper from very reliable sources, and all indications were that the results would show significant ancient population movements from West Asia and Eastern Europe to India, and not the other way around.

Moreover, thanks to already published ancient DNA from outside of South Asia, it's obvious that there were significant population movements from West Asia and Eastern Europe to India, and not the other way around. The one exception to this rule is the migration of the Romani (Gypsy) people from northern India to Europe, but this is irrelevant to the topic at hand, because it didn't have much of an impact on the genetic structure or linguistics of Europeans.

So why have I now decided to give Dainik Jagran my full attention? Well, because commentator Sanuj recently resurfaced in another comment thread and said this...

They have been hinting at the outcome, you are just not ready to listen to what they are hinting at, this Jagran article is a case in point. By the way Jagran is the most widely read newspaper in India, and is one of the most credible - rated by Reuters-BBC.

Yep, he's correct: Dainik Jagran is a huge and well respected newspaper.

Please note, however, that the chances of India being confirmed the Indo-European homeland thanks to the ancient DNA from Rakhigarhi are zero; not just low, not almost zero, but zero. Anyone with a generally healthy mind and the ability to be more or less objective in this matter has to admit that this is indeed the case. So why would one of the biggest and most respected Indian newspapers publish such utter crud?

It's an intriguing question to say the least. Moreover, was Niraj Rai actually interviewed by the reporter from Dainik Jagran? If he was, did he really say what he's claimed to have said, or was he grossly misrepresented? If the latter, has he sought a correction? If not, why not? Have the western scientists who are collaborating with Rai asked him what the fig is going on, and have they sought a correction? If not, why not?

Does anyone know if Dainik Jagran has since published a correction, or at least a letter from Rai straighting things out?

Admittedly, I have no idea what's going on now with the Rakhigarhi study and paper; the trail went cold months ago. But whatever it is, it's something peculiar. That's because I find it extremely unlikely that any newspaper, let alone one of the top newspapers in India, would be allowed to get away with misrepresenting and indeed inverting, either by design or mistake, the outcome of such a major international project.

See also...

Indian confirmation bias

The Out-of-India Theory (OIT) challenge: can we hear a viable argument for once?

Late PIE ground zero now obvious; location of PIE homeland still uncertain, but...

317 comments:

«Oldest   ‹Older   201 – 317 of 317
Nathan said...

@Nirjhar007
I have read the conclusions of Western Sanskritists and Indologists, Michael Witzel being one, and they say the society of the Rigvedic Aryans and the society of IVC were totally different.

Tom Rowsell said...

@Acharya Agnimitra

"Sounds ludicrous?Impossible! Right? Such a thing has never happened and never can."

May I offer the example of Anglo-Saxon invasion and the English language. The number of words in English of Brittonic origin is very small and many of these may be later borrowings from Irish or Welsh. Likewise the number of Celtic placenames in England is very small, especially in the East, but is quite high in the West Country. Even the Celtic placenames often have English suffixes or prefixes eg. Gloucester - derived from the Anglo-Saxon for fort (Old English ceaster) preceded by Celtic name, which derived from the Roman stem Glev- (pronounced glaiw).

A similar process must have occurred in India as in England.

postneo said...

@shah
Your translation is not proper. heres a better one. I dont need google

कि राखीगढ़ी में मिले ये कंकाल उन प्रजातियों के पूर्वजों के हैं जो इंडो-यूरोपियन भाषा परिवार की भाषाओं के वक्ता हैं और दुनिया में स्वयं को सर्वश्रेष्ठ प्रजाति के रूप में घोषित करने का दावा करते रहे हैं

that the skeletons found in rakhigarhi, are ancestral to those ethnic groups that are speakers of IE(today) and proclaim themselves as preeminent representatives(of these)

hmmm, so technically it does not claim OIT .... anyway the article is not worth so much attention

Chetan said...

Jaydeep's translation is pretty much accurate, word by word.

Chetan said...

Although it's not clear who the authors meant by "those making the claim of being the greatest race in the world". A reference to Brahmans? Because no Brahmin writings speak of themselves in racial terms. It's always a description of their spiritual excellence above the rest.

Rob said...

^ It must mean one of the most wide-spread lingusitic groups of the world - IE speakers.

Chetan said...

@Rob Certainly could be a reference to early 20th century racial views.

Rob said...

I don't know. In English at least, "Great' means epic , numerous , large. Like "Great War" = WWI.

Shahanshah of Persia said...

@postneo I didn't use google translate and Jaydeep's basically said what mine said but better. I trust Jaydeep's translation. The person who translated it for me is also a Hindi speaker.

postneo said...

Chetan its a poorly drafted sentence not gone through editing, the literal translation is meaningless unless they mean ancestral to the nazis.

also you have करते रहे हैं m... have been proclaiming David? just kidding... refer to my translation

Shahanshah of Persia said...

@Tom Rowsell The Anglo-Saxon invasion is not a good analogy in this case because unlike the Anglo Saxon invasion, the Aryan invasion left a huge impact on South Asian genetics. Though I highly doubt that the Vedic Aryans were purely steppe derived, I think that it's safe to say that the elites were 80% Steppe EMBA/Yamnaya/Poltavka like, with the rest being Neolithic Iranian and East Asian DNA, possibly. Tough to say for now. It's worth waiting and seeing what the Swat samples tell us. But David also agrees with my interpretation here. Now, let's wait and see. Modern South Asians are best modelled as Steppe EMBA not MLBA. I'm still wondering why that is, thoughts?

Davidski said...

The race they're claiming to be talking about is the white race. In other words, Europeans.

They mention Europe in the headline in a rather bizarre way, and I don't think that's the fault of the translation, and then also a few European countries in the text.

Nirjhar007 said...


Good translation from Jaydeep . But...instead of ''race'' perhaps generation or as postneo say ethnic group is more accurate :) .

Nirjhar007 said...

Although it's not clear who the authors meant by "those making the claim of being the greatest race in the world". A reference to Brahmans? Because no Brahmin writings speak of themselves in racial terms. It's always a description of their spiritual excellence above the rest.

No chetan its a reference to the Indo-Europeans . prajAti don't mean ''race'' :
http://spokensanskrit.org/index.php?mode=3&script=hk&tran_input=prajati&direct=au

Shahanshah of Persia said...

Another thing I wanted to say here is that Google's translation of this article is fairly accurate. I would say 90 to 95% accurate. So, there's a clear narrative being spun here by Jagran. Whether or not it's based in any truths or hard evidence will be seen when the study is out, though I highly doubt it. Out of India is not possible, based on the evidence we have, and even the slight chance of Steppe admixture being already present in the Harappan samples does not at all confirm the Out of India scenario. Regardless, I will say that this is highly unlikely, and in all likelihood Steppe ancestry arrived in India during the Late Bronze Age, from a population originating on the Andronovo horizon and resembling Steppe EMBA populations. There was also a clear attempt here to vilify Europeans, I will add.

Nirjhar007 said...

I have read the conclusions of Western Sanskritists and Indologists, Michael Witzel being one, and they say the society of the Rigvedic Aryans and the society of IVC were totally different.

You have to read the book , I used to do the same mistake .

Nirjhar007 said...

The Jagran article has IMO these main points :

1. People of Rakhigarhi had dna which can be considered as ''ancestral'' of Indo-Europeans today(that refer to the IE's worldwide).

2. The people had ''strong'' affinity with modern North Indian Brahmans .
..............................................................................................

But as said earlier the reporter got excited :) .

EastPole said...

@Jaydeep
“As soon as this discovery is published in International Journals, not only will there be a start of a new era of discussion, debate and brainstorming across the whole world but there is also going to come about a great change in the global history of human civilization.

From the discovery that has been achieved by the other scientists from different countries in Dr Rai's team, it has been found that the skeletons discovered from Rakhigarhi are the ancestors of those races who are the speakers of languages of the Indo-European language family and who keep making the claim of being the greatest race in the world.”


The question of “being the greatest race in the world” is no doubt linked with the question of “which culture was the origin of the world’s civilization”.

It is a very difficult question as the origin of religion and philosophy is debated:

https://s17.postimg.org/jd9tl9a9b/screenshot_337.png

“Universe and Inner Self in Early Indian and Early Greek Thought”

http://www.jstor.org/stable/10.3366/j.ctt1bgzdmh

No doubt that Greeks and Indians contributed the most, and there were some mutual influences as well as parallel developments, but I also think that elements of their religion, poetry and philosophy could have a common source:

http://eurogenes.blogspot.com/2017/10/tollense-valley-bronze-age-battle.html?showComment=1509036279732#c7168224625981191837

Ric Hern said...

Will there be claims that R1b people adopted Indo-European from R1a people ?

a said...

Ric Hern said...
Will there be claims that R1b people adopted Indo-European from R1a people ?

That will be a real headache in a narrative and chronological context. R1b-Z2103 is older than R1-Z93 5000+/- YBP[burial around Sintashta-Arkaim are stratified with older Yamnaya predating]. Majority of Yamanaya are R1b-Z2103-5700+/-YBP. They also practiced wagon burials, logically they must have had a word for wheel, replacing the vocabulary from India is rather problematic since some of the oldest known uses and burials and or of wheels are if Europe and Maykop.
Take Z2108-5700+/-[present in Dagestan and Gujarati] for example, found in Yamnaya under KMS67+
https://www.yfull.com/tree/R-Z2108/

https://www.yfull.com/tree/R-Z645/

Ric Hern said...

Can someone explain the process by which Indo-European apparently spread from India ? When, with who(DNA path) and through which Cultures did it spread all the way to Ireland ?

Chetan said...

@Ric Hern If Indo-European spread from India, then Vedic Sanskrit would have been the first to separate. That is so clearly not the case. And there is no evidence of any other IE language in India other than the Sanskrit-derived ones. A lot of fuss was made because of the presence of a few centum words in the language Bangani from the Himalayan foothills. Linguistically, that's all there is as evidence for an Indian PIE origin. In the case of the Anatolian hypothesis, at least there is a coherent model. None exist in this case.

Ric Hern said...

I understand Chetan.

Then who here is Pro-Out Of India that can explain their theory ?

From India to Ireland...

Salden said...

Again, there isn't enough South Asian DNA in West Eurasia to support Out Of India.

Jaydeep said...

Ric Hern,

For a linguistic model have a look at this -

https://id.scribd.com/doc/150388032/The-Epicentre-of-the-Indo-European-Linguistic-Spread-Useful

For some archaeological evidence have a look at this -

https://books.google.co.in/books?id=8a00AAAAQBAJ&pg=PA114&lpg=PA114&dq=mariya+ivanova+maikop+central+asia&source=bl&ots=_9-3rzpsXP&sig=tQScWsjZOtjyBcXjk5OzVvtLPg8&hl=en&sa=X&ved=2ahUKEwigqc_Ak4fZAhWGv48KHbSCBhEQ6AEwC3oECA8QAQ#v=onepage&q=mariya%20ivanova%20maikop%20central%20asia&f=falseYou




Ric Hern said...

@ Jaydeep

So the Out of India theory hangs upon a Maykop tread ?

So do you think R1a spread from India to Maykop and then onwards ?

Or do you think PIE was spread with Haplogroups L or J ?

Matt said...

@Ric: There is, as far as I can tell, no model for out of India.
There's no archaeological trail out of India, during the Bronze Age as would be supported by linguistic consensus, or earlier (the scanty archaeological evidence for a migration into India is plentiful by comparison with the obverse; advocates of OoI do not hold their evidence to the same standard, as mainstream scholars do).

There are no divergent early branching clades within India (unlike elsewhere, depending on the model). Nothing.

Whatever the unresolved questions of a model of expansion from somewhere between the Caucasus to Russia, India really offers no alternative (unless something *actually* comes up in the future).

There's nothing that anyone seems to point to other than "Well, there are some gaps in the trail from putative IE cultures on the steppe and it's impossible that a large urban civilization could shift languages to that of a smaller, more nomadic group (never mind that exactly this happened in Iran, Anatolia and later with the expansion of Arabic.)". No positive evidence, just claims of the absence of evidence for the dispreferred origin.

Ric Hern said...

@ Matt

Yes I understand

I'm just trying to figure out what they are thinking and what proof they will need to sway them.

I personally think even DNA evidence will not be enough. They will simply trade R1a for L or J to try and keep their point of view alive.

But maybe I'm wrong ?

postneo said...

" personally think even DNA evidence will not be enough"

what you people do not get, is that R1A, L, R1b have never been enough, its never been about that. AIT vs OIT debate is much older than genetics and has not been impacted much by it.

Also it is not the "them" that need to be swayed. AIT was mainstream even when R1a was said to have greater diverity/depth in India than Europe. The theory did not crumble because of genetics.

So if anything you need to convince mainstream indologists snd linguists from the west that DNA matters.

Dmytro said...

AIT is a scientific hypothesis, with lots of evidence but some remaining gaps. In Ockhamist terms it is not a "demonstrated" but it definitely is a "probable" theory. OIT is not even an "apparent" theory on those terms. It is a religion, and as such is incapable of being refuted. Discussion is pointless.

Santosh said...

@ Ric Hern, Matt and Jaydeep regarding the immediate few posts above made about the OoI models current

First of all, thanks very much for the above discussion!

It has been my personal imagination for quite some while regarding the broad umbrella of "indigenous" Aryans outlooks is that it has a place for a "remotely-indigenized/favorably-indigenized" Indo-Aryans as much as for "completely indigenous" Indo-Aryans. That is to say, a scenario that credits Indo-Aryans/Indo-Iranians for the Indus Valley Civilization even if it involved a migration of city-building Indo-Iranian/Indo-Aryan speakers from anywhere else, be it Maykop or Pontic-Caspian steppe, etc., just around the time of the transition of the neolithic stage Indus Valley to urban phase Indus, is also deemed acceptable perhaps. (This may be something to do with any attachments to the Indus Valley Civilisation that most Indians now infer to have been played quite some significant role in shaping the history of the subcontinent (at least successfully enabling the country to transition to the neolithic phase) on the part of all the powerful linguistic families of India (majorly Indo-Aryan and Dravidian); this may not be relevant here, but I just wanted to note what I have been thinking.) So, in that scenario, an IE family disseminated with the help of a few men carrying J or L into both Europe (I don't know and I haven't imagined too, the explanations posited for the subsequent transformations required in Europe) and northwest India from any intermediate location may also be considered favorable. R1a wouldn't absolutely be necessary, I suppose, in such a scenario. I don't know anything about how archaeology works in that regard too; perhaps the potential involvement of Maykop comes into picture here.

postneo said...

There is no archeological trail from the steppe to either Greece, Anatolia, Iran or India. We need to focus on that first not this shrill obsession with OIT. OIT is a bogeyman created by “you” folks. If you want to understand the actual stance of the anti AIT crowd it’s always been show me evidence for AIT and sometimes follow up with ...if that’s the quality of evidence We can also similarly push OIT but not intended as mainstream. Read Edwin Bryant who has done the in depth analysis of the debate.

Santosh said...

@postneo

"Also it is not the "them" that need to be swayed. AIT was mainstream even when R1a was said to have greater diverity/depth in India than Europe. The theory did not crumble because of genetics.

So if anything you need to convince mainstream indologists snd linguists from the west that DNA matters."

I agree. As a believer in the mainstream model based on linguistics and having found myself sufficiently convinced by my information (although personally not thoroughly evaluated) regarding the aDNA evidence becoming available since a few years that I perceived as pointing towards the late chronology and against the early chronology, I once emailed an Indologist singing the glories of aDNA and how it has "solved-off"/ is close to "solving off" the IE problem and got an extremely measured response outlining several of the problems perceived to be associated with a genetics based view of these matters and no indication of a positive or enthusiastic appreciation of my point at all.

Santosh said...

@postneo

"If you want to understand the actual stance of the anti AIT crowd it’s always been show me evidence for AIT and sometimes follow up with ...if that’s the quality of evidence"

Please correct me if I didn't understand you correctly, but the Indo-European problem needs to be addressed, no, some way or the other? That is, some sort of alternative models (internally developing and updating themselves) have to be tried to be generated and presented as contenders for the mainstream model at all points in time (or most points in time), no?

Ric Hern said...

@postneo @Jaydeep @Dmytro @Santosh

Thanks for clarifying things for me.

This will always be a Never Ending Story.

Nirjhar007 said...

Tom,
The example of Anglo-Saxon England is an interesting example :),but the demographic and geographic situation was very different. England is small and many Latin/Roman loanwords , especially names of towns, etc. remained, including Chester, while in Sanskrit and ancient North India we dont find similar traces of a lost language .
https://en.wikipedia.org/wiki/List_of_Roman_place_names_in_Britain
https://en.wiktionary.org/wiki/Chester

Matt said...

@Santosh, thanks. Re: ideas of Indo-Europeans as having an early presence in Harappa, as one of the contributing streams of ancestry and language (if not necessarily *the* culture and *the* language) coming in from elsewhere... with the proviso that I've not thought hard about this - I think at least there would be the problems for some linguistic trees (Indo-Aryan too late in Ringe's model), for some models of Indo-European common culture traits (with the reconstruction of early Indo-Aryan speakers with supposedly quite different cultural traits than the IVC) and of course the same archaeological ideas are a problem (is there a decisive archaeological trail between IVC and elsewhere, coming in?).

But at the same time, there is enough ambiguity in the linguistic trees that I'm not sure it's totally non-viable, on that consideration alone at least - Garrett-Chang's tree places divergence of Indo-Iranian at 3000 BCE and Indo-Aryan at 2000 BCE. (There is obviously something to be resolved in which Ringe's model has proto-II split off last and Garrett-Chang's split it off almost first from "Late PIE"). Genetically there is no reason to exclude it on the basis of what we currently know on the autosome, though the R1a divergence dates may be wrong...

(An implication would be that the Andronovo / Sintashta / Potapovka group would probably not be speaking anything Indo-Iranian, and we'd have no idea what they spoke!).

Al Bundy said...

The Russian steppes spread BaltoSlavic,Germanic and Italic and Celtic.Greek we need more samples, let's see what the Indian paper says.Anatolian adna is obviously huge.The Anatolian farmers spread too early and India is not the ultimate PIE homeland, that would be the Pontic Steppes or Caucasus.If IndoHittite only came from the Caucasus and nothing else I guess you can argue about calling it the PIE homeland or the IndoHittite homeland.I think that's basically where we are.

Anonymous said...

@Ric Hern

"This will always be a Never Ending Story."

With a bit a bad luck they'll have a Y-DNA R which they can't resolve any further. Or even worse: with both R2 and R1 markers, like Mesolithic Iboussieres from Mathieson having marker for J and I.

Anonymous said...

Most of the AMT/AIT posters here are not in favour of OIT. Their scenario more or less fits with an intermixing of ANE with an Iranian Plateau or Middle Eastern population in South Central Asia, that then entered the NW subcontinent to create the IVC.

Nirjhar can correct me, but her argument has been that there wasn't actually a steppe movement into South Asia. Instead, just like in the steppe, there was an admixture of ANE with CHG that produced a population that looked like EHG/Yamnaya, but wasn't.

I personally go with an early Steppe origin Indo-Iranian arrival in South Central Asia, a first pulse of I-A migration across Iran and the Indus Valley between 2500 BC and 2000 BC, with the Iranian speakers of Central Asia forming a second pulse of migration into the Iranian Plateau and back into Eastern Europe.

EastPole said...

@aniasi
“I personally go with an early Steppe origin Indo-Iranian arrival in South Central Asia, a first pulse of I-A migration across Iran and the Indus Valley between 2500 BC and 2000 BC, with the Iranian speakers of Central Asia forming a second pulse of migration into the Iranian Plateau and back into Eastern Europe.”

And what is the evidence for that migration from Iranian Plateau back into Eastern Europe. Could you link some genetic studies?

Anonymous said...

@Eastpole

I meant from Central Asia back into Eastern Europe. Scythian, Sarmatian, Cimmeria, Alan and other East Iranian speaking groups are the evidence.

Rob said...

@ Matt

Slightly OT, but you should read “Emergence of the Ideology of the Warrior in the Western Mediterranean during the second Half of the fourth Millennium BC” by C Jeneusse, if not done so. It’s an accurate piece of work which overcomes some of the errors and limitations of arguements available on anglophone literature.

Rob said...

@ aniasi

“Most of the AMT/AIT posters here are not in favour of OIT.”

Have you missed a word ??
Of course people who favour AIT don’t favour OIT !?

Anonymous said...

@Rob

Meant Anti-AIT. Oops.

EastPole said...

@aniasi

“I meant from Central Asia back into Eastern Europe. Scythian, Sarmatian, Cimmeria, Alan and other East Iranian speaking groups are the evidence.”

I do not understand the linguistic processes in Asia. I used to think that migrating Andronovo tribes were Indo-Iranians. Those I-I who entered India and mixed with locals became Indo-Aryans, and those I-I who entered Iran and mixed with locals became Iranians. Are you suggesting that Iranians existed before entering Iran and local not-IE population of Iran had no influence on the ethnogenesis of Iranians? And also that local not-IE population of India had no influence on the etnogenesis of Indo-Aryans? This would be strange as the norm is that languages change the most when people mix.

Acharya Agnimitra said...

@Tom Rowsell

Likewise the number of Celtic placenames in England is very small, especially in the East, but is quite high in the West Country.

Right? As you would agree, there are still Celtic and Latin place names that survived the Saxons. Why look any further than the River Thames!? The Celtic river Tamesas survived the Romans as well as the Saxons.

A fun fact- There is a river Tamasa even in India. A tributary of the Ganges. "The dark river" they both mean. The Celtic and the Vedic share even a river. Fascinating, no?


@Martin Clifford Styan
'Substratum in Sanskrit' is an outdated story sustained only by the refusal of it's propounders to acknowledge the facts. Witzel's examples of 'Para Munda Substratum' are dealt with here- (while they fail anyway when you consider the chronology of the RV)
https://www.scribd.com/document/64057781/Origin-of-Indo-Europeans-and-Witzel-Examined

And if Lubotsky had known the simple and basic fact about the chronology of the ten Rig Vedic books he would never have claimed BMAC 'substrate' exists in Sanskrit. Otherwise, his argument was pretty strong. (These words spare the oldest core of the RV)

Finally, if Herr Witzel actually believes the Sindhu is not an IA/IE river, he is losing it, or he's getting desperate.

Rob said...

@ Davidski

"And I don't know what you want me to help Salden with? He seems to be off topic, posting about some Italian paper of all things."

Salden keeps echoing the same Q; "Where's the South Asian DNA (in the Western Steppes)?"

So I asked him - what does he know about South Asian archaeogenetics. esp with regard to the Epipaleolithic - Neolithic precedents. Do i take your silence as suggesting miminal ?
Of course, other experts like Sein and AnthroS are welcome to comment..

Shahanshah of Persia said...

@Acharya Agnimitra Both you and Tom used questionable analogies, but both of you have made good refutations of the other's. I would say it's best to end this conversation here.

Anonymous said...

@EastPole

It isn't that simple. Proto-Indo-Iranian split into Indo-Aryan and Iranian somewhere between 2500 BC and 2000 BC. Andronovo is far too late to be Proto-Indo-Iranian, though it was an I-Ir culture of some sort. I would probably consider it Proto-East-Iranian, and ancestral to the later Iranian languages of the steppe and central asia.

Iranians existed outside Iran. The tribes I mentioned all spoke an Iranian language, but they weren't from the Iranian Plateau. Central Asia and Eastern Europe had Iranian speaking populations that weren't from the Iranian plateau, and didn't seem to have any origin there. Similarly, we have evidence of a Mitanni language and culture that was Indo-Aryan but not from India.

Some features within sections of these languages are form interactions with substrate populations, but that doesn't define the linguistic shift that create broader linguistic families. We think of Indo-Aryan as limited to the plains of South Asia, and Iranian with the Iranian plateau, but that is only because subsequent linguistic shifts affected other areas.

For example, there is evidence that the Iranian plateau had an Indo-Aryan speaking population before the Iranian expansion due to an IA substrate in Old Persian and Avestan. We know that there was a Gandhari prakrit spoken in areas that have now shifted to Iranian languages. We also know that much of Eastern Europe and Central Asia was Iranian speaking before the migration period and arrival of the Slavs, Germans, and Turks.

Matt said...

@Rob, will read if I get a chance. thanks.

@aniasi, though from what I remember this has been tested fairly exhaustively by Davidski and didn't really work very well or particularly in any of those models, as generally the outcome is underrelated to WHG and Levant_N, etc. (again from what I remember). Since the offsets are relatively small between what would be a EHG+CHG+lowAnatolian (Yamnaya) and ANE+CHG(?), and we have no firm model for the "other side" that they would be admixing with in South Asia (except using Iran_N and Onge / East Asian which are probably inexact), I guess it is hard to totally eliminate.

But in a way, wider consideration, what is the *point* of this theory? It seems to support no theorised proto-Indo-European homeland from historical linguistics or archaeology. For folk who reject the idea of a migration of people who are mostly descended from a North Eurasian HG population, bringing external ideas and religion, it offers nothing, apart from those people being marginally less related to Europeans. It's not really an "Anti-Aryan Migration Theory" but a "Slightly different Aryan Migration Theory", if we are to talk in those terms? There really are no present genetic observations it seems particularly required to solve.

So that would leave it as probably a generally less parsimonious idea that has the only effect is that it would serve to force localising IE language and religion into CHG...

Anonymous said...

@aniasi

"I personally go with an early Steppe origin Indo-Iranian arrival in South Central Asia, a first pulse of I-A migration across Iran and the Indus Valley between 2500 BC and 2000 BC, with the Iranian speakers of Central Asia forming a second pulse of migration into the Iranian Plateau and back into Eastern Europe."

Why? Based on what?

Anonymous said...

@Matt

That is their view, not mine. I think it has more to do with how the IVC is placed in relation to the IA migration. I'm pretty much down with the steppe framework.

@Epoch

1) Andronovo is too late and has EEF ancestry absent in IA.
2) The timeframe for the radiation of South Asian R1a is between 2500 BC and 2000 BC
3) This is debatable, but I go with the interpretation that Early Cemetery H (1900 BC) already shows IA influence, which indicates it was present in the region.

I don't believe that the IA has any part of the IVC, but rather that they were a nomadic group in the hinterlands and spaces between cities that rose to power as the IVC collapsed.

Santosh said...

@ Ric Hern, Matt and all others

Thank you very much for the responses.

@ Ric Hern

"This will always be a Never Ending Story."
Seems very true to me and personally, it feels like the situation is going to be like this for at least the coming 5-10 years or so.

@ Matt
"obviously something to be resolved in which Ringe's model has proto-II split off last and Garrett-Chang's tree places divergence of Indo-Iranian at 3000 BCE and Indo-Aryan at 2000 BCE."

Yes, these sorts of things are useful and very helpful and I wish people objecting to the mainstream model and having other perspectives do more amount of dirty work like mentioning these things in comments, etc. (on top of these types of arguments being presented in one of the papers the objectors link to and all that) (After all, people do not have anything against hard linguistics as opposed to philology, etc., do they?) If they have been doing so already and it was I who did not realise out of my bias, then I apologize lol.

@epoch2013

"With a bit a bad luck they'll have a Y-DNA R which they can't resolve any further. Or even worse: with both R2 and R1 markers, like Mesolithic Iboussieres from Mathieson having marker for J and I."

Lol! This is new to me! Do these sorts of things happen too? And having a confusion between R2 and R1 like this would be strange as hell and one of the worst things that can happen. What haplogroup will these sorts of guys be assigned to though? Could you enlighten me if you are knowledgeable?

@Acharya Agnimitra
"'Substratum in Sanskrit' is an outdated story sustained only by the refusal of it's propounders to acknowledge the facts. Witzel's examples of 'Para Munda Substratum' are dealt with here- (while they fail anyway when you consider the chronology of the RV)
https://www.scribd.com/document/64057781/Origin-of-Indo-Europeans-and-Witzel-Examined"

Could you link to any sources similar to the one you linked to above, that deal with the hydronymy of the northwestern rivers from a linguistic point of view? (I'm an amateur not personally very much involved in Indo-European linguistics except as I'm in Dravidian linguistics but I get tempted from time to time.)

And regarding Michael Witzel's writings, I don't know to what extent he is accurate is, because his writings feel quite informal, assertive and strange and not very organized and readable to me, many times. It also happens to be the case that any direct works of linguists (as much the mainstream ones as alternative ones and crack-pot ones) as opposed to philologists like Witzel are not accessible to me to help seeing things for myself, as I'm not very deeply involved with IE linguistics. But anyway, one of the takeaways from Witzel's papers is that some words like Śutudrī have no convincing IE etymologies and even change their form later on to acquire etymological basis likely out of folk-etymology, like Śatadru, apparently meaning 'running with a hundred streams'. I can't evaluate for myself if Witzel is completely accurate here, as again, I don't have access to writings of linguists regarding the issue. I'll keep in trying though now that this silly new topic has tempted me into it enough. I gladly appreciate any help from your side regarding the information about authors who offer pointed linguistic criticism to Michael Witzel's ideas about the northwestern river names. Thank you very much!

Davidski said...

@Rob

Salden is correct in pointing out that there is no South Asian admixture on the Eneolithic/Bronze Age steppe. This obviously has important implications for the direction of gene flow between the steppe and South Asia.

The only way that he is wrong, is if a Caucasus Hunter-Gatherer (CHG) like population was native to the northern part of South Asia, and lived there without incurring any Ancestral South Indian (ASI) admixture well into the Bronze Age.

But this seems highly implausible for several reasons which aren't even worth discussing.

postneo said...

@agnimitra, Santosh and others
While witzel and other indologists have listed substratum influences and loans, He is the first to acknowledge along with all others, that non IE loans in Vedic are anomalously small compared to other daughter languages. As for river names his old article in EJVS is meandering, there is nothing diagnostic in it.

Santosh said...

@ postneo

"While witzel and other indologists have listed substratum influences and loans, He is the first to acknowledge along with all others, that non IE loans in Vedic are anomalously small compared to other daughter languages."

Thank you very much!

Anonymous said...

@Santosh

Yes that happens. Have a look at sample Iboussieres25-1 from Mathieson 2017 marked J?. Someone mailed him and he stated they found markers for I as well as J but more J markers.

Or take a look at the hits Genetiker found in that Popovo 2 EHG sample. He called it J1 simply because it had markers for J and J1. But it also has one marker for R1b.

https://genetiker.wordpress.com/y-snp-calls-for-popovo-2/

My understanding is it goes like this: Haplogroups are named after mutations which are cumulative. E.g. R1b and R1a share all markers for R, and also for the one upstream of R called P. You could say R is a special case of P. If one finds only 1 marker for R1b but several for J and J1 the case for J1 is easily made.

Nick Patterson once stated here that you best not trust outliers too much. Also dating issues are known. See two Czech Bell Beakers samples that turned out to be early Middle-Age Slavs after re-dating.

That is what why, if they really find a R1a in Rakhigarhi, the regulars here will not start to fall apart but basically think it over, try to see for themself, and so on.

Anonymous said...

@Santosh

You also have to realise that when they "find" Y-DNA it is not complete. What is being done is look for enough mutations to nail it.

Rob said...

There are no dating issues if you Carbon date individuals , which should be a routine. With a study like this, we can be sure it’s be carbon dated, so we won’t need Epoch’s kind re-interpretation
Nor are there usually any issues with calls, unless coverage is poor . Sometimes the enthusiasts are much better at this part

Ric Hern said...

CHG is basically native to the Caucasus. For thousands of years they were living right nextdoor to the Steppe People.

So why would a CHG migration towards India and back to Maykop be necessary for Steppe people to mix with them for the first time ?

We already see a Haplogroup J in Karelia long before the Maykop Culture arrived. So what is the chances that CHG like people existed in pockets throughout Eastern Europe already during the Mesolithic or even Late Upper Palaeolithic ?

Rob said...

Jokes aside, there are now possibly 2 J1 in East Europe which are EHG, not CHG.
That means something to me.

Ric Hern said...

@ Rob

No like the Southern Urals and many other neighbouring cultures...

So basically your back pocket.

Ric Hern said...

@ Rob

So what do those 2 J1s EHG mean for you ? For me it means that the CHG was bred out of them through thousands of years of mixing only with EHG females....?

Arza said...

@ Rob
East-West J-I split among HG in Europe with Basal admixture being possibly as young, as TMRCA of J-Y6305?

https://www.yfull.com/tree/J-Y6305/
Note that Finn.

Anonymous said...

Poznik et al 2016 has the J1/J2 split at between 30 and 35 ky ago.

Anonymous said...

Considering the fact that CHG can be modeled as a tad WHG + EHG + Basal could have something to do with the J1 in EHG

Rob said...

@ Arza
Yes I think J, or some clades could have been originally some kind of Paleo-Eurasian UHG. I Think CHG formed considerably after the IJ split (c. 42 kya).

Matt said...

OT: Out of boredom, considering some basic models for modern Iranian population, using the West Eurasian PCA from Davidski's latest models.

Just based on simple extrapolation rather than nMonte: https://imgur.com/a/TrXCK

First image models present day Iranian_Persian average as 75% Chalcolithic and 25% X. You can see that X isn't represented by any population, and not by the Samartian-Scythian cluster who might be early Iranian speakers. X lands in a no-man's land between the Scythians and SE Europe. So Iran_Chalcolithic+X is not a good model unless anything close to X is ever found.

Second models Iranian_Persian average as 25% Samartian-Scythian. The outcome, Y, lands near Iran_Chalcolithic, but with extra ancestry from the Bronze Age Levant (probably 10%).

However, this doesn't consider that the Iron Age Iranian sample from 1000 BC (http://eurogenes.blogspot.co.uk/2017/08/the-iron-age-iranian.html) lands significantly west of the average for modern Iranian_Persian (though this population is variable), but not north or south.

This suggests that Iranian_Persian has experienced geneflow from directly east since the Iron Age (depending on how representative Iran_IA is).

So the third model provides a point Z that lands near the Iron Age Iranian, by modeling Iranian_Persian as 20% Pathan and 80% Z.

The final two images attempt to subtract Scythian-Samartian and Pashtun/Tajik related ancestry to model where populations from Iran may have been before an infusion of Scythian-Samartian early Iranian ancestry during the Iron Age, and then further South-Central Asian ancestry in history, by two ways (first by taking Pashtun and Scythian-Samartian related ancestry from Iranian_Persian, secondly by taking Scythian-Samartian related ancestry from Iran_IA). This lands at a point roughly equally between Iran_Chalcolithic and Jordan_EBA.

That would be compatible with Iran seeing significant gene flow from the west and Levant during the Akkadian empire or during the later Persian empires.

We will see how much this is validated by ancient dna.

Chad Rohlfsen said...

EHG likely has minor CHG. They are kind of significantly further from Ust-Ishim than WHG and the qpAdm/Graph works with this admixture. EHG is about as BEu as KO1.

Santosh said...

@ epoch2013 and Rob

Thank you very much for the discussion about Y-DNA haplogroups in ancient genomes!

Chetan said...

@aniasi To be honest I still consider the possibility of a pre-Andronovo Indo-Aryan migration to be slim. I think more Andronovo genomes, preferably from the southern horizon will yield pre-Indo Aryan like ancestry. If the Indo-Aryans did arrive before Andronovo in India or even Central Asia, then the current model of steppe migrations will require a change.

EastPole said...

“I believe Arya culture is a lot more ancient than Westernized Indology suggests; although how old I do not know. Arya is a culture, civilization, philosophy and faith rather than ethnicity. It was geographically spread from Iran, to the southern former Soviet Republics, Afghanistan, Tibet, Thailand, Vietnam, Loas, Cambodia, Malaysia, Sumatra, Java, Bali, Sri Lanka; although I don’t know what geography it was “founded” from so to speak. Its offshoots are the East Europeans, Germanic tribes, Romans, Greeks. Maybe they are more than offshoots and part of the original civilization.”

http://www.brownpundits.com/2018/02/03/ancient-arya-culture/

What do you think? Were the East Europeans offshoots or part of the original civilization?
Our genes and our languages seem to be more related to the original population, from which everything started, than the genes and languages of others.

CromwellCorner said...

Hi All,

Considering that the Anatolian neolithic populations that brought farming to Europe maintained a semi-distinct identity from the 'aboriginal' hunter gatherer populations of Europe for thousands of years, would it be a safe assumption to make that (1) the presumed Iranian Neolithic populations that spread farming to South Asia also maintained a distinct identity from the aboriginal populations of South Asia for thousands of years? It seems ancient DNA studies are clearly showing that pots were people...

If the above is a safe assumption, and considering that the Reich 2009 study concluded that prior to ~2000 BCE there was no admixture between ANI and ASI (*see below for caveat), can we assume that (2) the population of the Indus Valley would have been'Iran Neolithic' like?

*Although above I said there was no admixture between ANI and ASI before 2000BC in Northern India, it is understood that admixture occurred in Southern India prior to this. Can we assume that this would be caused by an 'Iranian Neolithic' like population moving deeper South in to India, and then ~2000 BCE ASI moves North and admixes into the ANI pops of Northern India?

Finally, wouldn't this 'Iran Neolithic' like population also naturally be a large proportion of the West Eurasian ancestry included in ANI?

If 1 and 2 above can be assumed, can someone use the tools available to determine how populations like Iran Neolithic, Iran Chalcolithic, CHG (I'm assuming these three populations are closely related) etc. are related to ANI? It would be worth investigating to see which populations make up the Indus Valley population before their admixture with ASI ~2000 BCE. It is understood that ANI includes western eurasian ancestry beyond 'Iran Neolithic' like that could have come from many different migrations in to South Asian, but if this is possible to investigate it might give us a better idea of what the Indus Valley Civilization population was like.

Also, I have a question regarding PCAs as I'm not sure how to interpret them. Would anyone be able to clarify why CHG samples cluster closer to SC_Asia pops than samples labeled Iran_Neolithic on many of Davidski's PCAs? Wouldn't Iran Neolithic be closer to SC_Asia as CHG is understood to be geographically more west of Iran Neolithic?

Thanks for any and all help. I think this would be an interesting area to explore further before the paper is released. Maybe Davidski can make a post about it?

Thanks in advance for any and all help. This is a very interesting topic.

Anonymous said...

@Chetan

The current models are actually fairly old. My main issue with it is twofold. First, the archaeological sites represent either a settled population or a nomadic one that is moving within a given area, instead of through it. In my view, Andronovo represents a cousin culture of those I-Ir who stayed behind. It is also far too lately dated, since the chariot appears amongst the Mitanni and Hyksos by the 16th century BC.

Second, it cannot explain the IA substratum in the Iranian Plateau, or the IA features found in cemetery H, or the fact that it doesn't show as much correspondence with the subsequent cultures of Swat, or the copper hoard culture.. It isn't until Federovo that there are fire altars and cremation, and that is far to close to the Painted Grey Ware culture.

Anonymous said...

@aniasi

Thanks. Interesting.

Vara said...

@aniasi

You pretty much nailed it. The pro-steppe people need to solve the Indo-Iranian, Tocharian and Hittite questions instead of bashing the almost dead OIT.

Can you explain the IA substratum in Iranian?

Chetan said...

@aniasi The chariots could have spread both east and west from the focal point origin making their way to Mycaenean Greece and to Anatolia around the same time. I'm not saying that the Indo-Aryans couldn't have brought the chariots to Mittani, but here is another possible explanation.

Anthro Survey said...

@Matt

Cool beans. Your analysis, in a way, supports my choice of using Iranian_Jews as a Monte input for Iranian_Fars, Persian and Mazandarani. As I've said before, there's a gap of ~2.5 millennia between Iran_Chl samples and presumed pulse of steppe-related ancestry into the region. Hence, there's no telling how much Mesopotamian influence the western plateau absorbed. More than likely, it was a considerable and ongoing process since Chalcolithic times.

Rob said...

@ David

"Salden is correct in pointing out that there is no South Asian admixture on the Eneolithic/Bronze Age steppe. This obviously has important implications for the direction of gene flow between the steppe and South Asia.

The only way that he is wrong, is if a Caucasus Hunter-Gatherer (CHG) like population was native to the northern part of South Asia, and lived there without incurring any Ancestral South Indian (ASI) admixture well into the Bronze Age."

The entire breadth of territory from the Near East to central Asia, Iran, Indus is lacking EUP, MUP and LUP data. I suspect that, from the clues we've got so far the area between Iran and the Indus was a well spring for UP Eurasian groups (Kostenki/ proto-WHG, the non-ENA part of ANE), sans Basal and ASI-Onge, which are later arrivals.

Davidski said...

@Rob

A very robust inference from the currently available ancient and modern data is that indigenous Paleolithic/Mesolithic foragers in South Asia were not CHG-like West Eurasians, or any type of West Eurasians at all.

In all likelihood, they were a Southeast Asian-related population with elevated Denisovan admixture, and if so, not relevant to the population prehistory of West Eurasia.

http://eurogenes.blogspot.com/2016/03/higher-than-expected-denisovan.html

Let's see if this inference is debunked by ancient data from South or South Central Asia. I doubt that it will be.

Pr V said...

why not? if you take a look at modern baloch or brahui samples who are from pakistan, many of them are 70-75% chg-like or iranian neolithic with the rest being ASI and steppe mixtures. During ancient times, they could have been closer to 100% chg-like or iranian neolithic. that shows that pure chg-like populations could have existed in northern india/pakistan area.

Rob said...

My guess is Iran Neolithic but less basal and more ANE/ EHG.

Nirjhar007 said...


It isn't until Federovo that there are fire altars and cremation
As per Kuzmina just cremation, no altars apart from domestic hearths...

Anthro Survey said...

@Nirjhar, Rob

Let's also not forget about Central Asia which could have harbored some ANE/EHG-heavy populations, and, quite possibly, exerted limited influence on Greater India and the Iranian plateau prior to 2000BC or so. It's much more parsimonious to suspect Central Asia influencing both India (and the steppe for that matter) than it is any OIT scenario.

The Srubna outlier myself and other are getting in models could mean either:
1)central Asian admixture picked up by IA migrants from "far eastern Europe" carried south or
2)a pre-existing EHG-Iran layer in India on top of the ASI-Iran_N layer.

I should point out that scenario 1)raises the possibility of PARA-IE linguistic influence in India prior to the "main" waves.

As for ASI---
I find it hard to imagine a lack of ASI in northern India prior to the Bronze Age. NorthEast India features a very similar climate as well as fauna/flora to that of South India and northeast isn't separated from northWest by any major mountain ranges, etc. Granted, there is the Thar desert which would have impeded *some* east-west movements, but the Punjab-Haryana corridor would have sufficed. So, imo, by Bronze Age times, there would have been an ASI equilibrium established across India with lesser %'s in "Pakistan". The significant presence of an ASI signal in Pashtuns and no plausible explanation for how it came to be there after 2000BC further diminishes the zero ASI scenario.

"while in Sanskrit and ancient North India we dont find similar traces of a lost language ."
Doesn't Sanskrit have a Dravidian substrate, though?

Davidski said...

@Pr V

During ancient times, they could have been closer to 100% chg-like or iranian neolithic. that shows that pure chg-like populations could have existed in northern india/pakistan area.

Yes, but not until the Neolithic, courtesy of this phenomenon...

http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0095714

Don't expect anything closely resembling West Eurasians, modern or ancient, in Mesolithic South Asia, even in what are now Pakistan and Afghanistan.

Expect a Southeast Asian-related population with some sort of unique twist, like, perhaps, admixture more or less related to the non-basal part of Iran_Neolithic.

@Rob

My guess is Iran Neolithic but less basal and more ANE/ EHG.

This seems like a low probability bet.

Anthro Survey said...

@Davidski

You know how, for ex, Mal'ta Boy scores on divergent components like South Asian, Amerindian, Baltic, North Atlantic, West Asian in your K13 unsup. calculator(built using modern populations)?

Well, in this context it clearly means that he was in the same clade, same ancestral stream that contributed to all those modern populations scoring high on those components.

So, applying this rationale, have you ever ever subjected steppe genomes to k13? If so, which ones score some South Asian and which ones score nil? In particular, how do LATER(MLBA, LBA) steppe genomes perform?

Rob said...

@ Dave

Well it's hard to have a strong opinion either way. Even the archaeology lags behind. But maybe youre right
BTW That article points out that farming spread via central & northern Iran, avoiding the south (which was too dry), into Afghanistan and central Asia. Seems like a northern route.
Its hard to guess what they encountered in Indus and Baluch region. My feeling is that ASI were much further south in the rain forests etc.
Hopefully we shall soon see.

postneo said...

@anthro"Doesn't Sanskrit have a Dravidian substrate, though? "

ASI is not a language marker. Its not a given that any ASI population in north India spoke dravidian.

As per the mainstream such as Witzel the dravidian substrate in vedic is negligible and even less significant than what he calls a small para-munda substrate.

based on modern IA features, I feel there was a para dravidian substrate in north India. it had many features alien to dravidian, word initial retroflexion, aspiration(voiced and unvoiced). voiced aspirates are an areal feature and there are plenty of non IE voiced aspirate examples. At the same time there are some features shared with dravidian too. e.g. ..verb negation in Bengali, lacking in other IA. verb conjugation based on gender In Hindi and westward is similar to dravidian but lacking in Bengali and eastern IA.

As for IVC, It would have been a sink for diverse language families.

Anthro Survey said...

@postneo

Oh, I wasn't suggesting the ASI-Dravidian connection. In fact, I'm more inclined to see the arrival of Dravidian accompanied by Iran_N ancestry. Then again, though, the issue of ANE-like non-basal ancestry of Iran_N potentially being present in India and/or later ANE from Central Asia making inroads complicates matters.

Davidski said...

@Anthro Survey

So, applying this rationale, have you ever ever subjected steppe genomes to k13? If so, which ones score some South Asian and which ones score nil? In particular, how do LATER(MLBA, LBA) steppe genomes perform?

Closest I got is this...

http://eurogenes.blogspot.com/2015/03/first-look-at-bell-beaker-corded-ware.html

Then again, though, the issue of ANE-like non-basal ancestry of Iran_N potentially being present in India and/or later ANE from Central Asia making inroads complicates matters.

The non-Near Eastern (mostly Basal) ancestry in Iran_Neolithic is probably best described as AG3-like. But there's also some other stuff in there that I haven't been able to characterize precisely. However, it appears to be more East Eurasian-like than EHG-like.

In fact, ANE/EHG-like is a very poor way to describe the non-Basal part of Iran_Neolithic, so I'm not sure why people keep expecting, or probably rather hoping, to find an EHG-like population in South Central Asia. This is a low probability assumption, so low in fact, that I'd call it a fantasy.

postneo said...

@anthro
"Oh, I wasn't suggesting the ASI-Dravidian connection. In fact, I'm more inclined to see the arrival of Dravidian accompanied by Iran_N ancestry"

again Iran-Neo Itself is not a language marker, such a population cannot be associated with a single language family. Given its age and spread they they could have spoken IE, dravidian and other extinct languages at the same time.

Iran Neo and even steppe like components in South Asia did not happen is a single shot but were diffusing alll the time. If the paper comes, out it will provide some idea.

Chetan said...
This comment has been removed by the author.
Chetan said...

@postneo _"based on modern IA features, I feel there was a para dravidian substrate in north India. it had many features alien to dravidian, word initial retroflexion, aspiration(voiced and unvoiced). voiced aspirates are an areal feature and there are plenty of non IE voiced aspirate examples. At the same time there are some features shared with dravidian too. e.g. ..verb negation in Bengali, lacking in other IA. verb conjugation based on gender In Hindi and westward is similar to dravidian but lacking in Bengali and eastern IA."


Isn't Munda part of the Austrasiatic family though? That means they would have entered the subcontinent from the east bringing rice cultivation with them. That makes it far too late for them to influence Indo-Aryan, especially the earliest layers of it attested from the north-west.

Also the gender conjugation of Hindi verbs can be shown to be an internal development from Sanskrit by the use of the Sanskrit participle for forming the present tense. Other than that, modern Indo-Aryan languages do show a lot of morphological influence from Dravidian

Matt said...

Davidski: The non-Near Eastern (mostly Basal) ancestry in Iran_Neolithic is probably best described as AG3-like. But there's also some other stuff in there that I haven't been able to characterize precisely. However, it appears to be more East Eurasian-like than EHG-like.

In fact, ANE/EHG-like is a very poor way to describe the non-Basal part of Iran_Neolithic, so I'm not sure why people keep expecting, or probably rather hoping, to find an EHG-like population in South Central Asia. This is a low probability assumption, so low in fact, that I'd call it a fantasy.


Follows from this that the whole idea of expansion of EHG like populations into the Middle East to form CHG/Iran_N, from Fu's newest paper, must be wrong? (So cannot be linked to presence on y-haplogroup J / R2?). Also CHG non-Basal Ancestry must be largely like Iran_N's... (whatever that is like).

postneo said...

@chetan
"Also the gender conjugation of Hindi verbs can be shown to be an internal development from Sanskrit by the use of the Sanskrit participle for forming the present tense"

Can you provide an example derivation of a participle leading to a conjugated verb. What I have fund so far is not convincing. What tendency drove the original verb to get dropped in the first place? why did it not happen in eastern NIA? When did the split occur? We are not just talking Hindi but all central and western NIA( the only exclusions might be dardic.. need to check)

@chetan
"Isn't Munda part of the Austrasiatic family though. That means they would have entered the subcontinent from the east bringing rice cultivation with them?. That makes it far too late for them to influence Indo-Aryan, especially the earliest layers of it attested from the north-west"

OK but how is that relevant? I am just paraphrasing Witzel on his para-munda substrate. As for potential Austroasiatic influence. How do you know its time frame? You just provided a direction not the time frame.

If you want to associate it with Rice, then Rice cultivation in IVC is already pushed back to 2200 bc at least so its well within the conventional vedic/IE time frame.

Davidski said...

@Matt

Follows from this that the whole idea of expansion of EHG like populations into the Middle East to form CHG/Iran_N, from Fu's newest paper, must be wrong? (So cannot be linked to presence on y-haplogroup J / R2?). Also CHG non-Basal Ancestry must be largely like Iran_N's... (whatever that is like).

I don't think this idea is wrong, but it probably only tells a part of the story, for one, because of the lack of Paleolithic genomes from Central Asia.

The non-Basal part of both CHG and Iran_N can't be explained just by the expansions of European foragers from, say, the Balkans and the steppes into the Near East, because they're not simply Basal/EHG-like/WHG-like. They both have something more eastern about them that will probably be traced back to Central Asia, and perhaps ultimately, at least in part, even further east.

And I don't think that CHG and Iran_N simply differ in terms of their Basal ratios. It seems to me that Iran_N has more of the said eastern input than CHG.

Matt said...

Yeah, but overwhelmingly CHG+Iran_N non-basal ancestry has to form a clade? (Compared to CHG non-basal mostly forming a clade with EHG rather than Iran_N non-basal). Whatever Iran_N non-basal mostly is, so is CHG non-basal.

Davidski said...

@Matt

Yeah, but overwhelmingly CHG+Iran_N non-basal ancestry has to form a clade?

Clades are a relative concept, and I'm seeing enough differentiation between CHG and Iran_N for them to become sister clades when more Paleolithic genomes roll in from around the Caspian and Central Asia.

Chad Rohlfsen said...

The best I could come up with way back when, is that the HG ancestry in Iran is basically mostly WHG-like, with a good chunk (~30-35%) of something between MA1 and Karitiana, as far as ANE/ENA cline. They routinely asked for about 15% from around the Ami line, near the line into Native Americans.

CHG is higher in WHG in that ratio, and of course, less Basal Eurasian.

Chetan said...

"Can you provide an example derivation of a participle leading to a conjugated verb. What I have fund so far is not convincing. What tendency drove the original verb to get dropped in the first place? "

I remember reading it from the Cambridge Survey of Indo-Aryan languages.

"OK but how is that relevant? I am just paraphrasing Witzel on his para-munda substrate. As for potential Austroasiatic influence. How do you know its time frame? You just provided a direction not the time frame."

We are not exactly sure about the date of the Munda expansion into India. But if it post-dated Indo-Aryan migrations, it's difficult to see how they could have left a substrate in early Indo-Aryan languages.


Jaydeep said...


Yes, but not until the Neolithic, courtesy of this phenomenon...

http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0095714

Don't expect anything closely resembling West Eurasians, modern or ancient, in Mesolithic South Asia, even in what are now Pakistan and Afghanistan.

Expect a Southeast Asian-related population with some sort of unique twist, like, perhaps, admixture more or less related to the non-basal part of Iran_Neolithic.


The above paper is based on assumptions and is extremely average. Here is what they simply fail to factor :-

1. South Asian Neolithic is not solely known through Mehrgarh but also through Neolithic sites in Haryana such as Bhiranna and Kunal. Bhiranna's dates are as old as 7300 BC i.e. it is a site as old as Mehrgarh.

Both Mehrgarh & Bhiranna show the use of domesticated Zebu cattle right from the start. Domestic cattle is not known to the Iranian Neolithic people. They were mostly goat herders while the people of Mehrgarh & Bhiranna were already using cattle, goats and sheep - a more complex Neolithic economy.

Here is what the excavator of Mehrgarh, J F Jarrige himself has to say on this matter -

"The similarities noticed between Neolithic sites from the eastern border of Mesopotamia to the western margins of the Indus valley are highly significant. A sort of cultural continuum between sites sharing a rather similar geographical context marked with an also rather similar pattern of evolution and transformation becomes more and more evident. But the Neolithic of Mehrgarh displays enough original features to imply an earlier local background which has so far not been documented. Nevertheless the cultural dynamism shown by the inhabitants of Mehrgarh as early as level I of Period I indicate that the Neolithic of Balochistan cannot be interpreted as the ""backwater'' of the Neolithic culture of the Near East."

Considering the fact that indigenous South Asian cattle, the Zebu, is already present both at Mehrgarh and at Bhiranna (sites several hundred miles apart) already in a pre-7000 BC context, it should be clear that the origins of South Asian Neolithic are older and they cannot be in a simplistic and shoddy manner be attributed to have originated from the Iran Neolithic.

-----------

As an aside, isn't it quite curious that the Iran Neolithic is so different to its neighbouring Levant Neolithic and Anatolian Neolithic, both in terms of genetics as well as subsistence, while far away in Mehrgarh, Baluchistan there is such close resemblance in subsistence strategies (in terms of housing etc.) and significantly enough, the modern inhabitants of Baluchistan, the Baluch & Brahui, show the highest amount of Iran_N component.

Could it be that South Asian Neolithic is not derived from Iran_Neolithic but that it is the other way around ? This looks a very distinct possibility also because the region where Iran_N sites have been found (in the Zagros region) was extremely inhospitable to life during the LGM and couldn't possibly have been a refugium.

https://www.nature.com/articles/srep31326/figures/2

--------------------

So to argue, based on that useless article that South Asian Neolithic came from Iran and therefore the West Eurasian ancestry in South Asia is only that old, displays a limited understanding of the subject.

(continued...)
--------------------------

Jaydeep said...

Here is something from genetics to show why West Eurasian ancestry in South Asia is atleast as old as the Mesolithic, if not much older :-

"We found no regional diversity differences associated with k5 (ANI/Iran_N/CHG) at K = 8. Thus, regardless of where this component was from (the Caucasus, Near East, Indus Valley, or Central Asia), its spread to other regions must have occurred well before our detection limits at 12,500 years."

It also has this to say regarding the putative origins of the ASI.

"In contrast to widespread light green ancestry, the dark green ancestry component, k6 is primarily restricted to the Indian subcontinent with modest presence in Central Asia and Iran. Haplotype diversity associated with dark green ancestry is greatest in the south of the Indian subcontinent, indicating that the alleles underlying it most likely arose there and spread northwards."

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3234374/

"Most likely events responsible for the current distribution of J2-M172 sublineages into Indian subcontinent could be any combination of 1) entry of herders from West and Central Asia/Middle East during late glacial maximum (LGM) of Holocene, 2) Neolithic demic diffusion from the West, and 3) Bronze and Iron age migration/admixtures."

http://eurogenes.blogspot.in/2016/01/y-hg-j2-has-deep-and-complex-history-in.html

"We analyzed 368 haplogroup Q3 samples from a diverse set of genealogical projects (see Methods for details) and assigned each to one of the Q3 branches...Figure 4 links the deepest Upper Paleolithic split (Q3e vs
the rest of Q3) to West and South Asia
: most Q3e samples (green) were found in Pakistan and India, although some were also found in Europe and North Asia (in Kadom Tatars to the west of the Volga river and in a single Evenk sample from South Siberia); these latter samples fall into the narrower Q3e-YP4500 sub-branch and might represent a later spread of Q3e, rather than its place of origin."

https://bmcevolbiol.biomedcentral.com/articles/10.1186/s12862-016-0870-2

"The hg U7 Bayesian skyline analysis (Fig. 3) shows a clear signal for an overall demographic expansion after the LGM. U7a drives the early stages of this demographic expansion, whereas the signal for U7b (the predominantly
European subclade of hg U7) occurs much later, ~8–5 kya (Table 1 and Fig. 3). The subclades of U7a that are common in the Near East and South Asia (U7a1, U7a2, U7a3, and U7a10) are characterized by coalescence dates and a growth phase prior to the Holocene (Supplementary Figure S1 and Supplementary Table S4). Among those, U7a3 is both the oldest (~19 kya) and most frequent throughout these two areas, whilst U7a1, U7a2 and U7a10 are older than 12 kya . Clades U7a2, U7a3 and U7a10 have individual components, specific to the Near East and South Asia, suggesting that U7a was already differentiated in both regions by the end of the Pleistocene."

https://www.nature.com/articles/srep46044

The combined evidence of all of the above clearly indicates that the West Eurasian ancestry in South Asians dates to atleast the late LGM around 15000 kya if not earlier.

So, the argument that it came with the Neolithic from Iran is hollow and without basis.

Rob said...

@ Matt

"Whatever Iran_N non-basal mostly is, so is CHG non-basal."

@ Chad

" the HG ancestry in Iran is basically mostly WHG-like"

Isnt that opposite to what was suggested in Fu & Lazaridis : CHG had VB like ancestry whilst Iran Neol was more toward ANE ?

Anonymous said...

@Vara

There are S sounds in some Old Persian, Median and Avestan words that should be H sounds in Iranian. They seem indicative of an Indo-Aryan substratum to some of the Iranian languages outside central asia.

https://openaccess.leidenuniv.nl/bitstream/handle/1887/2691/299_054.pdf

It also makes sense since Mitanni is IA, and the same expansion would have left other IA speakers on the Iranian Plateau before they were assimilated by the Iranic migration.

@Chetan

Fair point. However, Andonovo still comes too close to the 1st attestation of IA in the Middle East, and likely overlaps with the IA arrival amongst the Mitanni.

Chetan said...

@aniasi Are you saying the Indo-Aryans represented an initial wave from Sintashta-Petrovka into Central Asia around 2000 BC, but was then incorporated into the Andronovo horizon as it spread after 1800 BC. It's a good theory. I can definitely see it happening but we will have to wait for the genetic data. Why is it still not out I wonder.

Chetan said...

@aniasi That would explain the conservative nature of IA compared to Iranian wrt phonology. Most consonant cluster and vowels of Indo-Iranian are preserved better in IA than in the earliest Avestan text. Granted there is still a considerable time gap between the two.

And the chronology of borrowings into Uralic is late PIE > IA (or undivided IIr) > Iranian. Pretty straightforward. But I can't imagine a spread from the Catacomb culture or Poltavka. Poltavka is all R1b and Catacomb is too far west. I believe strongly the IA origins lie in Sintashta-Petrovka culture.

Anonymous said...

@Chetan

"Are you saying the Indo-Aryans represented an initial wave from Sintashta-Petrovka into Central Asia around 2000 BC, but was then incorporated into the Andronovo horizon as it spread after 1800 BC. It's a good theory. I can definitely see it happening but we will have to wait for the genetic data. Why is it still not out I wonder."

Close but not quite. I would say that the Proto-Indo-Iranian wave expanded through Central Asia before 2000 bc. Sintashta-Petrovka represents those that stayed behind. I'm inclined to see Andronovo as a nascent Proto-Iranian culture, which then expands through the Iranian plateau assimilating the previous IA-speakers, who are best represented by the Mitanni. I'm inclined to see the split into Indo-Aryan and Iranian as two linguistic waves, with the second Iranian wave representing new innovations in language and religion.

I think Sintashta is a good proxy, but all of these cultures represent Nomads who settle to some extent, or at least have an area that they continuously migrate within. These are the ones who stay behind long enough to build a settlement, or make a nice graveyard.

Chad Rohlfsen said...

@Jaydeep, they're Zarzian and not South Asian.

Rob,
Both are basicall between WHG and ANE on Dstats. Very minimal shift toward WHG for CHG. Very similar. Thing is, it doesn't look like ANE, but more ENA. I'll post something tonight.

Chetan said...

@aniasi This is how we can imagine it. Starting 2500 BC, the corded ware people start migrating eastward and southward, causing language shifts in the new areas of expansion. The Poltavka R1b speak a dialect directly derived from Yamnaya but is then replaced by the fresh wave of R1a migrants from the north. This leads to the development of the Catacomb culture in the west and the Potapovka- Sintashta culture in the east. Perhaps a spill over of Potapovka/ Sintashta R1a migrants speaking early Indo-Iranian/ Indo-Aryan expand into central Asia around 2000 BC and occupy small pockets there. Some even take over the fortifies cities of the BMAC agriculturalists. Soon the left over Petrovka people migrate south and east giving rise to the Andronovo horizon. This new wave speaks a distinctly Iranian language and is in competition with earlier pockets of Indo-Aryan left in Central Asia. Being not too separated in time, they still see each other as kin, but the Iranians religious rites clash with the Indo Aryans. Under pressure from the proto Iranian expansion, pockets of Indo-Aryans are pushed into South Asia (where they occupy the Swat valley first) and westward to Mesopotamia. But the Mittani Aryans could also have been hired mercenaries from Central Asia like how early Arab kingdoms hired Turkic mercenaries.

Anonymous said...

@Chetan

Fairly similiar to what I see, though I would say that the Indo-Aryans would be present in the Southern Caspian through to the Mitanni region. As I said, I tend to favour an earlier chronology since a) there are no signs of an invasion in the IVC cities and towns with b) IA and GGC influences in Cemetery H, which means that the IAs were already present and poised for a takeover.

I also don't see the Iranian religion changing until the rise of Zoroastrianism, when it probably still resembled the Rg Veda.

Chad Rohlfsen said...

Rob,

Here you go. This first one has a bad outlier Iran Ami Ami Ust Z>4.

https://drive.google.com/file/d/19MV5id9PU8-rTzcvZNBxjk_HciNybLW2/view?usp=sharing

This second one, is much better with no Z > 3. I could drop it down around just over 2, but this is sufficient to show what I am talking about.

The two branches that are at 0 are of no concern, since they follow shared drift branches.

https://drive.google.com/file/d/1g3wniCrLotScs4nqtqWesur6fkA2N66U/view?usp=sharing

Iran comes out about
41% B_Eu
34% UHG
12% ANE
13% ENA

Chad Rohlfsen said...

I'm going to add just one pop at a time to test this again if I can later.

Chetan said...

@aniasi Check out Adunqiaolu: new evidence for the Andronovo in Xinjiang, China in Antiquity if you haven't already. Certainly worth a read. And according to it, the expansion of the Andronovo horizon to the south (northern border of Central Asia) only took place after 1600 BC. So either the Indo-Aryans migration was part of an earlier expansion or their date of entry into India is later than 1500 BC. Since the second possibility is unlikely, I am now inclined to take the first view.

Alogo said...

Matt, that seems to be roughly the case. Compared to the one Iron Age sample, modern Iranians seem to need more IA steppe/central asia and secondarily Levant.

udaya udaya ranasinghe seneviratne said...

Are there any indications that Ethnic Lithuanians are very much closer to Indus population than North Indian Brahmins?

John Smith said...

From the discovery that has been achieved by the other scientists from different countries in Dr Rai's team, it has been found that the skeletons discovered from Rakhigarhi are the ancestors of those races who are the speakers of languages of the Indo-European language family and who keep making the claim of being the greatest race in the world.

Wow, does it really say that? Sounds completely unprofessional for a supposed credible news publication.

«Oldest ‹Older   201 – 317 of 317   Newer› Newest»