search this blog

Monday, May 23, 2016

The Rakhigarhi 15


Samples from fifteen Indus Valley Civilization (IVC) skeletons recently unearthed at Rakhigarhi have been sent to three different ancient DNA labs for analysis. The Tribune India reports that one of these labs is located at Howard University, USA (see here).

Why Howard University and not, say, Harvard University? Or did the reporter really mean the Howard Hughes Medical Institute?

It might be an important detail, because Harvard's David Reich is affiliated with the Howard Hughes Medical Institute. So if that's actually where the IVC samples are going, we're likely to see good quality genome-wide genotype data from them at some point, probably in the not too distant future.

In any case, I know that a lot of people are eagerly awaiting ancient DNA from the IVC, so a paper based on just basic uniparental markers from up to fifteen IVC individuals is really going to cause a stir. For instance, check out this discussion thread at a popular Indian forum: Out-of-India - From Theory to Truth: Part 2.

Holy shit. Let's be realistic here; the probability that ancient DNA from South Asia will make the Out-of-India theory (OIT) a credible proposition is nil. I'd say we have a better chance of finding extraterrestrial life in the next few years than the OIT ever becoming a viable option in the Proto-Indo-European (PIE) homeland debate.

If we eventually see genotype data from IVC samples, my guess is that they'll either resemble present-day Dravidians from South India, Brahuis from Pakistan with no discernible recent admixture from Africa and Arabia, Munda-speaking tribals from near the Himalayas, or a complex mixture of two or all three of these groups.

Unfortunately, knowing this won't tell us what language the IVC people spoke. The only way to learn that is to decipher the Harappan seals.

However, comparing the IVC samples to present day North Indians, particularly high caste North Indians like Brahmins, is likely to reveal whether South Asia experienced a large scale population shift during or shortly after the IVC collapse, especially at the top end of its society. If so, it's also likely to tell us what type of population(s) moved into South Asia at this time and basically laid the foundations for modern South Asia. This information is going to be very valuable in the PIE homeland debate.

So, now the waiting game begins.

Update 24/05/2016: I ran a series of D-stats of the form D(IVC_proxy,Test_population)(Source_population1,Source_population2) to try and profile the non-IVC ancestors of present-day high caste North Indians. All results with Z scores of > negative 2 are in bold print.

If we assume that the IVC people were a lot like present-day Dravidian-speaking South Indians, which I think is a pretty good assumption, then the newcomers came from the Bronze Age Eurasian steppe, and were probably very similar to the Poltavka outlier individual that I blogged about recently. See here.


If, on the other hand, the IVC people were basically like the least admixed Brahuis from present-day Pakistan, then the newcomers were very similar to the Afanasievo pastoralists of the Early Bronze Age (EBA) Altai region.


But, if the IVC people were like the present-day Munda-speakers from near the Himalayas, then the newcomers were probably a mixture of both Kotias-related agriculturalists from West Asia and pastoralists from the Eurasian steppe. Note the lack of any Z scores lower than -2. In other words, the results are very close to neutral.


See also...

Rakhigarhi ancient DNA paper probably a while away

110 comments:

Nirjhar007 said...

We have no idea of the potency of the aDNA that will come out of India, it was such a vast population and civilization also.

So, to be realistic, I think we better watch and see, before making conclusions favoring any theory!

Davidski said...

I'll run some tests tomorrow on South Indians, Mundas and Brahuis of the form D(IVC_proxy,Indian_North)(Source1,Source2) to see what shows up. I'll post the results in an update or in the comments here.

Nirjhar007 said...

Alright, but do you have the genome of Korku and Santali for munda, so you can compare? also Kurux for Dravdian with Southern Dravidian groups?.

Davidski said...

How about Paniya, Pulliyar, Ho, Kharia and Ori?

Nirjhar007 said...

The first two are from the same region , and later three also . We need to see what difference do geography create .

For Example Korku are a bit isolated from the rest of Munda groups, same is Kurux for Dravidians.

Karl_K said...
This comment has been removed by the author.
Nirjhar007 said...

Dave,

Ori? are you sure?. There is a small local group but I now think that, Its impossible that they got tested.

Karl,

Of course they have millions of living descendants , there is no doubt in that.

batman said...

The distribution of y-dna H1, H2 and H3 in modern populations are quite interesting as an overtoure to this repport.

Especially since the H1-outlier carried by the Romani famnilies obviously carried both y-dna and root-language from India to the Levant and Europe. (Probably before and during the Persian wars, simulatniously with the deportation of Hebrews to the Caspian plains, where the old spice- and silk-routes to Europe would pass.)

Moreover, H2 (P96) was found in Neolithic Iberia - somewhat proving the expected - that y-dna H have a paleolithic origin, as a descendant of the pale-arctic y-dna hg C/F.

While the major bulk of toadys hg H (as H1 and H3) are found in India, there seem to be some (relict?) lines of the H2-mutation left in Europe.

H2 (P96) is found at locations in France, Switzerland, Germany and the Netherlands, as well as Armenia, Iran, and India.

According to wiki; "It is hypothesized that Haplogroup H2 may have been one of the original patrilineal lineages of the earliest Europeans; and may indicate that ancient Europeans once shared a common ancestor with the Proto-Dravidian peoples of the Southern Indian Subcontinent."

The later influx of hg J in India may be a marker of a "brother-dynasty" that took over - became the new kingline and thus the subsequent nobility and male community - after the IV-collapse.

Seemingly there's another emigration out of 'indo-aryan' India, too - represented by a an outlier of y-dna J. Thus we may associate the spread of J from the Persian bay to the Levant with the Akkadian expension. Later J1 seem to have spread with the expansion of Achamenid Persia - which included Bactria, Trans-Caucasia, Anatolia, the Levant, Egypt and islands of the eastern Meds.

May that point to India and the IV-civilisation as the origin of the early semittic dynasties (y-dna) that appeared in southern Sumeria during LBA, from where it is known to have expanded into the Middle East and NE Africa (Wells 2003, etc.)?

May that also explain the reson for the combination of common words and vocals that seems to link the Asiatic and the African languages?

Davidski said...

J2 is much older in India than the IVC collapse.

The rapid expansion of R1a-Z93 is dated to just before the IVC collapse.

Nathan Paul said...

So All of you are betting that R* wont be found there. Lets see.

Davidski said...

The only way you'll see R* in those remains is if R2 isn't properly tested and it ends up looking like basal R.

This has happened in the past with modern DNA from the region.

Davidski said...

Check out the update!!!

Nirjhar007 said...

There was no collapse , but a compromise of social structure due to natural causes. Instead of any hypothetical migration from steppe , chances are likely that, there were movements from N India itself!. But it is related to Indo-Iranians , not IE's! .

Anyway, none of the above scenarios of David, are to be taken seriously , there is no data to suggest what he postulates. All we are getting is inferences of common ancestry, the gigantic holes due to lack of ancient genome, when gets filled, things will be radically different, that I promise you guys .

postneo said...

howard must be harvard

Davidski said...

Nirjhar,

Nothing will change.

There was a massive population movement to India from the steppe, but no population movement from India to the steppe.

postneo said...

this is an important population

http://www.nature.com/jhg/journal/v59/n1/full/jhg2013112a.html

postneo said...

Veddas are a relatively isolated HG population differentiated from the mainstream

Davidski said...

From your link...

Interestingly, West Eurasian contributions of 28.19, 25.33, 25 and 20% were detected in the Sri Lankan Tamils, Vedda people, Up-country Sinhalese and Low-country Sinhalese respectively, whereas only a 1.75% contribution was evident in the Indian Tamils.

There isn't much of a difference between Sri Lankans and South Indians, and in fact Paniyas from South India appear to have less West Eurasian admixture than any Sri Lankans that I've tested. They probably even have a bit less than Veddas.

I can't see Veddas being a better proxy for IVC than most Dravidian-speaking Indians and even some low caste Indo-Aryan Indians.

George Okromchedlishvili said...

Veddas might be good proxies for pre IVC HG Indians but I strongly doubt that they were the ones that lived in Northern India at the time of Aryan Invasion.

Davidski said...

Best bet for IVC people is something like a mixture between some Brahui and Dravidians from South India.

So it's likely that the invading population from the north wasn't exactly like Poltavka outlier, but maybe like a mixture of Poltavka outlier, Afanasievo and even Kotias.

Nirjhar007 said...

Hogwash, funny never the less...

Davidski said...

You need to be more objective and move with the data, no matter what your own personal preferences are on the topic. That's the scholarly way.

huijbregts said...

@ Davidsky

Only today I noticed that in your Yamnaya post you used a special Dstat-sheet. This is a nice sheet.
When I cluster it with R hierarchical clustering, method 'ward.D2' and k=7, I see some nice clusters.

[Bichon, ElMiron, Hungary_HG, Loschbour, Motala_HG, Villabruna]
average Dstat on column LaBrana1 is 0.4830500.

[Atayal, Dai, Itelmen, Munda, Nganasan, Papuan, Ulchi, Yakut]
average column scores Dai2= 0.4234875, Han2=0.4250500, Karitiana2=0.4105250

At k=7 I also find a cluster of 28 mostly Indian and Turkic pops. The ancients are AfontovaGora3,
Altai_IA, Karasuk_outlier, MA1 and Okunevo.
The best fitting column averages are Samara_HG=0.3894143 and Yamnaya_Samara2=0.3840714.
These are better than Satsurblia=0.3696321 and South_Indian2=0.3840714.
Selection of only the Indian populations did not provide a better column.
It seems we do not yet have the right reference column for the Indian pops.
Maybe you surprise us with your upcoming post.

Nirjhar007 said...

D,
I am exactly doing that bud...

postneo said...

Veddas are not un-admixed but they have MTDNA R and U groups in good contrast to tamils and sinhalese who are dominated by M.

The so called "west eurasian" lineages were probably in the Vedda before IVC. Need to check this but W, and U in general have deep history in the subcontinent. its unlikely that european women turned Vedda in recent times.

Grey said...

1) "If we assume that the IVC people were a lot like present-day Dravidian-speaking South Indians, which I think is a pretty good assumption, then the newcomers came from the Bronze Age Eurasian steppe, and were probably very similar to the Poltavka outlier individual"

2) "If, on the other hand, the IVC people were basically like the least admixed Brahuis from present-day Pakistan, then the newcomers were very similar to the Afanasievo pastoralists of the Early Bronze Age (EBA) Altai region."

3) "if the IVC people were like the present-day Munda-speakers from near the Himalayas, then the newcomers were probably a mixture of both Kotias-related agriculturalists from West Asia and pastoralists from the Eurasian steppe."

Could it be multiple events: near-steppe Kotias types first and steppe later?

Shaikorth said...

Vedda seem to be considerably more West Eurasian than Paniya, and most likely this is recent. They're basically like Chenchu and Hakkipikki.

http://www.cell.com/cms/attachment/2024885451/2044556513/gr1_lrg.jpg

Nathan Paul said...

Postneo,

'so called' is the right word.

R2 is Indian. Kurds and Chechenians have more than 20%
H is Indian Paleolithic. Now H2 in Europe also. So is Mt M.

This is all for somebody's ethnic ego satisfaction.

batman said...

@ Davidski:

"J2 is much older in India than the IVC collapse."

Sure. It's still somewhat enigmatic to explain both H and J as mesolithic, in India. Though, due to it's southern distribution it seems that the H-subclade of the paelarctic F, is somewhat older in India than it's 'brother-line' J.

What I adressed was the possible time of an exodus of hg J, from India to Mesopotamia - and an emergence of J1 there.

There was an extensive maritime trade network operating between the Harappan and Mesopotamian civilisations as early as the middle Harappan phase, with much commerce being handled by "middlemen" from the Perisan Bay, connecting the IVC and 'the Punic' to the Punic Sea, Egypt and Greece.

The only known larger expansion of new settlers in the southern area of the Persian Gulf seem to arise with the Akkadians, some 4.200-3.900 BP. Whether this coincide with the first migration of the Romani (y-dna H) too, or only with the westward migration of Brahmins with hg J, remain to be seen.

Thus the IVC-collapse, estimated at 4.000-3.800 BP could surely have been a part of the reason why the bronze-age migration 'Out-of-India' developed. Moreover, the agressive expansion described in the Sumerian tablets coincide both in time and place with the first known weapons construed for manslaughther ('weapons of war').

What seems clear is that an outliers of hg J came out of India and into Sumeria - from where they spread further west. The first exodus may have happened during the Akkadian period, establishing the new, Akkadian 'empire' in the old Sumeria. (Later this military force were able to expand further west, initiating the Trojan wars.) May this new akkadian dynasty represent the origin of subclade J1?

Thus I remarked that the western outlier of the 'Indo-Aryan' hg J - and maybe the outlier of hg H respectively - was a consequence of the Indus Valley Collapse (4.000-3.600 BP). Moreover I coin that to the Akkadian/Asian appearance in the Levant, Egypt and NE Africa at the same time - by 3.200 BP leading to the "Burnt Cities" of the Middle East and the "Bronze Age Collapse" of the eastern Med - some 3.200 years ago.

A further expanssion - west and north-west respectively - is known to have happened during the Achamenid ('Persian') empire, which conquered the vast areas north of the mountains - surrounding the Black Sea, the Caucasus, the Caspian Sea and Bactria - to controll (all) the continental trade-routes from Asia to Europe - some 2.600 BP.

batman said...

@ Davidski:
"The rapid expansion of R1a-Z93 is dated to just before the IVC collapse."

That may fit very well into the picture - as a third wave of immigrants from Indias northern neighbourhood. This time to settle the cattle-farming and create a north-indian population with a significant degree of lactose-tolerance...

Though it remains to clearify that this migration of cattle-herders - of the later, European stock - were not the first migrants with an I-E language. The first imigrants from mesolithic Eurasia/Caucasia to the Indian sub-continent - carrying y-dna H and J - were probably part of the I-E language-family, too. The impact of the first vawe - on the paleolithic (pre-dravidian) language of ice-time India - was obviosuly not overwhelming, since an end result can be seen as 'dravidian' - and reconned to be outside the I-E language-family.

Later the dravidian language seem spread westwards, anyhow, together with hg J and H. Which may explain how and why the language of paleolithic India became the stem substrate of the Elamo-Dravidian and the Afro-Asian language-groups.

aniasi said...

I notice that there has been little discussion of the major implication of David's test here. The Poltavka and Afanasevo cultures are much older than Andronovo and Sintashta, the presumed ancestors of the Indo-Aryans.

In addition the recent revision of R1a in India (http://eurogenes.blogspot.com.au/2016/04/signals-of-ancient-population.html) places it's expansion at least a few centuries before the collapse of the IVC.

With that in mind, does it not seem to make the most sense that there was an Indo-Aryan presence in South Asia, before the IVC collapse, and that this was not an influx from Central Asia after 2000 BC? I am not proposing that they were the dominant culture of the IVC, but were still present as pastoralists in the countryside. They only become dominant as the IVC collapses, and migrants from cities and farming settlements fall under their control.

At the very least, it helps deal with the Sarasvati conundrum. The description of the river is far to grand to be located in Afghanistan, but now it looks like genetic dating puts the Indo-Aryans somewhere in the region before the the Hakra dries up. It also seems to explain why the Rig Veda has such a northwestern geography, multidirectional migrations, and doesn't mention an external homeland.

postneo said...

Veddas are an important fossil population period.... like the Kalash. They don't play much of a role as a proxy for IVC, that was not the idea or intent.

They are not only geographically isolated but even culturally isolated probably more so than the Kalash. You cannot pretend to be an HG on a whim, its very hard to fake it over the centuries and raise surviving kids. Similarly Chenchus are an ancient isolated population very hard to fake and become one although in their case there are more hooks.. buffaloes etc..

Vedda's isolation is borne out by unshared drift as stated in the paper. interestingly their dominant R mtdna is R30b which is frequent in panjab and r8a frequent in orissa and andhra predesh, far to their north. Other interesting vedda lines are u1a and u7 found scattered in the europe and the west asia. u1a in india is not common in india. u7a peaks only in gujarat, caucasus and the Vedda of all people.

The Vedda are thus anomalous and north shifted compared to their neighbours.

andrew said...

If we assume that the IVC people were a lot like present-day Dravidian-speaking South Indians, which I think is a pretty good assumption,"

FWIW, I think that this is a horrible assumption that is extremely unlikely to be true. The IVC and South Indians were almost completely isolated from each other on a population genetic basis (e.g. probably not one individual exchanged mate per generation) until sometime around the middle Rig Vedic era (ca. 3500 years ago).

The available evidence suggests that the ANI-ASI distinction predates the arrival of Indo-Europeans by thousands of years and may date to the early Holocene or even earlier.

Davidski said...

The available evidence suggests that the ANI-ASI distinction predates the arrival of Indo-Europeans by thousands of years and may date to the early Holocene or even earlier.

West Eurasian admixture arrived in India during the early Neolithic at least, maybe earlier. That's why there are basal clades of Y-DNA J2, mtDNA U1a, and so on, deep in India, and even in Sri Lanka.

But the major admixture event between West Eurasians and Indians only took place after the Indo-European migration, and this is what dominates the admixture signal between the so called ANI and ASI in modern DNA.

So it's highly unlikely that the IVC people were ASI. But it's also unlikely that they were like, say, Brahuis without any ASI.

That's because there's evidence that the IVC may have been Munda or part Munda speaking, and Mundas, like for instance Kharis, are essentially mostly ASI but with significant slices of CHG and East Asian admixtures.

And even if the IVC was not even part Munda, then it wasn't far away from the nearest Munda-speaking region, so some admixture with Mundas is plausible.

Davidski said...

Guys, first clues about Harappan DNA.

- Neolithic migration from Iranian plateau

- later admixture via maritime links (with Arabia?)

See the concluding remarks in the image here...

http://shinpaleopathology.blogspot.com.au/2016/05/new-presentation-harappan-burial-sites.html

Screams Brahui to me.

FrankN said...

While also pertinent to the discussion here, this fresh paper (Monday) seems actually worthwhile a separate thread:
Daniel Shriner, Fasil Tekola-Ayele, Adebowale Adeyemo & Charles N. Rotimi: Ancient Human Migration after Out-of-Africa
http://www.nature.com/articles/srep26565

To understand where they are coming from, see the same authors' 2014 paper
http://www.nature.com/articles/srep06055

Highlights:

1) Kalash as 49% North European - 51% Indian admix in first approximation - with more migration edges, the NEurop. admix is shown to a/o incorporate a Levantine-Caucasian component.

2) Indian admix into the West Eurasian (NEur, SEur, Arab, Levant-Cauc) root; that admixed root has contributed some 55% of Cushitic ancestry, with the remainder being related to Nilo-Saharan and Omotic (Ari~Mota). Interpreted as among others representing "Back to Africa" 60 kya (plus a lot of other migrations later on).

3) 15% Native American ancestry in North Europeans - distinct from Steppe/Yaqmnaya, and also not Siberian (which has provided additional admix into some North Europeans)

4) Connection between Sub-Saharan Africa and a node ancestral to Chinese, Japanese, Melanesian, Native American, Siberian, and Southeast Asian, possibly reflecting an OOA migration some 70 kya.

Davidski said...

That paper isn't worth commenting on.

imam-din said...

So is it official now that Baloch/Brahui component in Harappaworld calculator (Gedrosia in dodecad) tracks the deep indus valley ancestry in modern south asian populations. Can we also conclude that Brahui are relatively less admixed remnants of ancient indus valley civilization people? This would have implications for south indian dravidian population as it would mean that brahui like dravidians invaded south india to spread their language there. Punjabi and sindhi languages always sound to me as having a brahui like dravidian substrate in their phonetics particularly high incidence of retroflex r sounds in these languages.

Davidski said...

It's not official yet, because we haven't seen any DNA results.

To me it seems that they're giving a hint in that poster presentation about what we should expect from the DNA results that they have for their four samples.

And you shouldn't project what you learned a few years ago from admixture calculators onto ancient DNA results. Just totally ignore the admixture calculator stuff.

Grey said...

Nathan Paul

"So All of you are betting that R* wont be found there. Lets see."

Personally I think there was probably multiple flows with
- a CHG like one first
- possibly a small R1b one in the middle connected to metalwork and not a big tribal thing
- the main R1a one
so my guess is if it's too early for the main one then small chance of R1b if the graves are connected to metal working.

pure guess though

Nirjhar007 said...

David,

Are you a retard? they speak of 1st millennium bc , its more likely related with the break around ~600 BC as recorded via archaeology!!!

Nirjhar007 said...

and the Neolithic migration from Iran will be ANE+CHG type..., that occured around 4500-3800 BC AGAIN recorded by archaeology..

Davidski said...

So what? None of this explains steppe ancestry in South Asia, which is obviously there unless you don't want to see it.

Davidski said...

These clowns aren't going to release these first Harappan DNA results at least until they also have the DNA of the 15 newly excavated skeletons, because they didn't see what they were expecting to find, as in strong continuity with present-day upper caste Indians.

So in other words, there's no R1a-Z93 in these four samples. And this is a problem.

Nirjhar007 said...

There are few thing bud-

1. The so called steppe ancestry is most likely a reflection of the archaic common ancestry , the aDNA from Iran, India and even Bactria , as we know coming, will show that and that is almost certain.

2. With no archaeological support the idea of the second millennium bc invasion is discarded by every serious researcher , as I have told probably million of times . There was also , do not forget actual steppe related migrations with Scythians and others from around 2000 YBP , in the Subcontinent area . That migration , which actually did happen has inflated that ancestry in notable manner in some populations such as Gujarat, which was under Saka influence also around Pamir of SC Asia. A systematic thinking with actual data is very needed.

BTW, I am atleast glad to see that you support SSVC will be ANI dominant :) ...

Davidski said...

The so called steppe ancestry is most likely a reflection of the archaic common ancestry.

This is complete nonsense.

The steppe ancestry I'm talking about only formed on the steppe during the Chalcolithic and Bronze Age, and none of its elements are from South or even Central Asia. This package also includes R1a-Z93, which shows a rapid expansion during the Early Bronze Age.

So how and when did it get to South Asia? And did it use a time machine?

Nirjhar007 said...

The steppe ancestry I'm talking about only formed on the steppe during the Chalcolithic and Bronze Age, and none of its elements are from South or even Central Asia.

That is what we call nonsense , because :

1. we need aDNA from Southern areas to conclude how ''stepp exclusive' that ancestry was , so without aDNA from C and SC Asia etc its retarded to think that way.

2. R1a-Z94 is an Asian marker which is now found from Steppe cultures like Yamnyay,Sintashta etc, its almost absent in Europe now days , any guy with a dim bit of wit will agree with that .

3. To prove that Z-94 is not Asian in origin , the only way is to see the aDNA from where it is frequently found i.e. India,SC Asia etc.


Davidski said...

All of the elements that make up Bronze Age Steppe ancestry are native to earlier Eastern European peoples. You know, like EHG, CHG and EEF. Nothing to do with South Asia.

And R1a-Z94 isn't native to Asia, it just moved there from the steppe, just like R1b-M269 moved to New York, Buenos Aires and so on, and is really common there now.

Surely any guy with a dim bit of wit can understand that, unless of course he's living in fantasy la la land.

Seinundzeit said...

By now, I think it's rather obvious that South Central Asians (Pamiri Tajiks, Kalash, Nuristanis, Kohistanis, Pashtuns, etc) and upper caste northern Indians, not to forget Punjabi/Haryani Jatts (who aren't technically upper caste, but are usually more European-shifted than Brahmins from their own regions), have very substantial amounts of steppe Indo-Iranian ancestry. And R1a-Z94, which peaks in South Central Asia, is obviously from the Bronze Age steppe, and probably originates in Eastern Europe.

The only issue is that South Central Asians obviously can't be 100% BA-steppe, yet seem to have equal/more ANE ancestry than BA-steppe populations. But this is easily explicable, since South Central Asians also have heavy amounts of extra CHG, which is mainly ANE + Basal Eurasian. Also, South Central Asians have substantial indigenous South Asian ancestry, which seems to have some sort of affinity to MA1 (we still don't know what that means, perhaps this IVC aDNA will help with the whole MA1-South Asia puzzle).

Regardless, the amount of ancestry from the Bronze Age steppe is going to vary, depending on the method. With the d-stats nMonte sheet, an excellent model for the Kalash/Pashtuns has them at 40% Andronovo + 40% Dravidian South Indian + 20% Kotias-related. So with d-stats, almost half of Pashtun/Kalash ancestry comes from the Bronze Age steppe. In that same sort of model, Tajik_Ishkashim comes out 60% Andronovo + 35% Dravidian + 5% Kotias-related, while the other Pamiri populations also turn out 60% Andronovo, but with less Dravidian (Ishkashim Pamiri people seem to be as South Asian-admixed as Pashtuns and Kalash, but as BA-steppe admixed as other Pamiri populations. Which makes them somewhat strange, but very interesting, as one wonders what cultural/historical situation exists with regard to these different Pamiri groups). So in this context, the Pamiri people have more than half their genetic ancestry from the BA-steppe.

With qpAdm (and with TreeMix), I'm sure we all recall that Pashtuns/Kalash turned out to be 65% BA-steppe, while Pamiri Tajiks were construed as 75%. In hindsight, I think that is too much. But it did demonstrate that this is an unambiguously real (and very substantial) signal.

At the end of the day, I'm sure Pashtuns/Kalash are anywhere from 30%-50% LN/EBA European, while the Pamiri Tajiks are anywhere from 40%-60% LN/EBA European. Now, what we really need is a proper explication of what the non-LN/EBA European ancestry of South Central Asians is, in general genetic terms. Hopefully, these IVC aDNA samples will shed some light on these questions.

Nirjhar007 said...

David Reich, Ron Pinhasi ,

With all due respect to the author of this blog, I think you know now, what I was talking about him..

Regards...

bharatiya said...

As an Indian this is all very exciting. Hope this will lay to rest OIT and other such pointless speculation

Karl_K said...

"That's because there's evidence that the IVC may have been Munda or part Munda speaking"

I don't think you can call it evidence.

"These clowns aren't going to release these first Harappan DNA results at least until they also have the DNA of the 15 newly excavated skeletons."

Wow. That is total speculation on your part. They are certainly not clowns. If they are waiting on more analysis, it is because it will make a more accurate story. That way they will have ample evidence to respond to unreasonable interpretations or accusations by people with their own agendas.

Karl_K said...

"The so called steppe ancestry is most likely a reflection of the archaic common ancestry , the aDNA from Iran, India and even Bactria , as we know coming, will show that and that is almost certain."

This is not anywhere close to certain. Maybe in the la la land I have heard mentioned recently.

Davidski said...

@Karl

I don't think you can call it evidence.

Only if you think that it's not possible to produce evidence with methods employed by historical linguists.

If so, that's pretty harsh, considering that historical linguists produced the most sensible models for the PIE and Proto-Indo-Iranian homelands, which are now being corroborated by ancient DNA.

You can certainly say that it's not empirical evidence, but it's still evidence. Whether it's correct and useful or not is another matter.

They are certainly not clowns.

How much do you know about these people? Would it surprise you to learn that they don't believe in the Aryan invasion/migration theory?

Karl_K said...

"How much do you know about these people? Would it surprise you to learn that they don't believe in the Aryan invasion/migration theory?"

It doesn't matter what they believe. They are collaborating with people at several well respected institutions in multiple non-India countries. Major news agencies have increased awareness of the testing. This can't and won't be covered up by some imaginary conspiracy in anyones's mind.

If the initial data doesn't fit a previous model, then often scientists do additional testing so that they can come up with a model that makes sense without any doubts.

It doesn't mean that they are fools, it means that their previous models were wrong (even if obviously misguided).

I have been in science for several decades and have met very few people who would hide data that disagrees with their politics or previous publications. Usually they just pretend like their previous completely wrong publications always agreed with the new 'adjusted' model.

Davidski said...

@Karl

I don't think you understand the situation here. This is very political. We're not dealing with the usual types of agendas.

For the last decade or so at least some Indian institutions have been trying to build a consensus that there was no Aryan migration, or indeed any significant migration, into South Asia for something ridiculous like 12,000 years or more.

They've basically succeeded. Even in that Moorjani et al. paper on the ~4,000 YBP ANI/ASI admixture in India the authors ended up with a lobotomized version of the truth, in which it was apparently plausible that ANI and ASI populations had lived in India for many thousands of years, but did not mix until the Bronze Age.

It's a fucking joke. Thank god for ancient DNA. But I'm telling you, don't expect this to run smoothly. If anything doesn't fit the Indian narrative, it'll take a while to see the light of day.

Karl_K said...

"If anything doesn't fit the Indian narrative"

There is not a single 'the Indian narrative'. This is controversial within India for a good reason. Mostly because there are millions of very intelligent and educated Indians who are only interested in the historical truth and couldn't give a shit about fairy tales other than that they provide for exciting festivals and holidays. Otherwise, there are 90% of the people who are oblivious that there is any kind of debate or data to have a debate about, and a very small minority of powerful people who have a political agenda.

Davidski said...

Why don't you go over some papers by Indian authors on Indian population history from the last 15 years, and show us how many different Indian narratives there are?

Thanks in advance.

Karl_K said...

It doesn't matter what the narratives in the published papers are.

"Even in that Moorjani et al. paper on the ~4,000 YBP ANI/ASI admixture in India the authors ended up with a lobotomized version of the truth, in which it was apparently plausible that ANI and ASI populations had lived in India for many thousands of years, but did not mix until the Bronze Age."

And so what? We can all see what the data actually shows. Why do you care what some 'academics' who wrote the papers put in the discussion section? If labs at Cambridge and 'Howard' are involved, then the raw data will be available.

They can say whatever they want. I say, the more data the better. Although I prefer open trickles of data as it emerges, in the end, the more the better.

Davidski said...

Get some popcorn and watch the shit storm that comes along with the ancient DNA from India.

Nirjhar007 said...

As an Indian this is all very exciting. Hope this will lay to rest OIT and other such pointless speculation

By pointless I hope you meant KH , otherwise you are a buffoon.

All,

The only reason that we are seeing a global effort, with more sampling being done regarding the SSVC aDNA , is that, they know how valuable its going to be, just not for the Eurasian population history, but also PIE one.

I assume they have succeeded and they have circulated , the news is very significant and that's why its growing day by day.

Davidski said...

Only one buffoon posting in this thread currently, and it's not bharatiya or me.

Karl_K said...
This comment has been removed by the author.
batman said...

The Indian subcontinent have obviously been furnished with haplogroups which origin can be traced to the palearctic Eurasia.

According to the present taxionomy that would include (any) offsprings of y-dna makrogroup C-F, such as hg H, J and K2/R1. A similar pattern should be found on the female side of the Indian gene-pool, where daugthers of the pale-arctic Ursula, Tara and Helena seem to have accomanied the various y-lines from the post-glacial forefather F.

Before the arrival of these sub-arctic people there must have been a native population in the sub-tropical/tropical parts of India that had survived the Dryas-periods down south, in a rather large number.

The population of this tropical, paleolithic refugia would undoubtly have developed their own toungue, different from the memebers of the arctic refugia. Thus the first introgerssion of palearctic genes may be ascribed to a small group of northerns emigrants, arriving some 11.000 years ago, as the pale-faced Eurasians started to reproduce and spread, obviously seeking contact with tropical survivors of the late Paleolithich (LGM-YD).

Their impact on the indian gene-pool may still be seen, eventhough their lingusitical impact was rather small. Thus the Dravidian language would remain basically different from the 'proto-I-E' of the northern Caucasians.

A stronger impact is seen in northern India, corresponding to the spread of y-dna J, a group frequently found among I-E-speakers - from India to Spain and Scandianvia.

Y-dna J's 'brother-lines' (hg G, I and their cousins of hgs R1a/b) are also based in areas where the I-E language-family have been dominant. In this respect y-dna H, peaking in the Dravidian India, represents an 'outlier'.

batman said...

It should be added that the genetic impact of y-dna H and J in the Indian subcontinent may have been the result of civil connections and intermarriages between native and caucasian nobilities. With an ancient cast-system that was pre-dominant in the reproduction-cycle of the entire population a change of y- og mt-dna at the top of the pyramid would have an impact on the entire population, five generations down the road.

Thus a dynastical change - given the patrilinear principle - would have a direct impact on the male side of the comming generations, from centre to periphery. Modelling this principle as it grow a certain population/culture through time would thus explain the facination for the (exact, equilibrate) megaliths - like cumuli, stupas and pyramids - made by the ancient civilizations. The ziggurats and pyramids found along the equatorial belt of the globe seems to prove that the arcitecture of this specific symbolism spread due to the knowledge and skills of an ancient school of masons with inter-continental connections.

Such as the 'aryans' of the historical records, such as the Vedic litterature. Which may well correspond to the modern phrase 'Caucasians' - as the common denominator of the post-glacial Europeans that descended from the arctically adapted Cro-Magnons of the Solutrean/Magdalenien/Hamburg cultures.

Undoubtly, as the survivors of an arctic climate and environment, the ice-age survivors of NW Europe, known as the Ahrensburg-Swidrien-culture, would have a variety of skills and knowledge unknown to their tropical cousins, that survived the Younger Dryas in large 'refugias' within the tropical climates of Africa, India, SE Asia and S America.

The proto-Dravidians must have belonged to the Indian 'refugia', which obviously made up a distinct group during the late Paleolithic. Thus the question if one can find other and older y-lines from the Indian sub-continent - from the chieftains reigning before any aryan/caucasian diplomats or noblemen arrived from the 'high north' ("Mt Meeru") arrived. Especially on the y-side, where the 'dynastical' realm of the ancient civilizations seems is able to change the y-side of entire popultions in no mnore than five generations...

The female part of the ancient breeding-system must have played out somewhat different - as it is rather obvious that the paleolithic mt-groups from S and SE Asia is still dominant in these regions.

Shaikorth said...

Karl, you're quoting Basu et al which used ADMIXTURE to infer ASI and ANI resulting in inflated numbers compared to Moorjani et al's formal testing.

Karl_K said...

@Shaikorth

Sorry, I didn't mean to put that quote. How about...

"Our analysis documents major mixture between populations in India that occurred 1,900–4,200 years BP, well after the establishment of agriculture in the subcontinent. We have further shown that groups with unmixed ANI and ASI ancestry were plausibly living in India until this time."

And I have nothing else to change about what I said.

I just re-read the Moorjani et al. paper. To what were you referring as a lobotomized version of the truth? I can't see where they ever disagree with an Indo-Aryan migration. They are careful to only refer to dates where admixture ended, but they don't say when it began. And how would they be able to say that without ancient DNA anyway? Even you can't accurately say that.

Is this the statement?

"On the contrary, a recent study that searched for West Eurasian groups most closely related to the ANI ancestors of Indians failed to find any evidence for shared ancestry between the ANI and groups in West Eurasia within the past 12,500 years (although it is possible that with further sampling and new methods such relatedness might be detected)." ?

Davidski said...

@Karl

Here's the comic relief from the Moorjani paper that I was referring to...

It is also important to emphasize what our study has not shown. Although we have documented evidence for mixture in India between about 1,900 and 4,200 years BP, this does not imply migration from West Eurasia into India during this time. On the contrary, a recent study that searched for West Eurasian groups most closely related to the ANI ancestors of Indians failed to find any evidence for shared ancestry between the ANI and groups in West Eurasia within the past 12,500 years3 (although it is possible that with further sampling and new methods such relatedness might be detected). An alternative possibility that is also consistent with our data is that the ANI and ASI were both living in or near South Asia for a substantial period prior to their mixture.

Even at that time it was already plausible that there was gene flow from close to Europe to South Asia during the Bronze Age because the phylogeography of R1a was strongly suggesting it. And I wrote about it at the time.

Here's my non-diplomatic assessment of the data presented in the paper: South Asia has seen multiple waves of population movements from West and Central Asia since the Neolithic, including the Indo-Aryan invasion during the Bronze Age, which reshaped the genetic structure of the region in a remarkable way. Indeed, the Aryan invasion introduced into South Asia one of the most common Y-chromosome lineages there today: R1a-Z93 or R1a1a1b2*. Obviously, scientists working on the problem of the peopling of South Asia really need to become aware of this marker, and in particular its very close relationship to the Northern and Eastern European-specific R1a-Z283.

http://eurogenes.blogspot.com.au/2013/08/moorjani-et-al-on-recent-population.html

And yeah, you're quoting the Basu paper, which isn't worth quoting. You seem to have too much faith in these things. Most of the time it's just garbage.

Jaydeepsinh Rathod said...

This is exciting news. All in all, potentially we could have aDNA of 19 samples from the IVC site of Rakhigarhi. If this goes through well and we have a paper in collaboration with the Reich lab in some months, it would be momentous. Rakhigarhi is now officially the largest known Harappan site ever. According to its current excavator, Vasant Shinde, it probably is as large as 550 hectares, which is twice as large as Mohenjo Daro, the next largest. It also dates to a slightly earlier period than Harappa & Mohenjo Daro. Rakhigarhi could well have been the capital of the Harappan Civilization. As far as genetics is concerned, Rakhigarhi is in the state of Haryana, which along with Punjab, is the place of the Jat people. If I am not mistaken, among Indians, the greatest affinity towards the steppe population is displayed by these Jat people.

One more point, the recent paper "Punctuated Bursts...", dated the expansion of R1a-Z93 in South Asia between 3300 BC - 2000 BC. This, if correct, is earlier to the so-called Aryan Migration in South Asia. And it is parallel to the highpoint of the Harappan Civilization which is dated from 3500 BC - 1900 BC. The Z93 people could well have the elites of the Harappan Civilization. Evidence of Z93 presence in the steppe as early as 3000 BC is not inconsistent with this scenario.

The Harappans were a very widely spread out civilization. Its contacts with the Central Asians date to the 4th millenium BC at the very least and some movements from Central Asia into the steppe cannot be ruled out. Michael Frachetti has already argued for a Inner Asian Mountain Corridor, which connected SC Asia to the steppe, all the way upto Altai, from around the 3rd millenium BC. This corridor may have existed even earlier but not much of archaeological research into that earlier period has been done, either along the IAMC or Central Asia or in South Asia.

One last point - The South Korean scholar seems to indicate Neolithic contacts with the Iranian plateau. This is of importance. If one looks at the Y-Dna of Iranian people, it is astonishing to how much of HGs the Iranians share with South Asians - HGs C, G, Q, H, L, T, J1, J2, R1, R2. So it appears that people to people contacts with the Iranians & SC Asians seems to have been quite extensive and from atleast the Neolithic. Now, it should also be made clear that along with the Near East, the region East of Zagros, has also been considered by many as the second center of animal & plant domestication. The exact center of domestication east of Zagros is not clear. However, it seems to argue that the influence of this second center of domestication should have spread across the Iranian plateau and the Indus civilization. Hence the Indus civilization & the Iranian plateau should be considered as important centers of early human civilization on par with the Near East. It was not a backwater by any means.

Parallel with this is the fact that the earliest clades of R1a are found in Iran. Since the people of Iranian people were already in contact with early Indians since the Neolithic, some of this R1a should have spilled into South Asia (if R1a was not already present in South Asia) through Iran quite early. The presence of R1a-Z93 into South Asia therefore, does not need a steppe input.

Jaydeepsinh Rathod said...

Finally, the archaeological evidence for steppe migration into South Asia does not exist. Let me quote for everyone's benefit, the words of James Mallory,

"...I fear that there would still be a sufficient assemblage of words to indicate that both Europeans and Asiatic Indo-Europeans shared inherited words for both livestock and arable agriculture (if someone can prove they did not, this would make things easier for many of the homeland models). Thus, any solution to the homeland problem must be able to explain how we can recover cognate terms associated with farming from Ireland to India."

"This is indeed the problem for both the Near Eastern and the Pontic-Caspian models and, following the logic of this analysis, the Bouckaert model appears to be in the same boat. All of these models apparently require the Indo-European languages (including their attendant agricultural vocabulary) to be superimposed/adopted by at least several major complex societies of Central Asia and the Indus."

"In any event, all three models require some form of major language shift despite there being no credible archaeological evidence to demonstrate, through elite dominance or any other mechanism, the type of language shift required to explain, for example, the arrival and dominance of the Indo-Aryans in India."

"But all theories must still explain why relatively advanced agrarian societies in greater Iran and India abandoned their own languages for those of later Neolithic or Bronze Age Indo-Iranian intruders."

http://cyberleninka.ru/article/n/twenty-first-century-clouds-over-indo-european-homelands

The fact is, the Iranian plateau & the Indus was one of the two early centers of plant & animal domestication. Why would these people abandon all their agricultural vocabulary and accept agricultural terms of the invading steppe people who were certainly less advanced than them in agriculture ? Does it make sense ?

Nirjhar007 said...

Hey Jaydeep,

http://njsaryablog.blogspot.in/2016/05/jp-mallorys-current-assessment-on-proto.html

Davidski said...

@Jaydeepsinh Rathod

A lot of the time you're ignoring basic facts that would otherwise render your theories untenable. Here are two examples.

One more point, the recent paper "Punctuated Bursts...", dated the expansion of R1a-Z93 in South Asia between 3300 BC - 2000 BC. This, if correct, is earlier to the so-called Aryan Migration in South Asia.

No, they didn't. They dated the expansion of Z93 to 3300 BC - 2000 BC. Using modern DNA it's impossible to say where this expansion happened. What if all the samples came from colonists on Mars? Would you then claim that that the authors dated the expansion of R1a-Z93 on Mars between 3300 BC - 2000 BC? Surely not.

The steppe is where the expansion of Z93 first began, and it's a long way from South Asia. So the date of the entry of the Aryans into India at around 1500 BC looks safe.

Parallel with this is the fact that the earliest clades of R1a are found in Iran.

This is irrelevant because...

a) You don't know where those clades were 4,000 years ago.

b) 99% of R1a is M417 derived, and M417 looks very much like it expanded from Eastern Europe during the Bronze Age.

Nirjhar007 said...

This is irrelevant because...

a) You don't know where those clades were 4,000 years ago.

b) 99% of R1a is M417 derived, and M417 looks very much like it expanded from Eastern Europe during the Bronze Age.


From first point , yes, we need to SEE FIRST were it was at all absent in Iran , India etc.

Regarding, the second , without aDNA from those areas . Its plain hogwash...

Davidski said...

Nirjhar,

Read my post from 2013 very carefully. Try your best to get your head around it. This is pretty much it. You'll have to accept it at some stage.

Here's my non-diplomatic assessment of the data presented in the paper: South Asia has seen multiple waves of population movements from West and Central Asia since the Neolithic, including the Indo-Aryan invasion during the Bronze Age, which reshaped the genetic structure of the region in a remarkable way. Indeed, the Aryan invasion introduced into South Asia one of the most common Y-chromosome lineages there today: R1a-Z93 or R1a1a1b2*. Obviously, scientists working on the problem of the peopling of South Asia really need to become aware of this marker, and in particular its very close relationship to the Northern and Eastern European-specific R1a-Z283.

http://eurogenes.blogspot.com.au/2013/08/moorjani-et-al-on-recent-population.html

Nirjhar007 said...


You'll have to accept it at some stage.
Only If I get drugged . No sane person will accept that conclusion of yours, at least, at the moment and am pretty much sure in any distant future as well! .

BTW, lets forget about something about which we can't do anything right now. Can you please show me some d-stats or ADMIXTURE of the ''Ori'' group you mentioned ?.

I suspect they are actually these folks-
https://en.wikipedia.org/wiki/Kurukh_language

George Okromchedlishvili said...

It's no point arguing with guys like this. In the culture he comes from this whole ancestry thing is of huge importance and he won't accept the reality that shatters his ethnocentric ideas.
I've tried to have a discussion with some Ossetian nationalist who claim G2a is an IE marker and Ossetians are direct descendants of Steppe nomads. Needless to say it was futile - their ethnic pride lasts on the idea of ethnic and cultural continuity of their nation even though all facts point to the contrary (needless to say I've mentioned that Georgians are also not "pure" and probably speak a language that was not initially carried by CHG-type "native" Caucasians).

Davidski said...

N,

Google is your friend in cases like this. But here...

http://mbe.oxfordjournals.org/content/28/2/1013.full

Can't vouch for the usefulness of the analysis though.

Nirjhar007 said...

George,
I don't support OIT ,

David,
You weren't specific enough .Thank you for the paper !.

bharatiya said...

While this team of researchers might come to some sloppy inferences, it would be due to their ineptitude rather than a result of manipulating data under political pressure. Most of those interested in the AIT/AMT (both against or for) out of political considerations care little for the nuances of genetics. After all most of political statements on the AIT are based on outdated versions of the theory from the early 20th century or even earlier. Ultra-Traditionalists are going to remain steadfast on dates in the remote past. Christian Missionaries are going to continue targeting tribals and lower castes calling for bans on Hindu festivals that supposedly celebrate the genocide of Black aborgines. As for the leading OIT theorists (like Elst & Talageri), they have already disowned the utility of genetics in studying the expansion of linguistic famillies

Shaikorth said...

Frank,

"2) Indian admix into the West Eurasian (NEur, SEur, Arab, Levant-Cauc) root;"

This migration edge can also mean that there is West Eurasian admix in India since Treemix edges can be flipped.

In fact, the supplements suggest this is the case rather than the other way around. They do f3 admixture tests between all their components, and none of those four components give a significant negative signal as a mix of Indian and something else. Meanwhile the Indian component gives significant signals as a mix of any of them and something else (Chinese, Southeast Asian, Melanesian etc).

imam-din said...

I somehow agree with David that there is a strong "indian national narrative" that everything related to Sanskrit language and Vedas is indigenous to modern india and so were the sanskrit speaking folks. Some even think that Vedas and Sanskrit in india is more than 10,000 years old and all that kind of fairy tales.

But in reality the Sanskrit language and composers of Vedas which are nothing but sanskrit poems composed by ancient sanskrit speaking saints called "rishi" who were supposedly the offsprings of the rigvedic gods mentioned by the same poets/rishis in the rigveda. One thing is clear that Sanskrit language is not developed anywhere in modern india, pakistan or afghanistan. If you listen any rigvedic sanskrit hymns on youtube , the language sounds too much like russian or ukrainian language especially with high incidence of consonant clusters at the beginning of the words such clusters like "prch" or "prsh" etc. If you want to know how Rigvedic sanskrit sounded like, just listen to modern dardic languages such as Kalasha, Shina or Kashmiri (Koshur) they all sound very different from punjabi/urdu/hindi etc. type languages because the former don't have much dravidian substrate and are still relatively close to rigvedic sanskrit in their vocabulary, phonetics and grammatical structures.

imam-din said...

Can we get rid of obsolete terms like "indian subcontinent" as this term is not used by any modern academic or political world organization. If someone wants to refer to modern country of india it is better to stick with india only. And if the purpose of usage is to the broader region then "south asia" is the standard contemporary term. Someone here is also using terms like "SSVC" which not one can have a clue, again the indian narrative at work instead of calling "indus valley civilization IVC" someone is trying to bring his narrative by developing his own terms like "SSVC" which I believe would be "Sarasvati Sindu valley civilization", and then the same people say that there is no "indian narrative" involved here.

bharatiya said...

@imam-din: While the present geographical region is best described as South Asia, to avoid confusion with The Republic of India, it is perfectly acceptable to call the entire culturul and geographic region India while talking about its past. It was known as such to the Greeks, Persians, Romans and other people in the past. While proto-Indo-Aryan might have its roots elsewhere Sanskrit was developed in South Asia. The Rigveda was composed in the basin of the Sindhu and the Saraswati (Ghaggar-Hakra) with a few possible reminiscences of Afghanistan. At its earliest stage it had adsorbed numerous distinctly South-Asian features. While it is true that the Dardic languages preserve several archaic features lost in most other branches of IA they possess influences from the substrates spoken in those regions besides numerous phonetic and structural innovations. Dravidian was never spoken in the regions currently speaking Panjabi or Hindi. Instead the substrate language of the Gangetic Basin (Masica's language x) possibly was related to Kusunda. By no stretch of imagination does modern Russian sound closer to sanskrit than the modern IA languages.

postneo said...

I think they will find z93 and perhaps ANI in a more unadmixed state. chances of finding ASI will be less since this is Haryana.

"The South Korean scholar seems to indicate Neolithic contacts with the Iranian plateau."

This I think is based on much older work by Shinde not specific to Rakhigarhi and is based on trade artefacts from other sites.

Davidski's inference of "BA steppe" is actually very late and involves a sudden "Sintashta migration" in 1900 BC. I think it will be proven false. Its based on taking a some endogmous populations e.g gujarati A and gujarati B and saying that one is steppe like. You could take any bunch of endogamous groups in India and do that. You could claim the Vedda are closer to andronovo compared to sinhalese or some such junk.

On the other hand Svante paabo and Reich's inference that ANI and ASI started mixing during the bronze age is worth noting. Its based on segment lengths.

Davidski said...

I think they will find z93.

What do you base this statement on, any evidence or just religious faith?

Rami said...

Davidski the expansion of R1a-Z93 is now dated to 2.5-2.0 Kya BC in South Asia, not 1500 BC, that changes the dynamics considerably. You keep pushing this narrative as if the IVC ended because of them as if your grand dad had some part in it.

Davidski said...

@Rami

The expansion of R1a-Z93 is now dated to 2.5-2.0 Kya BC in South Asia.

When the steppe people started expanding on the steppe 2.5-2.0 Kya BC, they eventually ended up in India with their Y-chromosomes.

Take some time to think that over before replying.

Chad Rohlfsen said...

CHG is not BE plus ANE. It is no closer to ANE than WHG. Higher CHG than Yamnaya would make SC Asians have much less affinity to MA1 than Yamnaya. I would bet more on CHG plus a lot of MA1, then that MA1 decreases with LBA steppe groups coming in.

postneo said...

"What do you base this statement on, any evidence or just religious faith?"

I have answered this before. its because z93 has no specific structure in south asia, when you look at downstream clades. But outside the subcontinent there is structure. Unlike you, I attach no significance to z93* in modern populations, to me It actually shows people at the periphery. All this cannot be explained by a sudden Sintasha like incursion in 1500 BC.

Anyway we will know soon ... who knows

Imam Din: I would not get too obsessed with vedic and sanskrit. I don't think vedic was the only IE language. There is evidence of non vedic IE in india, perhaps older or parallel dialects.

for vedic, I agree with your statement on compounds but there are surprising also resemblances in speech cadence with modern dravidian lacking in current north indian languages. In terms of retroflexion though its unlike dravidian. in voiced aspirates it gets more firmly rooted in parts of eastern and central india away from the northwest. perhaps they were located further west during IVC day but moved.

Anyway the vedic language is its own beast and cannot be stereotyped in simple terms.

Davidski said...

All this cannot be explained by a sudden Sintasha like incursion in 1500 BC.

Yes it can. Every single Z93 lineage in South Asia today is from the Bronze Age steppe.

None of the arguments you have provided here to counter this are valid.

Davidski said...

This is very funny.

I found the paper referenced in the concluding remarks of the Harappan poster presentation. Starting at the bottom of page 144...

Our previous studies suggested that a second biological discontinuity occurred in the history of the Indus Valley after the collapse of the Harappan Civilization around 1700 B.C. ... One of the sources that has been posited for this gene flow has been central Asia. The results of this study and that of others suggests that a central Asian source does appear to be supported by several lines of biological data.

How did they miss this bit? Kinda seems important to me. Here's the link...

http://pages.uoregon.edu/jrlukacs/Dr.%20John%20R.%20Lukacs%20Website/downloads/Document47.pdf

postneo said...

http://pages.uoregon.edu/jrlukacs/Dr.%20John%20R.%20Lukacs%20Website/downloads/Document47.pdf

You seem to have read selectively.

In Imangaon tribal god and uppr caste maratha turn out more autocthonous and lower caste mahars seem intrusive. discontinuity btw neolithic and chalcolithic is drastic whereas 1700 bc is gradual etc. Nobody has missed anything, Every locality will have its own trajectory and more importantly neither imangaon or mehargarh are guranteed to have homogenous populations to start with.

As for the weak conclusion vis a vis central asia, Its conventional to say this in all publications regardless of evidence in the last 50 years at least.

Davidski said...

Not sure about weak. In any case, it's looking pretty good nowadays.

bharatiya said...

@Postneo
For most of the last fifty a years linguistics was the only major evidence for an Aryan migration/invasion, with most of the archaeological evidence being inconclusive. Genetics has changed all this, particularly recent ancient DNA findings in the steppe. A bronze age migration from the steppe is almost certain as per our current knowledge. While certain details maybe uncertain and the chronology may not exact looks like the AMT happened after all

Jaydeepsinh Rathod said...

Dear Nirjhar,

I am well aware of your blog as well as that of Giacomo. Thanks for the link anyway. The new 'throne of truth' webblog seems interesting.

David,

"No, they didn't. They dated the expansion of Z93 to 3300 BC - 2000 BC. Using modern DNA it's impossible to say where this expansion happened. What if all the samples came from colonists on Mars? Would you then claim that that the authors dated the expansion of R1a-Z93 on Mars between 3300 BC - 2000 BC? Surely not.

The steppe is where the expansion of Z93 first began, and it's a long way from South Asia."

You say that using modern DNA its impossible to say where the expansion of Z93 or any lineage for that matter took place. But nor can we do that by only having ancient DNA from the steppe without aDNA from South Asia or West Asia.

Secondly, what you're implying in essence is that, all of the Z93 lineages present in South Asia already underwent expansion while in the steppe before coming to South Asia. It implies a population explosion of Z93 on the steppe. So why is Z93 so scarce on the steppe today ? Did almost all of it go down South ?

You're also implying that once the Z93 lineages reached South Asia and spread across the length & breadth of it, it did not undergo any lineage expansion because all of it had already happened in the steppe. But how is it even possible ? How could have Z93 spread so much in South Asia without any lineage expansion ?

(conti...)

Jaydeepsinh Rathod said...

(...continued)

"a) You don't know where those clades were 4,000 years ago.

b) 99% of R1a is M417 derived, and M417 looks very much like it expanded from Eastern Europe during the Bronze Age."

True that we don't know where these clades were 4000 years ago. But you want to assume that they were not there either in Iran or in SC Asia. I would say unless we have aDNA from this region, there is no point in denying such a possibility. Assuming HG R originated in Siberia, we already know that it was at the other end of Europe in Italy in the form of R1b already by 14000 BP. In short carriers of R HG were quite a mobile people. So why could R subclades not have gone down South into Central & South Asia already in the Mesolithic ?

After my last comment, I reread Underhill's paper as well as some earlier studies on Indian populations. One of the deficiencies of the Underhill paper is the less than adequate sampling of India & Pakistan which collectively hold 1.5 billion of humanity. They only had about 1000 samples representing 1.5 billion people while from Iran they had nearly 2000 samples when its population is only 1/10th of the subcontinent. This makes no sense when you are actually trying to find out the rare basal clades of R1a.

Hence I went through 2 earlier reports on Indian populations.

http://www.nature.com/jhg/journal/v54/n1/full/jhg20082a.html

According to the above paper, India has the presence of paragroup R*, R1* as well as R1a* (Suppl. Table 1)

According to this other study,

http://www.krepublishers.com/02-Journals/IJHG/IJHG-08-0-000-000-2008-Web/IJHG-08-1-2-001-256-2007-Abst-PDF/IJHG-08-1-2-097-08-349-Trivedi-R/IJHG-08-1&2-097-08-349-Trivedi-R-Tt.pdf

there is presence of paragroups R*, R1 & R1a in South Asia (Table 2A).

According to this,http://dispatchesfromturtleisland.blogspot.in/2015/05/the-genetics-of-kalash-people.html, the Kalash people have relatively high incidence of R*(6.8 %) & R1*(2.3 %) in its population besides the presence of R1a.

When one looks at all of these collectively, how can one dismiss the possibility that West Asia or South Asia could have harboured basal clades of R1a (including M417) in the prehistoric period as well ? Let us also not forget that the SC Asia has the CHG/Gedrosian component which likely entered the steppe during the Yamnaya phase or earlier during the Eneolithic phase. It is unlikely to have come from the Caucasus due to absence of R1a in the Caucasus but it could still have come to the steppe from SC Asia.

If there was any migration from the Sintashta people into South Asia, why does admixture not detect EHG component into SC Asia ? Or for that matter the ENF component that is present in Sintashta ?

Jaydeepsinh Rathod said...

"I found the paper referenced in the concluding remarks of the Harappan poster presentation. Starting at the bottom of page 144...

Our previous studies suggested that a second biological discontinuity occurred in the history of the Indus Valley after the collapse of the Harappan Civilization around 1700 B.C. ... One of the sources that has been posited for this gene flow has been central Asia. The results of this study and that of others suggests that a central Asian source does appear to be supported by several lines of biological data.

How did they miss this bit? Kinda seems important to me. Here's the link..."

This is an old study. A more recent and much more comprehensive study is the following

https://www.academia.edu/8627502/Grades_Gradients_and_Geography_A_Dental_Morphometric_Approach_to_the_Peopling_of_South_Asia?auto=download

As you can see, there is no affinity of the samples from Timargarha (1400 BC - 800 BC) towards Central Asian BMAC samples. There is no evidence of any movement around 1500 BC from Central Asia into South Asia, based on the above. However, studies based on Dental morphology or odontometrics have their limitations and they cannot arrive at any definite conclusion.

If you're interested there is this other study, co-authored by Victor Mair

https://www.academia.edu/7065071/Horse-Mounted_Invaders_From_the_Russo-Kazakh_Steppe_or_Agricultural_Colonists_From_Western_Central_Asia_A_Craniometric_Investigation_of_the_Bronze_Age_Settlement_of_Xinjiang

Karl_K said...

@Jaydeepsinh

I totally agree.

Just because several Corded Ware samples carried the M417 and Z645 mutations (which are ancestral to Z93), and the Sintashta and Srubnaya cultures (both of which appear to have been rich in R1a-Z93), didn't live in South Asia or have any obvious South Asian specific genetic signatures?

It just makes sense that we need a lot more data to have any clue what happened.

It is not only plausible, but highly likely (based on the clear lack of evidence from key areas), that the Indo-European languages and the R1a-Z93 haplogroup arose in South Asia very early. And from there they spread to Europe and the steppe.

Davidski said...

@Jaydeepsinh Rathod

Hence I went through 2 earlier reports on Indian populations.

Those papers are useless. If you can't find the evidence that you're looking for in the latest papers with deep Y-chromosome sequence data, then you don't have the evidence yet.

If there was any migration from the Sintashta people into South Asia, why does admixture not detect EHG component into SC Asia ? Or for that matter the ENF component that is present in Sintashta ?

Of course there's both European HG and European Neolithic admixture in South Asia among the Indo-European speakers there. I don't know how you came to the conclusions that there wasn't. Just have a look at the Admixture bar graphs from the Haak et al. paper from last year.

Karl_K said...
This comment has been removed by the author.
imam-din said...

I often read on these genetic forums that Steppes groups don't have "south asian" genetics when in reality the correct description should be that ancient steppe groups don't have "paniya/Paliyar" type genetic contribution. Otherwise every component in south asia that is other than "paniya/paliyar" is also present among ancient steppe populations, e.g. ANE, CHG/ENF and some minor east Eurasian components.

I think before making a "definite statement" that the genetic flow was only from Steppe towards south asia we should wait until we have got equivalent Bronze age, Neolithic and Mesolithic aDNA from south asia including areas south east of caspian sea (BMAC). I am increasingly thinking that to connect a whole language group to a particular "haplogroup like R1a1" is ridiculous. Think about it, is there any area in modern world where a particular haplogroup has a monopoly on a language, obviously not. Languages are spoken by functioning societies irrespective of how many different kinds of male haplogroups live in those societies. It has always been like this since human being developed societies after domesticating plants and animals with the advent of Neolithic revolution.

postneo said...

"I often read on these genetic forums that Steppes groups don't have "south asian" genetics when in reality the correct description should be that ancient steppe groups don't have "paniya/Paliyar" type genetic contribution."

actually there is an ASI/papuan component that shows up for the first time in yamnaya and then in europe from Haaks paper. Even David has been forced to deal with it and labeled these under the "farmer wives umbrella"

You can compute components with any which way ... central asia would be more ANI vs ASI just due to geography. Similarly north india would have more central asia and south india would have less CA. Its just common sense geography not overly significant.


Z93 may have originated in the steppe and come to south asia but a massive adna replacement in 1500 BC is ruled out.

Language may be only loosely correlation with all this. But coming to linguistics, the assumption that vedic is the root of all IA itself needs to be examined. The germanic word "naked" is closer to bengali and marathi versions vs intervening sanskrit hindi and latin. This might mean that the PIE reconstruction on the indic side is simplistic and made artificial assumptions as vedic being the sole root of IA.

Davidski said...

Actually there is an ASI/papuan component that shows up for the first time in yamnaya and then in europe from Haaks paper. Even David has been forced to deal with it and labeled these under the "farmer wives umbrella"

There's no ASI/Papuan admixture in Yamnaya. You can't use any little bit of noise that you find to make an argument. At least try harder if you don't want to look like a complete idiot.

And what the hell is "farmer wives umbrella"? Ever heard of female exogamy in LN/EBA Europe?

You can compute components with any which way.

No you can't.

postneo said...

Ever heard of female exogamy in LN/EBA Europe?

Yes its a useless trope, pretty much the the norm in most sexually dimorphic primates and mammals, nothing particularly european or even human.

Davidski said...

News just in: the lab at Copenhagen Uni also got the samples, which means full genomes are on the way. See last paragraph here.

http://www.tribuneindia.com/news/haryana/democracy-has-roots-in-harappa/233774.html

Nirjhar007 said...

Thats not a new news bud..

Rami said...

David you have your head so far up your pompous ass sometimes you cannot tell left from right. Expansion of R1a Z93 in South Asia is not the same as expansion on the Steppe, and the implications are quite different from your golden chariot wet dream invasions. Also the the Indo Iranians who arrive on the scene in South Asia were a mixed up lot , as these people admix with the locals they come across, in this case the BMAC peoples in Central Asia. Indo Aryan related tribes were in South Asia long before Iranic tribes wiped out whatever proto Indo Aryan groups were on the steppe. Just as the Sintashta were quite removed from the original Yamnaya Indo Europeans, the same can be easily said for the Avestan Iranics who moved to Iran and the Indo Aryans who entered South Asia.

Davidski said...

Expansion of R1a Z93 in South Asia is not the same as expansion on the Steppe.

It's the same thing you idiot.

Modern South Asian Z93 lineages are from the steppe, every one of them, and they contain information about the initial expansion of Z93 on the steppe around 4,000 years ago.

You're claiming that this expansion 4,000 years ago happened in South Asia. But that's because you're confused, and can't get your head around the fact that there was no one in South Asia with Z93 at that time.

Krishna Rao said...

Your article is making conclusions before the evidence is out. Not sure where you learned to write, but I would wait and see before making such bold statements about OIT. We don't know yet. Wishful thinking does not make up for evidence.

Davidski said...

@Krishna Rao

Problem is that you don't understand the evidence that we already have from the available ancient and modern DNA.

The game is practically over for OIT. It's just that some people like you don't want to believe it until the final nail is in the OIT coffin.