The Genetic Origin of the Indo-Europeans

[J Man]

The new Eneolithic era I11828/I31755 J2a (J-M319) sample from the Lower Don Krivyanskiy-9 grave 19 site is really interesting. According to the paper autosomal wise this sample is genetically quite close to Yamnaya although of course he pre-dates Yamnaya. It will be very interesting to see this sample's G25 results once the data comes out.

Also this sample is described as being similar to "Serednii Stih" samples in terms of burial pose, artifacts and date and his CHG like ancestry is apparently more similar to Mesolithic CHG compared to Aknashen.

Anyway interesting to see a J2a sample from the Eneolithic Steppe context.


I11828 6257 Krivyansky Russia T2a1b J-M319 J2a1a1a2b1b

[Jerome]

Finally, we plot in Fig. S 11 the proportions of Table S 29 which visually demonstrates all our inferences. The contrast between the Don-Dnipro and Volga in terms of Ukraine_N vs. EHG ancestry; the presence of Aknashen ancestry in Maikop, Remontnoye, the Serednii Stih, and the Yamnaya, but not in Golubaya Krinitsa or the Volga; and the contribution of a population like BPgroup of ancestry in populations across the entire region except the far north of eastern Europe (Karelia) and starting from at least the mid-6 th millennium date of the Golubaya Krinitsa individuals, one of which has it (GK1) while the other one does not (GK2).

Interesting,so what they are saying is that CHG weren't PIE speakers but rather aknashen was.
They very discretely put this in footnotes by saying 'that CHG was present in golubaya/don, khvalynsk but these people still weren't PIE speakers" What creates PIE speakers is the 'Aknashen ancestry In yamanaya,progress,sredny stog,remotnoye to the exclusion of already present CHG"

This is just an intermediate hypothesis by which aknashen ancestry moving to north Caucasus creates PIE.

So south Caucasus ancestry moving north=PIE.

Basically,PIE is from south Caucasus but also from north Caucasus.

[Jerome]

They leave all scenarios open, but claim based on the current evidence that hypothesis B is a bit more likely. I highly suggest you read the last 20-40 pages of the supplement, everything will become clearer.
The way I understood some of the rationale, is that BPgroup is mix of Khvalyns + Caucasus. And BPgroup + SShi (Sredni Stih) gave raise to Core Yamnaya. "and Serednii Stih people had therefore not only BPgroup-related ancestry from the south (as people on the Volga did), but also some Aknashen/Maikop-related ancestry." And BPgroup inflow is necessary to model Anatolian BA. 

Between the North Caucasus piedmont and Lower Volga population, mixtures were taking place during the 5 th and 4th millennia BCE between populations with ancestry characteristic of the Neolithic and Eneolithic Caucasus (represented by Aknashen and Maikop) and Lower Volga-North Caucasus Eneolithic populations (represented by BPgroup), both of which had substantial proportions of CHG ancestry from the earlier hunter-gatherer periods. Steppe ancestry was present in the later Maikop (Table S 9) and Remontnoye (Table S 6, Table S 7) individuals from this area. It was also present further south in Chalcolithic Armenia at Areni-1 cave, but there it was added to a different, Masis Blur Neolithic population (Table S 11). 5 Thus, the Caucasus area was seeing admixture in both south-north and north-south directions: Remontnoye had Aknashen Neolithic/Maikop ancestry and Maikop and Armenian Chalcolithic had steppe ancestry.

People out of this Caucasus-Lower Volga admixture zone people flowed outwards: along the Volga where all the “southern” ancestry can be well explained as of BPgroup origin alone (Table S 14) and into the Don-Dnipro area where it interacted with the descendants of hunter-gatherers of the Dnipro-Don forming the Serednii-Stih cline. A “Pre-Yamnaya” population quite like the Core Yamnaya was at the other end of 183 the cline (Table S 20) and Serednii Stih people had therefore not only BPgroup-related ancestry from the south (as people on the Volga did), but also some Aknashen/Maikop-related ancestry.

We have thus argued that Yamnaya was formed when people of “eastern” Aknashen-Maikop/BPgroup origins (of proximate Remontnoye-related origins) moved westward and admixed with people of the Serednii Stih culture. Motivated by this, we tested whether we could use this framework to jointly model all these populations involved in our reconstruction:

Observe also the previously mentioned contrast between Eneolithic populations of the Don-Volga who did not require Aknashen/Maikop ancestry but can be modeled with BPgroup/PVgroup ancestry alone (Fig. S 2, Fig. S 3) with those of the Serednii Stih culture of the Don-Dnipro area that also had Aknashen-related ancestry (Fig. S 9).

In Table S 27 that only models involving Aknashen_B/Azerbaijan_C and BPgroup/PVgroup and Igren_o ancestries have no losses in the tournament, the full results of which are shown in Table S 28. While no unique solution emerges out of this tournament, it is useful to weigh our confidence in the different models. With the knowledge that we have only partially sampled the genetic variation of the Caucasus, Lower Volga, and Dnipro-Don areas, it is nonetheless interesting that the 3-way models directly recapitulate the conclusions we reached by exploring 2-way models: that all three of these areas contributed to the formation of the Yamnaya.

Finally, we plot in Fig. S 11 the proportions of Table S 29 which visually demonstrates all our inferences. The contrast between the Don-Dnipro and Volga in terms of Ukraine_N vs. EHG ancestry; the presence of Aknashen ancestry in Maikop, Remontnoye, the Serednii Stih, and the Yamnaya, but not in Golubaya Krinitsa or the Volga; and the contribution of a population like BPgroup of ancestry in populations across the entire region except the far north of eastern Europe (Karelia) and starting from at least the mid-6 th millennium date of the Golubaya Krinitsa individuals, one of which has it (GK1) while the other one does not (GK2).

Hypothesis B harmonizes with all known facts and the results of our reconstruction of Yamnaya origins strengthen it, as the Yamnaya do indeed have ancestry from the south: both early ones via their BPgroup ancestors which experienced gene flow from the Caucasus and contributing to the Serednii Stih and Volga clines; but, also later ones via the migration of Remontnoye-related people (who also had Maikop/Aknashen ancestry). What was only indistinct before (the CHG ancestry in the Eneolithic steppe and the extra Anatolian-Levantine ancestry in the Yamnaya4 ) has now come into better focus.

Hypothesis B: a Caucasus-West Asian origin of Indo-Anatolian origins is strengthened by the finding of early migrations from the Caucasus into the Volga/Don-Dnipro Eneolithic populations followed by later Maikop/Armenian Neolithic ancestry into the ancestors of the Yamnaya. This 214 hypothesis also maps to the transformation of Chalcolithic and Bronze Age central/western Anatolia which saw half to all its Neolithic population replaced.

It can all get very confusing. Had to read the analysis part multiple times, and still not sure I understand it right.

Yes,it's interesting how they are doing semantic jugglery for later future data and studies.

Pretty decieving and confusing.
This type of langauge seems to be very 'lazarid-ish,since lazaridis often does this kind of juggling and obfuscation,it's very apparent on his Twitter.

[old europe]

What they seem to be saying is that EHG isn't pie but Something like aknashen/Mesopotamian+Kelteminar/TTK+CHG+EHG is PIE.?

I think what they are trying to do is just mix up/fit the southern arc with david anthony's hypothesis of Volga -caucasus and Sredny

They are just trying to keep all parties happy,so they just mix it all up and force-fit theories.

I disagree with you on same issue but I think that with this post you nailed it. This paper defends a revisted version of the southern arc thesis. They only move the allegedly PIE genetic cluster from south to north. That genetic cluster  did not work south of the caucasus ( the authors admit that) and IMHO will not work if you move it 500 hundred miles north

[Southpaw]

I wonder how do we interpret this?

"Pre-Indo-Anatolian languages were spoken by at least some of these diverse ancestors, but living Indo-European languages trace their ongins to the Yamnaya expansion ca. 3300BCE out of (3) and the earlier Indo-Anatolian expansion ca 4400-4000BCE out of (1)"

Does the <"out of" one> have the meaning of "bar"/"except". And if so would this mean that Armenian, Greek and Albanian (3) and Anatolian IE (1) somehow diversified from the stem pre Yamnaya?

To me it looks clear: all the living IE languages, including Armenian, Greek, and Albanian, stem from 3, and only Anatolian stems from 1. It means that after the phase 1, Pre-Proto-Anatolian branched off, and only after the phase 3, all the other branches (possibly excl. Tocharian, which is not living any longer) branched off.

so basically their take that area number 1 is the urheimat is based esclusively on the anatolian language  split.
Well anatolian language arriving from the balkans ( and later on) would have been brought by a population that was basically nearly 100% EEF so quite difficult to spot them.

Lordy these R1bV136 folks are the greatest language teachers in history. Despite being a very tiny minority among every branch of IE living languages they are considered by the authors as the sole creators of PIE. Look a lot like BS to me frankly

I know i can google, but where is R1bV136 more common today? Do we have any ancient historical population aDNA where this clade was common?

[Jerome]

Actually in their supplementery,they also perfectly model Anatolia BA using aknashen +Neolithic Anatolia without any steppe and they just say 'There are various possible competing hypothesis and all of them can be true.....'
Which is pretty strange,they maintain an ambiguous position although in the image they show progress in Anatolia,but in the supplementary they don't insist on that.

I think they model Anatolia BA with the BP group just to give their newfound samples more importance and credibility to their theories.

Let's see what they have in wait for future,they also mentioned in supplementary about sampling Neolithic/mesolithic central Asia and Siberia more.
Would be interesting to get these samples.

[Jerome]

.....

[VladMC]

Target: TUR_Kaman-Kalehoyuk_MLBA:MA2200
Distance: 0.6849% / 0.00684904 | R5P
24.6 IRN_Hajji_Firuz_N
23.2 UKR_LN_o
22.2 TUR_Arslantepe_EBA
20.8 RUS_Maykop_Late_EBA
9.2 TUR_Tell_Kurdu_N


———-
Target: TUR_Ikiztepe_EBA:IKI016
Distance: 2.7653% / 0.02765270 | R5P
49.8 TUR_Camlibel_Tarlasi_En
20.4 RUS_Maykop_Novosvobodnaya_EBA
16.6 TUR_Catalhoyuk_Meso_Ceramic
8.8 GEO_EMeso_CHG
4.4 IRN_Seh_Gabi_LN

Target: TUR_Ikiztepe_EBA:IKI017
Distance: 3.1545% / 0.03154492 | R5P
45.4 RUS_Maykop_Novosvobodnaya_EBA
20.6 IRQ_Cayonu_EMeso
18.8 SRB_LN
8.4 TUR_Camlibel_Tarlasi_En
6.8 DEU_LBK_HBS

Target: TUR_Ikiztepe_EBA:IKI036
Distance: 3.1427% / 0.03142677 | R5P
36.4 TUR_Camlibel_Tarlasi_En
20.2 RUS_Maykop_Novosvobodnaya_EBA
19.0 DEU_LBK_HBS
15.4 IRQ_Cayonu_EMeso
9.0 TKM_Geoksyur_EBA

Target: TUR_Ikiztepe_EBA:IKI037
Distance: 4.1005% / 0.04100480 | R5P
45.4 TUR_Arslantepe_EBA
19.0 IRQ_Cayonu_EMeso
16.8 RUS_Maykop_Novosvobodnaya_EBA
13.2 TUR_Camlibel_Tarlasi_En
5.6 TUR_Tell_Kurdu_N

[Jaska]

There are some laws for the pathernal or patriarchal type of society.
- -
The mens in IE society are also some kind of solders, so they may need to go and fight.. But the home of IE family is generally controlled by the mother of such man. So this is the reason why the language is the same - because the mother of the man is controlling all.

The situation gets complicated when these people spread to the region of different societies: how they can spread their language outside their family homes? And how about the daughters of the Indo-Europeans, who are moving away, among a community of another language speakers? Indo-European speaking spots here and there might not be enough to replace the older language in the whole region.

The best argument for PIE being an EHG language is that out of the main three male lineages, 3 out of 3 are derived from them and not the Caucasian side or the assimilated people of the West.

Well, that is a correlation, yes.
But it is also a fact that the regions in Northeastern Europe were inhabited exclusively by the EHG populations until the Corded Ware times, and their languages – which were replaced by the advancing Uralic speakers from the 2nd millennium BC onward – were not Indo-European but Paleo-European.

Of course, it is theoretically possible that these EHG languages were distantly related to Indo-European. There were tens of thousands of years of time for these languages to diverge beyond recognition. Recent more resolute results have shown that there were several distinguishable EHG ancestries in different regions, probably also different languages.

And naturally, there are also other language families – even still living – which can be associated with the Caucasus HG ancestry. All I am saying is that it is very difficult if not impossible to try to find out where the Indo-European language lineage originally came from. Considering the Lower Volga Region, there was also a considerable root of Central Asian ancestry; in Majkop, a considerable root of Western Siberian HG ancestry – we cannot exclude the possibility that Indo-European was originally an Asian language.

[Jaska]

now I see their hypothesis. Pay attention everybody. Basically what they are saying is that WHG/EHG on the Dneper Don and even full fledged EHG on the Volga were not PIE. For them PIE starts when the CHG/TTK kicks in to form the northern caucasus-lower Volga cline. So the cluster that makes PIE is the CHG/TKK
so what the  paper is all about is a revisited "southern arc hypothesis" For them PIE is born north of the caucasus but thanks to a southern caucasian influence.

They do not explicate where the language came from: they only talk about Proto-Indo-Anatolian after the admixture. But it is true that the more western river valleys can be excluded at this point. So the potential candidates for bringing in the Indo-European language are the Lower Volga EHG ancestry, the Central Asian ancestry, and the Caucasus HG ancestry (perhaps even the Levantine ancestry, concerning their hypothesis B).

The carrying capacity of a territory limits the manpower that can be mustered by the inhabitants to defend themselves from invasion or to invade foreign territories. - - However, if you asked me to consider the likelihood that any population of present-day Europe might expand their territory and spread their language over the entire (sub)continent in the future, the Kalmyks would not be the first people to come to mind.

My point is that it is not predictable: there are many factors for language expansion, one of them always being the luck. Manpower is not always so crucial: not all languages spread via hostile takeovers. We know now that there was a population bottleneck even in the Yamnaya population, when the effective population size was very low.

[old europe]

now I see their hypothesis. Pay attention everybody. Basically what they are saying is that WHG/EHG on the Dneper Don and even full fledged EHG on the Volga were not PIE. For them PIE starts when the CHG/TTK kicks in to form the northern caucasus-lower Volga cline. So the cluster that makes PIE is the CHG/TKK
so what the  paper is all about is a revisited "southern arc hypothesis" For them PIE is born north of the caucasus but thanks to a southern caucasian influence.

They do not explicate where the language came from: they only talk about Proto-Indo-Anatolian after the admixture. But it is true that the more western river valleys can be excluded at this point. So the potential candidates for bringing in the Indo-European language are the Lower Volga EHG ancestry, the Central Asian ancestry, and the Caucasus HG ancestry (perhaps even the Levantine ancestry, concerning their hypothesis B).

The carrying capacity of a territory limits the manpower that can be mustered by the inhabitants to defend themselves from invasion or to invade foreign territories. - - However, if you asked me to consider the likelihood that any population of present-day Europe might expand their territory and spread their language over the entire (sub)continent in the future, the Kalmyks would not be the first people to come to mind.

My point is that it is not predictable: there are many factors for language expansion, one of them always being the luck. Manpower is not always so crucial: not all languages spread via hostile takeovers. We know now that there was a population bottleneck even in the Yamnaya population, when the effective population size was very low.

I'm on genetic forum since 7/8 years and as far as I remeber every debate by professional and amateurs alike on the PIE was all about uniparentals. That was because the uniparentals pointed in the direction of the steppe hypothesis. Now genetics has demonstrated that at least two PIE markers (R1a and I2a)  were present not in the open steppe but in the forest- steppe/ forest rich in WHG/EHG. The very same people decided that the methodology should be changed. Now it is all about the autosomal and the presence of R1a and I2a is for them completely irrelevant to the PIE debate.  
How can you play a fair football game when the adversary is both the referee and the player?

[Riverman]

The new Eneolithic era I11828/I31755 J2a (J-M319) sample from the Lower Don Krivyanskiy-9 grave 19 site is really interesting. According to the paper autosomal wise this sample is genetically quite close to Yamnaya although of course he pre-dates Yamnaya. It will be very interesting to see this sample's G25 results once the data comes out.

Also this sample is described as being similar to "Serednii Stih" samples in terms of burial pose, artifacts and date and his CHG like ancestry is apparently more similar to Mesolithic CHG compared to Aknashen.

Anyway interesting to see a J2a sample from the Eneolithic Steppe context.


I11828 6257 Krivyansky Russia T2a1b J-M319 J2a1a1a2b1b

That sample is very significant, because I speculated in that past that around the Lower Don, where we have a higher cultural development and see Southern inputs in the archaeological context, leading up to Sredny Stog, that there might be lineages involved in this fusion, which didn't make it due to later founder events after the takeover of the EGH clans.

That sample is the first evidence for this hypothesis being correct, that the diversity might be highest in the Lower Don region.

Those sites are relevant for the issue:
Krivyansky:Rostov Oblast, Lower Don group
Mariupol:Mariupol Neolithic Necropolis

Its a shame they got just one site from the Lower Don group from a too late time period, but at least one, better than none. Are there more samples from that site with yDNA assignment? I find the table not very neat.

Thus, while we can conclude that low
amounts of European farmer ancestry entered the UNHG population (from the western neighbors of the
NPR hunter-gatherers), it is possible that for at least some of them there was CHG-related ancestry as
well (from the east). Such ancestry was also detected in the GK1 subset at Golubaya Krinitsa Neolithic in
the Middle Don and at the Krivyansky Eneolithic (ref.3) in the Lower Don and may have thus extended
further west into the Dnipro region.

One of the R1b1a1b1a1a2c1a3a2 is from Mariupol (I27983).

Also very important is the "Proto-Yamna", so still no developed Yamnaya, from Bulgaria, which is R1a1a1 (R-M417)

I1456 DUR1 Durankulak, Kurgan
F, burial 15 (main burial)2

This debunks the idea of "R-Z2103 elites" only in the burials, and shows that R1a came form a different subset of only Yamna-related (!) people.

The developed Yamnaya are, like usual, nearly all R-Z2103.

[Archetype0ne]

Actually in their supplementery,they also perfectly model Anatolia BA using aknashen +Neolithic Anatolia without any steppe and they just say 'There are various possible competing hypothesis and all of them can be true.....'
Which is pretty strange,they maintain an ambiguous position although in the image they show progress in Anatolia,but in the supplementary they don't insist on that.

I think they model Anatolia BA with the BP group just to give their newfound samples more importance and credibility to their theories.

Let's see what they have in wait for future,they also mentioned in supplementary about sampling Neolithic/mesolithic central Asia and Siberia more.
Would be interesting to get these samples.

The thing is they don't just use one model. They ran multiple tournaments of various component for most populations. They provide the tables, the scores, and then based on the best fits try to make sense of it.

If you only focus in reading the last 40 pages, especially the first 20, before the last 20 I think they are not trying to fit anything by forcing it. As either way this is a departure from the Armenia being the cradle to proto-IE to BPgroup being the tracer die of IE, which would mean a back-migration of the Aknashen rather that a starting point... Don't know if I am making any sense lol.

[Jaska]

I'm on genetic forum since 7/8 years and as far as I remeber every debate by professional and amateurs alike on the PIE was all about uniparentals.

Debate always concerns things about which people disagree. So?

That was because the uniparentals pointed in the direction of the steppe hypothesis. Now genetics has demonstrated that at least two PIE markers (R1a and I2a) were present not in the open steppe but in the forest- steppe/ forest rich in WHG/EHG. The very same people decided that the methodology should be changed. Now it is all about the autosomal and the presence of R1a and I2a is for them completely irrelevant to the PIE debate.

Who are THEY and how do they relate to our discussion?
All I am saying is:
1. Language could be associated with any genetic level.
2. Language cannot be seen from the DNA or the genetic results – only linguistic results can tell when and where certain language was spoken.

[Archetype0ne]

It took me quite a few hours of reading and re-reading... and it might look ugly... but I hope this helps others (and that I did not make any mistakes).


Cline SShi <-> DonVolga
SShi = Don-DniproHG + BPgroup
BPgroup = KhlopkovBugor + CHG/PVgroup
DonVolga = BPgroup + PVgroup
DonDniproEN =  BPgroup + PVgroup +ASH
Remotnoye = BPgroup + ASH (44.6%)
A: CoreYamnaya = SShi + Remotnoye
B: CoreYamnaya = Aknashen + BPgroup + Igren_o (EHG?)


Now...

Deconstructing A:
CoreYamnaya = ((Don-DniproHG +  (KhlopkovBugor + CHG/PVgroup)) + (( KhlopkovBugor + CHG/PVgroup) + ASH)

Deconstructing B:
CoreYamnaya = Aknashen + (KhlopkovBugor + CHG/PVgroup) + Igren_o