Central and South Asian DNA Paper

I think that J2b2-L283 just happened to be more common in the Balkans today due to a founder effect in the Illyrian population. Otherwise J2b2-L283 is found pretty much all over Europe, even if at lower frequencies in the north.

I agree. There sure were founder effects after the arrival in illyria, yes. And L283 probably were in small numbers, yes.

But i was not claiming L283 came in big numbers, but rather that there is no proof they were just commoners. Or at least the chance that they were commoners is just the same as for any of the 3 other Hg's mentioned. Both M417, Z2103, L51 and L283 experienced founder effects after the dispersals from the steppe. And all of them can be found all throughout europe in residual amounts. I dont know much about M417 and L51, but both R1b-Z2103 and J2b2-L283 have been found in high status bronze age graves, and now also "high" status neolithic graves in Iran.

Maybe it was me who misunderstood your first post, but it just seemed like L51, Z2103 and M417 were presented as high status lineages, while all other steppe lineages (J2b-L283 and other minor R1b/R1a clades) were presented as 'commoners'.
Small numbers of J2b2-L283 in the steppe migrants does not mean that they were insignificant. Remember for a 1000 soldiers there is 1 king, not a 1000 kings for 1 soldier.

Anyways these are just minor details, and its just my opinion.
In the big strokes we pretty much agree on what happened; J2b2-L283 came from the east during the bronze age, and multiplied their numbers in the balkans during the ethnogenesis of the Illyrians, as well as after the ethnogenesis. While in the rest of Europe, the amount of J2b2-L283 stayed residual. (until roman expansionism began later of course. Which with the help of balkan soldiers would raise the level of J2b2-L283 once again in places like spain, britain and Italy.)
 
I was trying to reconcile some genetic evidence pointing to migrations from Steppe to southeast Europe to Anatolia, with the Anatolian languages not being in the same branch as Balkan languages. So maybe some early IE languages already existed there, straight from Armenia/Iran, then when the Steppe migration arrived, their relatively new IE form mixed with the old one, which would explain the archaic elements in there, and the seemingly Steppe ancestry. Seems a little complicated though.

Anatolian probably branched off - according to most glottochronological estimates I've seen - before 3,500 BC or even 4,000 BC, and the later southward waves of steppe or steppe-related migrations that impacted the Balkans took place much later, including some that went on as late as the Iron Ages (that was certainly the case of at least some Dacian and Illyrian tribes). If IE languages akin to Anatolian were spoken in the Balkans much earlier than 2,000 BC, they must've been swept away by Greeks, Phrygians, Thracians and many others along the subsequent centuries, or even earlier by languages related to the Copper Age/Bronze Age CHG expansion that possibly brought languages like Minoan. IMO the steppe admixture probably increased significantly only after these later Bronze Age/Iron Age IE migrations, especially those after the spread of Proto-Greeks.
 
Every body was expecting to find a mythical place north caucasus full of R1bs and actualy the oldest (m269) is found in northwestern Iran.

We don't have any ancient Y-DNA from the North Caucasus (e.g. Maykop) yet. Anyway I always said that R1b-M269 originated south of the Caucasus and crossed north between 6000 and 5000 BCE (most probably as R1b-L23 and some of its subclades).

Who on this forum expected R1b-M269 to have originated in the North Caucasus?
 
Sorry if it's a silly question - I don't know linguistics - but could there have been many IE speaking peoples, even before the Hittites? We don't have written records from them but they could have existed. Then of course there were huge migrations of an IE Steppe population in the Bronze Age, but maybe they weren't the only IE speaking people at the time.

The Hittites actually arrived quite late (1650 BCE), many centuries after Proto-Celtic R1b-P312 and Proto-Germanic R1b-U106 spread all over Western Europe and Proto-Indo-Iranian R1a-Z93 spread over western Siberia and Central Asia. It was about the time of the Indo-Aryan conquest of India and when R1b-U106 entered Scandinavia (Nordic Bronze Age). Their language was archaic, but as a people they appear quite late in the context of Indo-European migration. That's why they must have split early but remained isolated (like in the mountainous southern Balkans) for a long time before entering the historical record.
 
Anyway I always said that R1b-M269 originated south of the Caucasus and crossed north between 6000 and 5000 BCE (most probably as R1b-L23 and some of its subclades).
What tipped you off back then, that R1b-M269 came from south of the caucasus?
Personally i did not see it coming. But i love when samples like this comes around and challenges peoples beliefs and conceptions.
 
What tipped you off back then, that R1b came from south of the caucasus?
Personally i did not see it coming. But i love when samples like this comes around and challenges peoples beliefs.

That's because R1b-V88 is found all over the Levant and spread from there to Africa, and because the highest diversity of old R1b clades was reported to be in the Middle East. Both African R1b-V88 and Steppe R1b-M269 are/were cattle herders and cattle were domesticated halfway between the Levant (V88) and South Caucasus (M269), so my logic is that they must have a common origin there.

What I did not see coming was that there would be R1b-P297 and R1b-V88 in the Balkans (Iron Gates), Ukraine and Latvia in the Mesolithic. For me R1b was only in the Middle East in the Mesolithic and domesticated cattle around the modern border of Syria and Turkey, then moved to the South Caucasus, then crossed to the Steppe. I underestimated the propensity of Mesolithic HG to travel long distances and settled a bit everywhere. If the West African E1b1a in Mesolithic Iran in this paper isn't a mistake, that is another remarkable example.
 
There is a substantial amount of R1a, particularly in the northwest, and yet in the 41 samples from 1200 to 800 BC there is 10-15% steppe ancestry, but not a single R1a. Even in the Iron Age, there is only one R1a, and that's from 400 BC.

Either this is an unusual area, and R1a arrived via another route(?), or R1a was just a founder effect for some reason, or a lot of it did come with various SAKA groups.

It's like there's a disconnect between the conclusions and the data.

Also, the steppe MLBA percentages are pretty darn low in these samples. I mean, they vary, but a lot are in that 10-15% range.

I have to read the Supplement again, I guess. :)

Wouldn't Saka groups be partially East Asian admixed (and modern Indians are not, except for speakers of Munda languages in the east).
 
IMO the steppe admixture probably increased significantly only after these later Bronze Age/Iron Age IE migrations, especially those after the spread of Proto-Greeks.


something i was thinking about too. i think the study that looked at minoans and myceneans also tried to model myceneans as a mix of armenian populations and EEF and it would work pretty well. so the amount of "steppe" they had would not really be steppe in that case. maybe the same thing happened in italy.
actual steppe might come from later migrations south and slavery.

though when we speak about steppe admixture what exactly is it? EHG and CHG? as far as i know the exact propotions of CHG and EHG in the total amount of "steppe" is not clearly defined. wouldnt it be better to just look at EHG and CHG admixture seperatly to look at steppe and caaucasus movements since CHG might have come in two independent routes? or can we clearly distinguish what came from what direction?
 
That's because R1b-V88 is found all over the Levant and spread from there to Africa, and because the highest diversity of old R1b clades was reported to be in the Middle East. Both African R1b-V88 and Steppe R1b-M269 are/were cattle herders and cattle were domesticated halfway between the Levant (V88) and South Caucasus (M269), so my logic is that they must have a common origin there.

What I did not see coming was that there would be R1b-P297 and R1b-V88 in the Balkans (Iron Gates), Ukraine and Latvia in the Mesolithic. For me R1b was only in the Middle East in the Mesolithic and domesticated cattle around the modern border of Syria and Turkey, then moved to the South Caucasus, then crossed to the Steppe. I underestimated the propensity of Mesolithic HG to travel long distances and settled a bit everywhere.

Nice. Great prediction.

I was thinking maybe the mesolithic P297 and V88 in Iron gates and eastern europe, are a hint towards that the R1b population was in fact already rather big in the middle east by the time cattle were domesticated. As the supply of animals to hunt began to run low, some R1b hunters would migrate far(maybe via the black sea) to hunt somewhere else (Iron gates, latvia etc.), while some V88 and M269 stayed and domesticated those few animals that were left in that part of the middle east where they resided. So those domesticators did not need to move away like the rest, as they now had a reliable food source.
But i dont know, i am just speculating. They are rather peculiar those mesolithic samples of the Iron gates, Ukraine and Latvia.
Would be interesting to see ancient DNA from the mesolithic southern coasts of the black sea and northern anatolia, to see if that could have been the migration route of those hunters.

If the West African E1b1a in Mesolithic Iran in this paper isn't a mistake, that is another remarkable example.
Yes that sample will indeed be telling if it turns out its not a mistake. Let's hope they'll release the final paper soon.(or at least an updated spreadsheet)
 
The problems with craniometrics are that:

1) the variables are too limited (dolicho vs brachy, high vs low face) compared to DNA

2) the same skull shapes can be found in completely unrelated populations (e.g. dolichocephalic Nordics and Berbers)

3) cranial traits recombine when two different populations mix and can give rise to intermediate traits that are different from both source populations and closer to other unrelated populations. You can't extract percentages of admixtures from the original populations based on craniometics.

It's like mixing colours. If you mix one type of yellow with one type of blue, you get one type of green. With DNA you can see what the original colours were and what percentage of each is present in the mixture. With craniometrics all you know is that you have green, but don't know what the original colours were, nor how many were used and in what proportions. In the case of the colour green, you know that either you already had some green and could have altered it with adding another colour, or you could make it from scratch with blue and yellow. But if the colour is brown, it's much harder to know what was mixed originally.

I 'll answer this in an appropriate thread when I have time; just keep in mind serious cranial study implies metrics and tries to taxinomy; means are not sufficient, extremes and their interpretation are to be taken in account, and evolution in time when possible, of means but also of individuals - the shape is the result of very numerous dimensions, made sometimes by the only eye of the observator but depending on theses dimensions all the way - I consider it can help to capture brutal enough changes in a place peopling as a proof of historical moves, and as I said already it is more useful to separate pops than to settle their common origin - it's a complement, not the ideal tool, but sometimes it give us a more detailed sketche than does auDNA (anDNA research itself has different levels of accuracy and depth according to surveys, and can also give way to oversimplifications or pure errors)
 
Actually there is. The only admixture that clearly starts to appear and incrase whenever and wherever IE languages started to appear, except for West Asia (until now), is the Bronze Age steppe one. I'm perfecty fine with the hypothesis, increasingly supported by evidences, that the earliest form or PIE (maybe even still pre-PIE) was in West Asia, south of the Caucasus and near the Caspian and/or Black Sea, but that is a really long way from claiming that the steppe peoples had nothing to do nor were "necessary" at all for the spread of IE languages much, much later, when the Late PIE that expanded the most was certainly very different from that Earliest PIE if we consider that even the Anatolian branch, which is itself an evolution of that, and not a fossil that preserved that Early PIE intact, is already so distinct from the non-Anatolian IE branches that some once found it even hard to believe it was really IE and not a closely related language family. To talk about the spread of IE languages in the Bronze Age trying to link that phenomenon exclusively with Shulaveri-Shomu south of the Caucasus around 5,000 BC would be more or less like trying to establish how Portuguese and Spanish spread to South America relying just on the expansion of Italic tribes in Iron Age Italy.
The steppe component was probably NOT the source of PIE, but to say it was not "necessary" and had nothing to do with the spread of the vast majority of IE branches is simply unsubstantiated if you look at the ample evidences and the conclusions of virtually every studies based on evidences. People mix, change their genetics, absorb others, and the subsequent generations, even if totally transformed genetically, may keep the language and culture alive.
Of course, admixtures and Y-DNA haplogroups don't transmit languages, people do, so you and no one can expect to find a completely seamless and always intensive association of steppe admixture with the arrival of IE-speaking peoples, but there is undoubtedly a very strong correlation with the sole exception, until now, of Bronze Age Anatolia (as for Iberia, we have no reason to believe IE languages were there very early, all the IE languages documented there were very clearly not that divergent from Central European IE tongues to have been spoken there since the early Bronze Age).
There is no need for steppe to explain Pie in Anatolia, armenia ....that is what i meant.
But being at that, Pie (if shulaveri) went as far as balkans... What.would have been the impact of thiose?
What if bell beaker came from there and not steppe and spoke pie?
What if Myceanian was a Pie nothing to do with steppe?
 
Wouldn't Saka groups be partially East Asian admixed (and modern Indians are not, except for speakers of Munda languages in the east).

Perhaps twenty percent from only one group might have drifted out of the population.

The only other possibility I can see is that the West Siberian hunter-gatherers carried R1a as well.

Another question for investigation is which other route it might have taken.
 
The Hittites actually arrived quite late (1650 BCE), many centuries after Proto-Celtic R1b-P312 and Proto-Germanic R1b-U106 spread all over Western Europe and Proto-Indo-Iranian R1a-Z93 spread over western Siberia and Central Asia. It was about the time of the Indo-Aryan conquest of India and when R1b-U106 entered Scandinavia (Nordic Bronze Age). Their language was archaic, but as a people they appear quite late in the context of Indo-European migration. That's why they must have split early but remained isolated (like in the mountainous southern Balkans) for a long time before entering the historical record.

This story that Hittites arrived 1650 BC makes no sense, it's like saying Italics only arrived in Italy when Rome became an empire, you can see autosomally that Anatolia becomes Indoeuropean in the Chalcolithic.
GQ66BLE.png
 
Wouldn't Saka groups be partially East Asian admixed (and modern Indians are not, except for speakers of Munda languages in the east).

yes, that is what they say in the paper
their conclusion is that the bulk of steppe admixture arrived in India with steppe MLBA
 
Perhaps twenty percent from only one group might have drifted out of the population.

The only other possibility I can see is that the West Siberian hunter-gatherers carried R1a as well.

Another question for investigation is which other route it might have taken.

there was R1a in Kitoi, but IMO it was R1a-YP1272, the combed ware branch which is supposed to originate in that area
that makes sense too, the Kitoi for sure didn't speak IE
 
I agree with everything, except perhaps with the Hittites moving early from the south west Caspian into Anatolia. The Hittites were close relatives of the Luwians (from Troy), Carians, Lydians and Lycians, who were all living in western Anatolia. The Lycians are said to be originally from Crete. That's why I think that an early migration from the Steppe to the Balkans is more likely. Furthermore, the R1b-L51 branch (Centum or Proto-Italo-Celto-Germanic) only moved to central Europe between 3000 and 2500 BCE, but Steppe people started invading the Carpathians and Balkans ("Old Europe") from 4200 BCE. What happened to all those Steppe people who settled in SE Europe for over one thousand years? I believe that they were the Anatolian branch and that later Steppe invasions pushed them to western Anatolia (and probably also Greece and Albania).

OK, here we agree.
 
This story that Hittites arrived 1650 BC makes no sense, it's like saying Italics only arrived in Italy when Rome became an empire, you can see autosomally that Anatolia becomes Indoeuropean in the Chalcolithic.

that's right, but on the other hand, before the Hittites conquered the land of the Hatti, ca 1900 BC, they were just a minor tribe, other non-IE languages were spoken in that area
 
There is no movement towards the Anatolia on David Reich's whiteboard. Arrow from Central Anatolia (Node A) goes probably to Northern Iran (Hajji Firuz) and from there to north, Caucasus and then to Ukraine. If this is not only an assumption and he has Hittite data, Central Anatolia may be the PIE (Indo-Hittite) home. I can't post the link now but you can look at the whiteboard from the NYT article.
 
Nick Patterson interjects:

"Nick Patterson (Broad) said...@Lenny Dykstra:

I am extremely grateful to my South Asian collaborators
(and this goes for the whole Reich lab.); they are honest,
excellent scientists. There's a simple reason we have not
published on DNA from the IVC proper -- we have no useful DNA, and as you might guess this is not from lack of trying. "

Well, that's that.
 
There is no need for steppe to explain Pie in Anatolia, armenia ....that is what i meant.
But being at that, Pie (if shulaveri) went as far as balkans... What.would have been the impact of thiose?
What if bell beaker came from there and not steppe and spoke pie?
What if Myceanian was a Pie nothing to do with steppe?

There is not a shred of evidence that Indo-Europeans were in Anatolia or Armenia before after 2000BC (for Hittite and Mitanni), and long after that for Armenian and Iranian. And there is mountains of evidence to the contrary for transmission from the steppe.

Imagine if we found some particular G2a clade responsible for farmer expansion into the Balkans in 10000BC Iberia. Would we then have to say that Anatolian farmers actually came from Iberia? This is what you're trying to do with this 5500BC Iranian R1b-M269 guy.
 

This thread has been viewed 286574 times.

Back
Top