search this blog

Tuesday, July 4, 2017

Out-of-India chickens coming home to roost


Razib has posted a spacious but none-too-technical review of the ongoing Aryan Invasion Theory (AIT) controversy, along with some personal anecdotes and predictions about how ancient DNA from South Asia might shape the debate in the near future (see here).

It should be a useful guide to the topic for those of you who aren't quite as excited about reading about my latest adventures with qpGraph as many of the regulars in the comments here.

One thing that I'd perhaps add to Razib's post is that the ancient DNA record now boasts Late Neolithic Yamnaya-like Corded Ware Culture individuals from the East Baltic region that belong to Y-haplogroup R1a-Z645. And that's usually as far as their lineages go (see here).

This is important, because the Z645 mutation is directly and recently ancestral to the pair of likely post-Neolithic mutations that define the two R1a subclades most common in Europe and South Asia today: Z282 and Z93, respectively.

So not only are the "European" R1a-Z282 and "South Asian" R1a-Z93 relatively young sister clades, but their ancestral clade has now been found in ancient samples from Northeastern Europe that probably predate their appearance by only a few generations, if that. Of course, the upshot of all of this is that R1a-Z93 could not have originated very far from the East Baltic, which makes South Asia look about as likely as the homeland of this subclade as the goddamn moon. Conversely, it makes AIT look very plausible indeed.

However, granted, this might seem very confusing to anyone who hasn't been studying the R1a topology for years, and perhaps better left out of the more mainstream debates on AIT for the sake of simplicity. By the way, I found this part of Razib's post especially intriguing:

One scientist who holds to the position that most South Asian ancestry dates to the Pleistocene argued to me that we don’t know if ancient Indian samples from the northwest won’t share even more ancestry than the Iranian Neolithic and Pontic steppe samples. In other words, ANI was part of some genetic continuum that extended to the west and north. This is possible, but I do not find it plausible.

I suspect that this scientist's rather fanciful suggestion (which really flies in the face of very solid models based on ancient genomic data from Europe and surrounds) is a hint of the direction that the debate will take right after the publication of ancient genomes from South Asia. Because when that happens, obfuscators like this guy (usually hopeless Out-of-India proponents) will either have to concede defeat and quit the debate, or ramp up their obfuscations to spectacular new highs.

And please don't mistake my confidence on this issue for bluff and bluster. It's not exactly the best kept secret out there that ancient samples from India and Pakistan are now ready, and...oops I probably can't say more than that for now. Pity.

See also...

Ancient herders from the Pontic-Caspian steppe crashed into India: no ifs or buts

Indian confirmation bias

179 comments:

Davidski said...

Great exchange here between a typical academic obfuscator and an Indian journalist.

“Archaeology has been indeterminate on the Bronze Age migration question, but there is no archaeological discovery that will militate against genetics joining linguistics in favouring migration as the right explanation.”

http://www.thehindu.com/opinion/op-ed/the-problematics-of-genetics-and-the-aryan-issue/article19165320.ece

Palacista said...

What are the odds of a two stage population turn over in the IVC area, one about 2500 BCE and the second about 1500? The first Iranian, the scond steppe.

Davidski said...

Dunno, but the odds of Yamnaya being native to both Eastern Europe and North India are zero.

Grey said...

i wonder if a similar thing happened twice i.e.

- very early in the farmer expansion some high density wetlands HGs around Black Sea adopted some of the encroaching farmer package

- much later in time some high density wetlands HGs around Baltic / North Sea adopted some of the LBK farmer package

Ryan said...

@Pal - "What are the odds of a two stage population turn over in the IVC area, one about 2500 BCE and the second about 1500? The first Iranian, the scond steppe."

Uh. What do you mean by Iranian? Like, linguistically? Because that would also be steppe.

Palacista said...

Using Iranian as shorthand for the modern area of Iran, which in 2500 BCE was most certainly non steppe.

Ryan said...

Ah, so Elamite / Dravidian or whatever. Isn't 2,500 BCE a bit late for that? I'd think 3,500-4,000 BCE would be a better candidate for large demographic change - just before the rise of the Indus Valley Civilization rather than at its peak.

DDeden said...

Stupid question: Did "chicken" derive from chick-hen or so?

JohnP said...

People are putting too much faith on Harappans having some significant/big ASI "bright green" admixture like 60/40 or 50/50. I think them much more likely to be a Iran_LN population at their earliest and slowly acquiring ASI admixture as millennia passes until their collapse, having 60/70 Iran_LN and 40/30 ASI at this time.
But I could be wrong too, whatever. At least no R1a will be found.

AJ said...

No point in talking about that when there is no aDNA from South Asia and Central Asia from various periods - Paleolithic, Mesolithic, early Neolithic. Worry about your steppe theory for now.
______ ______ _______ ________

As for your dead steppe theory : Matieson & Reich et al 2017 points out in recent paper, no steppe migration to Bronze Age Anatolia.

The Genomic History Of Southeastern Europe Mathieson et al. 2017

"No evidence of Copper Age Balkans-to Anatolia migration.

One version of the Steppe Hyothesis of Indo-European language origins suggests that Proto-Indo-European langages developed in the steppe north of the Black and Caspian seas, and that the earliest known diverging branch -Anatolian- was spread into Asia Minor by movements of steppe peoples through the Balkan peninsula during the Copper Age aroung 4000 BCE, as part of the same incursions from the steppe that coincided with the decline of the tell settlements.

If this were correct, then one way to detect evidence of it would be the appearance of large amounts of the characteristic steppe ancestry first in the Balkan Peninsula, and then in Anatolia - However, our genetic data do not support this scenario.

While we find steppe ancestry in Balkan Copper Age and Bronze Age individuals, this ancestry is sporadic across individuals in the Copper Age, and at low levels in the Bronze Age.

Moreover, while Bronze Age Anatolian individuals have CHG/Iran Neolithic related ancestry, they have neither the EHG ancestry characteristic of all the steppe populations sampled to date, nor the WHG ancestry that is ubiquitous in southeastern Europe in the Neolithic. This pattern is consistent with the seen in northwestern Anatolia and later in Copper Age Anatolia, suggesting continuing migration into Anatolia from the East rather then from Europe."

Basically : They found no steppe admixture in Bronze Age Anatolia but B.A Anatolians had CHG and Iran Neolithic admixture.
_______ _______ ________ _________ _______

Most likely/possible trail currently:

Hittite migration to Anatolia <---- CHG/Maypok/Caucasus culture---> language shift in steppes

"The foreign objects in the North Caucasus reveal no connection to the upper reaches of the Euphrates and Tigris or to the floodplains of Mesopotamia, but rather seem to have ties to the Iranian plateau and to South Central Asia."

There is no "Yamnaya culture" archaeological trail to South Asia, even if steppe admixture in S.A prefers Yamnaya over other steppe groups. Material cultures of Neolithic and Bronze Age South Asia shows more similarities with Maykop/Caucasus cultures/neighboring BMAC than anything else.

Davidski said...

Those Bronze Age Anatolians are most likely Hattians, but whatever they are, they can't be linked to any plausible Indo-European culture: too early and wrong place.

So why would they have steppe ancestry if they weren't Indo-Europeans?

Wouldn't this throw into doubt the association between steppe ancestry and Indo-European languages?

Davidski said...

And I didn't bother to reply to any of your comments about South Asia because you're rambling nonsense.

The fact that you're ignoring the currently available ancient DNA evidence from Eastern Europe plus the modern DNA, including Y-chromosome, evidence from South Asia speaks volumes about your motivations and mental ability.

Rob said...

I had no idea Hattians lived in southwest Anatolia

Davidski said...

These Anatolians aren't from extreme southwest Anatolia. Their site is mapped as part of Hattian territory on some maps and the date fits better than for Hittites and Luwians.

Roy King said...

Those Bronze Age samples are definitely not in Hattian territory which is encircled by the Halys River. They are likely part of the Western Anatolia substrate population that is neither Hattian or Hittite or Luvian. Later the area was inhabited by Luvians. They could be the pre-Greek substrate language speakers or Leleges.

Ryan said...

@JohnP - "People are putting too much faith on Harappans having some significant/big ASI "bright green" admixture like 60/40 or 50/50."

Sundadonts early in Mehrgarh suggest some sort of ASI affinity there IMHO. I'm not sure if Y-haplogroup L suggests West Eurasian or East Eurasian (ie ASI) affinities though.

Rob said...

@ "Those Bronze Age samples are definitely not in Hattian territory which is encircled by the Halys River. They are likely part of the Western Anatolia substrate population that is neither Hattian or Hittite or Luvian. Later the area was inhabited by Luvians. They could be the pre-Greek substrate language speakers or Leleges."

That sounds rather speculative (and somewhat contrived to fit a predetermined dogma), especially if the Leleges were thought to be "autochthonic Aegeans", why would they be CHG/ Iran -rich ?

Ryan said...

@Rob - "That sounds rather speculative (and somewhat contrived to fit a predetermined dogma), especially if the Leleges were thought to be "autochthonic Aegeans", why would they be CHG/ Iran -rich ?"

Aren't the earliest Anatolian Neolithic samples already part CHG? Maybe there was some structure within Anatolia.

Rob said...

@ Ryan
"Aren't the earliest Anatolian Neolithic samples already part CHG? Maybe there was some structure within Anatolia."

One or two of the Tepecik individuals look like they had some CHG, not too surprising because it is further east than Mentese or Barcin; but the change in the Bronze Age seems pronounced. Certainly, the genetic results hinted in the Mathieson paper correlates with archaeologcal and palaeoecological picture.

Davidski said...

CHG-like ancestry was in Anatolia at variable levels during the Neolithic. So whatever was bringing it to Anatolia at that time probably just brought more of it during the Bronze Age as populations grew and expanded. Nothing special.

Davidski said...

Plus the CHG brought with it some ANE (but no Yamnaya-like admixture!)

CHG admixture in early western Anatolian farmers

http://eurogenes.blogspot.com.au/2016/02/chg-admixture-in-early-western.html

AJ said...

@Davidski

"The fact that you're ignoring the currently available ancient DNA evidence from Eastern Europe plus the modern DNA, including Y-chromosome, evidence from South Asia speaks volumes about your motivations and mental ability."

Hilarious. Obviously, there are no regions past Eastern Europe worth testing for various Paleolithic and Mesolithic aDNA to test this theory, good thing geneticist dont think like you.

Given that steppe admixture in South Asians prefers Yamnaya over Sintasta/Andronova, that itself shows steppep admixture in S.A is proxy for something Mesolithic in Central Asia and let's not forget that your Mesolithic EHG is admixed 75% ANE coming from Siberia or Central Asia.

"Those Bronze Age Anatolians are most likely Hattians, but whatever they are, they can't be linked to any plausible Indo-European culture"

They picked Bronze Age samples for a reason. Hittites came through either from Balkans (unlikely now) or Caucasus (most likely now). PIE is still far from settled but there is no steppe admixture in Bronze Age Anatolia for now, period.

Ryan said...

@AJ - "Given that steppe admixture in South Asians prefers Yamnaya over Sintasta/Andronova, that itself shows steppep admixture in S.A is proxy for something Mesolithic in Central Asia"

No it does not.

AJ said...

@Ryan

You should take that to Lazaridis et al. and Reich, not me.

Rob said...

@ Dave

Not quite. As I said, some early Anatolian farmers had minor CHG, quantifying at ~ 10-15%, and seen in your PCA as lying slightly shifted in your PCA.
The Bronze Age individuals, on the other hand, a massively shifted.
Claiming this is due to internal expansion isn't going to hold much water.

Davidski said...

@AJ

Yamnaya-related admixture cannot be a proxy for Mesolithic ancestry in South Asians, because the Yamnaya genotype only formed during the Eneolithic/Bronze Age in Eastern Europe.

Formal models certainly don't allow room for what you're suggesting. In other words, South Asians have Bronze Age Eastern European ancestry, and both Lazaridis and Reich have argued this.

And the Y-chromosome link between Eastern Europe and South Asia clearly also dates to the Bronze Age. I explained this in my blog post.

Are you trolling or mentally deranged?

Davidski said...

@Rob

Claiming this is due to internal expansion isn't going to hold much water.

It actually makes perfect sense.

The Chalcolithic and Bronze Age Anatolians cluster pretty close to the most CHG-shifted Neolithic Anatolian.

All that happened was a homogenization of genetic components within Anatolia, from west to east.

Rob said...

No it doesn't. It fails basic mathematics.
The CHG admixed Neolithic farmers from Anatolia were a minority, and then only ~ 10% level. If you homogenize them with the rest of the early Antolian farmers (who had even less CHG), the overall result is less CHG, not a marked shift of ~ 20-30% more.

The only feasible explanation is frank migration from the east.

The "Chalcolithic" Anatolian should not be groups with the early ANF, she came from 3600 BC, is clearly a divergent population which came after a 700 year hiatus at the site. She was the earlier version of BA Anatolians.

AJ said...

@Davidski

"Lazaridis and Reich have argued this."

Right...as Lazaridis says "useful direction of future research is a more comprehensive sampling of ancient DNA from steppe populations, as well as populations of central Asia (east of Iran and south of the steppe), which may reveal more proximate sources of the ANI than the ones considered here, and of South Asia to determine the trajectory of population change in the area directly."

I'll take their word for it. Without Paleolithic, Mesolithic and early Neolithic aDNA from Central Asia and South Asia, there is no point in arguing about this with you.

"And the Y-chromosome link between Eastern Europe and South Asia clearly also dates to the Bronze Age. I explained this in my blog post."

Bahaha

That link is most likely Central Asia. In other words, Eastern Europe <--- Central Asia ---> South Asia. All we need is Paleolithic & Mesolithic aDNA from the region.

Silva 2017 study : "R1a in South Asia most likely spread from a single Central Asian source pool, there do seem to be at least three and probably more R1a founder clades within the Subcontinent, consistent with multiple waves of arrival."

Davidski said...

@Rob

The only feasible explanation is frank migration from the east.

Yes, but initially very gradual, starting during the Neolithic.

Davidski said...

@AJ

The Silva paper says nothing about any migrations from Central Asia to Europe. It just points out that South Asian R1a derives from Central Asia, because it obviously does if it came from the steppe.

Of course Eastern European steppe populations didn't enter South Asia directly. They obviously had to cross Central Asia in order to do that.

Some of them, like the ancestors of Afanasievo and Andronovo, even settled in Central Asia. It's these kinds of Yamnaya-related Central Asian groups that moved into South Asia, you stupid troll.

AJ said...

@Davidski

No point in arguing about this without Paleolithic, Mesolithic and early Neolithic aDNA from Central Asia and South Asia.

Davidski said...

@AJ

No point in arguing about this without Paleolithic, Mesolithic and early Neolithic aDNA from Central Asia and South Asia.

That's because you're unable to make basic inferences from available data. You don't even understand when these papers are spelling out to you what these obvious inferences should be.

How the hell is Paleolithic, Mesolithic and early Neolithic aDNA from Central Asia and South Asia going to change the fact that Mesolithic Eastern Europeans carry R1a, and that the ancestral clade to South Asian R1a was sitting in the East Baltic and surrounds during the Bronze Age?

Do you even know how to read a map? Look, there were migrations from Eastern Europe to Central Asia during the Bronze Age (the big fucking arrows pointing right).

https://drive.google.com/file/d/0B9o3EYTdM8lQdzRlWXBQaXlnMGc/view?usp=sharing

Central Asia and South Asia were sinks not sources of ancestry at this time.

Ric Hern said...

The South Asia Mesolithic origin of R1a theory is rather confusing. If R1a originated there who precisely were responsible for the introduction of Indo-European into India ?

How can Sanskrit preserve so many similarities to Baltic and Slavic if their Common Ancestor split up during the Early Mesolithic and they would have been separated from each other for more than 10 000 years ? Why 10 000 years ? Because R1a was already found in the Ukraine at +10 000 years ago.

So far there is no evidence of R1a and R1b in Neolithic Anatolia and Iran that could point to a migration from India during this period.

So does all hopes of a Western and Southern Asia origin of PIE hang on the Maykop Culture ?

postneo said...

@Davidski
#Wouldn't this throw into doubt the association between steppe ancestry and Indo-European languages?"

Actually it won't throw it in to doubt especially when its not proven in the first place. Its already under doubt as are all other hypotheses.

"How the hell is Paleolithic, Mesolithic and early Neolithic aDNA from Central Asia and South Asia going to change the fact that Mesolithic Eastern Europeans carry R1a"

Yes ..so you froth..Mesolithic R1a is found in Siberia as well, northeast of the Altai Far East of afnasievo even, its not confined

The supposed LBA expansion of Andronovo in to Central Asia is looking weak. With more sampling Fedorovo cultures are turning out with dates as early as Andronovo.


As for Reich , Lazardis they have slightly better insight than most of us, but at the end of the day they are still groping in the dark. Mainly because such little sampling has been done.




Ric Hern said...

@ postneo

Where does Central Asia begin and end ? According to some maps, Southern Siberia, Afghanistan and Pakistan is not included in Central Asia....And even if we are talking about Central Asia, there is a huge ecological difference between Northern and Southern Central Asia.

I personally think it is a grave mistake not to specify more accurately precisely where within a broad area we talk about. Central Asia can mean anything from Scorching Hot Deserts to Snow Covered Expanses which if interpreted only as being the One or the Other could lead to misleading assumptions.....

Rob said...

Wasn't there some tweet after the Meso-Neo Altai genomes talk that they found a better/ more recent proxy for "ANE" than AG/ mal'ta?

That this comes on the back of some recent excavations (noted in Russian news) from Kyrgyzstan is perhaps noteworthy.

Anthro Survey said...

@Rob

It seems the CHG-zation of Anatolia during the copper age coincided with extensive CHG-zation of the rest of the Near East during that time. We've also seen this in Lebanon, Jordan, and ancient Egypt so far. This begs the question as to whether the westward expansion of the eastern basal-rich lineage can be connected to the spread of metalworking. Kind of like demic diffusion corresponded w/adoption of agriculture in the Neolithic.

postneo said...

@Ric hern
Central Asia is not a useful term. It's bandied about as a way of saying we don't know anything.

Most of Central Asian pop density is in the bounding mountain regions.

The spatial separation of z645, 93, 282 Is profound whereas temporally they are almost contemporaries.

It's an anomalous situation. Almost as if they were cut off. It probably correlates with some kind of geographical feature that induced this.

Rob said...

@ Anthro.

My feeling is that there were several movements in that period occurring in the Near East/ Caucasus / Anatolia, all of which to varying extents would involve 'CHG', but the details are not entire clear as yet.
I also don't think it can all be correlated in any neat way with any particular technique or craft, for which there were multiple lines, even a couple of Corded Ware type axes have been seen in the Levant.
But they might all indeed share the feature of being more pastoral based and metalurgically inclined than earlier, Neolithic groups in Anatolia & the Levant.

Davidski said...

There weren't any tweets about better ANE proxies than AG3. The tweets were about more extreme East Asian Siberians than present ones.

And metallurgy is not a Proto-Indo-European thing. It was introduced into Indo-European societies by foreigners.

postneo said...

"The South Asia Mesolithic origin of R1a theory is rather confusing. If R1a originated there who precisely were responsible for the introduction of Indo-European into India ? "

There is no point claiming a Mesolithic origin of r1a anywhere. So far they have found it in vastly separated spaces.

Also IE was never confined by y clade. Why make such restrictive assumptions when no one really has a clue. sarmatians did not have much R for example.

Davidski said...

@postneo

Quit playing dumb and finally take a good look at all the ancient DNA from Europe, it includes Z645, Z93 and Z283.

postneo said...

Conclusion without sampling is obfuscation/censorship

z93 from Europe even the ancient ...however close to Tmrca etc is still statistically a straggler

Xiaohe at 2000 bc has z93 minus. Neither ancients nor mordern populations have been sampled adequately outside Europe.

The few ancients sampled don't have much r1a in corresponding moderns so it's not the right place.

Davidski said...

Eastern Europe has R1a to Z93 from 8,000 BC to 2500 BC.

So it's the right place. And there's nothing special about R1a elsewhere.

Now go and find another hobby.

Ric Hern said...

@ postneo

If R1a and R1b were not responsible for the spread of Indo-European, then who were ? Which other Haplogroups had such a large impact everywhere where Indo-European are spoken today ?

Ric Hern said...

Just like the Basque debate...You see +-1,5 Million Basque Males with R1b in a relative small area making the percentages look very impressive,and many hammer on this.

However we forget that there are almost 2,7 Million Irish, 19 Million French, 15 Million Spanish and 20 Million Germans that are R1b Men.

So how does this make Basque a proxy of how and from where R1b migrated or originated ? Using outliers like Sarmatians etc. confuse people and pull the picture skew unnecessarily.

Grey said...

@Ric Hern

i may have remembered this wrong but IIRC one of the things about the Basques is they have the EHG part of the EHG+CHG "steppe" component but not the CHG part - which implies they got that component before the EHG+CHG mix or from a population which never had that mix. I think that would be a significant clue (if correct).

Ric Hern said...

Yes I understand what you say, however 99% of Western European R1b Men speak Indo-European Languages so why would 1% be significant in determining which language the original R1b Men spoke ?

Alberto said...

@Ric

This has also been explained many times. Basques are an exception today. But they were the norm 2500 years ago. In Iberia, most of the population before the Roman conquest spoke non-Indo-European languages. And it's reasonable to think that they were R1b heavy.

Etruscans spoke non-Indo-European, and I'd bet that when we get DNA from them they're going to have as much R1b as Romans, if not more.

Western Europe speaks uniformly IE only recently. Much, much later than the arrival of R1b (Celtic is in all likelihood an Iron Age expansion, and we have no idea of what was spoken before in France or Britain). So for now, there's no strong connection between R1b and IE languages in Western Europe. Maybe more aDNA will say otherwise, but current data doesn't support R1b=IE in any special way.

Ryan said...

@AJ

Claiming that Reich and Lazaradis back up your argument is nonsense. You're making baseless assertions here.

Dmytro said...

The connection between genetic marker and language is absolutely fortuitous, even where

Dmytro said...

(Continuation I pad glitch...)massive. The specific ydna determines nothing. It is an interesting historical accompaniment of processes stemming from altogether different factors. Unless one is ideologically disturbed...

Ryan said...

Clearly Chadic languages are the original R1b language. :3

Samuel Andrews said...

@Alberto,
"Basques are an exception today. But they were the norm 2500 years ago. In Iberia, most of the population before the Roman conquest spoke non-Indo-European languages."

Didn't most of Iberia speak Celtic and other IE languages?

"So for now, there's no strong connection between R1b and IE languages in Western Europe"

R1b links Yamnaya, Bell Beaker, and IA Celtic speakers.

Anthro Survey said...

@Alberto

You do realize that Basques(and Catalans to a large extent; plus, many of them are former "Basques") are a bottlenecked/drifted population, right? Besides, they're not exactly the most steppic population in Europe.

The argument is that IE languages spread concurrently with steppic ancestry. In Western Europe, R1b was co-desseminated with this. You're missing the forest for the trees.

Anthro Survey said...

@Alberto

Also, I don't think we can say non-IE speaking populations were the norm across Western Europe as a whole---mainly just in Iberia and Sardinia. North Italy was almost entirely IE-zed. Recall that Etruscans underwent a period of expansion to attain their territory of greatest extent. Whose territories did they expand onto? Onto those of Villanovan-derived peoples and/or Umbrians. Besides, we don't have a good reason to think that Etruscan replaced IE outside of the core territories. Sections of Tuscany probably continued to use Ligurian and Umbrian in daily affairs, as was probably the case with Etrurian Padania and whatever additional IE langs. were spoken there.

I have a theory that North Italy received its full steppic package at a much earlier date than Iberia and that this has a lot to do with the persistence of non-IE there. Recently had a discussion with Rob about this which you will find most pertinent. See comments here:
http://eurogenes.blogspot.com/2017/06/matters-of-geography.html?m=1

Rob said...

The Basques are neither drifted nor bottlenecked. They're a group which formed only recently, not in the Stone Age

Anthro Survey said...

@Rob

I know they didn't form in the Neolithic---they have substantial steppe admix, after all. What I mean is that with respect to R1b reaching 90% there, their formation resulted from a demographic expansion of a subset of (Vasconic?) peoples living there. As for whether this was BC or early AD, I'm not sure.

Rob said...

Yes I agree- and it seems to mostly the DF27 variety
So it's completely established that R1b is a recent founder effect in that region. But I think Albertos point is we don't quite know what language the BB network spoke. Western indo-European certainly a leading hypothesis but really we have no attesstation to ground this on, and the earliest evidence in Western Europe comes thousands of years later when a lot of things could've happened subsequent to that, That is why the ADNA form Greece and India are gonna be very important , But also more aDNA from the latter Bronze Age and Iron Age from all parts of Europe

Matt said...

@ Davidski, would you mind running this qpGraph if possible: https://pastebin.com/k7nDg1TU?

Ryan said...

@Sam - "Didn't most of Iberia speak Celtic and other IE languages?"

It was more or less half and half as of the Romans getting there. I'm not sure where you're getting "most" from. Tartessian, Turdetanian, Iberian and Aquitanian are not IE languages. And R1b predates any IE language in Iberia - going back to the early Neolithic. I don't suppose you're going to suggest Chadic languages are IE too? I'm even sure of the argument you are making here - that a Basque elite assimilated a bunch of Celts maybe?

"R1b links Yamnaya, Bell Beaker, and IA Celtic speakers."

And Celtic speakers have a strong and noticeable non-IE substrate, as do German speakers. I'm R1b-M222. I speak English and French. Go back 500 years and my paternal ancestors spoke neither of those. Languages can be learned. Linguistic shifts happen.

R1b links Yamnaya, Bell Beaker and IE Celtic speakers, but it also links Basques, Berbers, Uralics, Turks and many other non-IE groups. And the link isn't as direct as you might think.

The first Bell Beakers did have R1b, but they lacked steppe admixture of any sort, and none of the Yamnaya samples were L51 - ie the samples from Yamnaya do seem to be associated with IE R1b clades, but not the clades that dominate western Europe. Maybe those samples are yet to be found. Either way, I don't think there's any plausible way to suggest that the first Bell Beakers spoke IE or that they were linked to Yamnaya except perhaps sharing ancestry from common non-steppe sources.

We know that once the Bell Beakers reached central Europe they mixed extensively with IE speakers - particularly CWC folks with mostly R1a Y-DNA, but Bell Beakers existed well before this. I think the most reasonable expectation is that since the later Bell Beakers showed cultural continuity with the earlier ones, that we should consider later Bell Beaker as likely not speaking IE either.

The question then is how the later Bell Beakers became so overwhelmingly R1b. Either they assimilated some unsampled R1b elites from CWC, or it descended from earlier Bell Beakers, but either way it doesn't seem likely that Bell Beakers showed such strong continuity with non-IE predecessors without keeping that non-IE language. If "western" R1b originated in Bell Beakers and not the steppe though, that would make much of the R1b in Yamnaya intrusive and recently assimilated from the west (which I personally think is likely but I realize most here would disagree).

Ryan said...

@Rob - We know at least the Iberian BBs did not speak IE. I wouldn't call that an open question at all.

Chad Rohlfsen said...

Ryan,

Iberian Beakers did not L51. There is no demic movement from Iberia. You should read the papers.

Davidski said...

@Matt

https://drive.google.com/file/d/0B8XSV9HEoqpFNHlaODVrNi1CU2c/view?usp=sharing

Samuel Andrews said...

@Ryan,
"And R1b predates any IE language in Iberia - going back to the early Neolithic."

R1b P312 seems to be linked to IE languages, not R1b V88 or other forms of R1b1a that have existed in Iberia since maybe the Paleolithic.

Samuel Andrews said...

@Ryan,
"Go back 500 years and my paternal ancestors spoke neither of those. Languages can be learned. Linguistic shifts happen."

You and all of us here are linguistic exceptions. Just about all of us can speak English because of recent British empires. The spread of English, French, and Spanish outside of their homeland can't be used as examples of how Celtic languages spread.

Excluding Latin, the distribution of IE languages in 1500 AD was probably the result of organic, folk migration.

Modern English really do have Anglo Saxon ancestry, modern Russians really do have proto-Slavic ancestry, modern Icelanders really do have Medieval Scandinavian ancestry. Why would Iron age Celtic speakers have been any different?

Language directly correlates with genetics because we learn how to speak our 1st language from our family and community(who are probably our relatives). Yes, language doesn't perfectly correlate with genetics but there's almost always some type of correlation.

Samuel Andrews said...

@Ryan,
"R1b links Yamnaya, Bell Beaker and IE Celtic speakers, but it also links Basques, Berbers, Uralics, Turks and many other non-IE groups. "

I was referring to R1b L23. It's a very specific form of R1b that both Northern Bell Beaker and Eastern Yamnaya belonged to at frequencies of 90%+. It's similar to how Andronovo and Corded Ware belonged to R1a Z645 at 90%+.

Rob said...

I'd like to see some more CWC haplogorups from Netherlands and Jutland

Arza said...

@ Samuel Andrews

Are you aware that in Lech Valley a double burial of a woman with a CWC-level Yamnaya ancestry (E09537_d) and a farmer man with Y-haplo G2a2a1a2a1a (E09538) was found?
What language their children have spoke?

https://3.bp.blogspot.com/-k1DcF3k75hw/WROgZOb6rAI/AAAAAAAAAIo/U7ntqrJ3AdYbbHUHlHVqyxLutH2O9bQogCLcB/s1600/LechLady.png

(He's just below her)

Don't you find interesting that she is the second most Yamna-like sample among Bell Beakers after I2787 who was most likely just a random migrant from the steppes to the Balkans, just like his older colleagues from the SE paper?

BTW Anyone have any news on Lech Valley paper?

Grey said...

@Ric Hern

"so why would 1% be significant in determining which language the original R1b Men spoke ?"

i don't think it does. i think it implies there's at least two separate stories.

Samuel Andrews said...

Who here doesn't speak the first language of their parents? The fact is families/communities pass on language more than anything else.

Yes there are many exceptions but exceptions are never the rule.

Samuel Andrews said...

Another thing, language is one of the things that distinguishes ethnicites from each other. What makes Spanish people Spanish? They speak Spanish.

And ethnic groups create arenas for breeding and barriers from breeding with people who speak different languages.

To say language and genes are unrelated things is crazy.

Chad Rohlfsen said...

I've got a pic of the Iron-Age Finns presentation by the author. If anyone wants to see it, I can post it.

Arza said...

Who here doesn't speak the first language of their parents?
Probably someone whose parents are of different origin and he/she speaks the language of one of the parents.

To say language and genes are unrelated things is crazy.
So don't say this.

Yes there are many exceptions but exceptions are never the rule.
What is rule now, then was an exception.

Strontium and oxygen isotope ratios disclosed more than half of the females to be non-local, while there were only single occurrences among the male and subadult individuals. This striking pattern of patrilocality and female exogamy prevailed between about 2500 and 1700 BC. It was independent of individual sites and their archaeological assignments to the Bell Beaker Complex or the Early Bronze Age. While the males ensured settlement continuity in a spatially limited area, the results indicate that the females were driving forces for regional and supra-regional communication and exchange at the dawn of the European Metal Ages.

Arza said...

@ Chad Rohlfsen
I kindly ask and thank you in advance!

Samuel Andrews said...

@Chad,

Pretty please.

Ric Hern said...

@ Grey

Yes exceptions isn't the rule. That is my actual point. Some people want to make too much of exceptions and therefore confuse themselves and others.

And some need to read what people wrote and especially the genetic papers posted, to form a better picture of what was actually said and reply accordingly.

Example: I was talking about "Western European R1b" which is clearly dominated by R1b M269 and more specifically R1b L51 men of which 99% speak Indo-European Languages today....etc.

Thanks for the reply.

Gioiello said...

@ Ryan
"If "western" R1b originated in Bell Beakers and not the steppe though, that would make much of the R1b in Yamnaya intrusive and recently assimilated from the west (which I personally think is likely but I realize most here would disagree)".

I am supporting that from at least ten years with thousands of posts mostly deleted now from blogs owned from my oppositors. At least one person agrees with you now.

Rob said...

I cannot quite make out, but the saami plot between EHG and Yamnaya (?)

Ric Hern said...

@ Ryan

The original Bell Beakers were I2a and G2a. So how could Western R1b have originated in Bell Beaker ? R1b were influenced Culturally but not Genetically by the Bell Beakers if I understand the Paper correctly.

This basically tells me that most genetic admixture R1b Men underwent was in Central Europe with Neolithic Farmers.

The Irish Paper also shows more affinity towards Central Europe rather than Southwestern Europe. So this also tells me that Bell Beakers did not have a significant genetic influence...




Gioiello said...

@ Ric Hern

"@ Ryan

The original Bell Beakers were I2a and G2a. So how could Western R1b have originated in Bell Beaker ? R1b were influenced Culturally but not Genetically by the Bell Beakers if I understand the Paper correctly.

This basically tells me that most genetic admixture R1b Men underwent was in Central Europe with Neolithic Farmers.

The Irish Paper also shows more affinity towards Central Europe rather than Southwestern Europe. So this also tells me that Bell Beakers did not have a significant genetic influence..."

Don't forget that these said scholars tested everywhere but not from Tyrrhenian Italy, Liguria, Southern France through the Rhone-Rhine line and of course Switzerland. It seems that aDNA from Sion is waiting. Let's wait we too... before saying more.

Matt said...

@ Davidski, thanks for that. It looks like it gets down to a worst Z down to 2.936, which is comparable to the best previous graphs I managed that covered the same populations, and with a lot less complexity in ancestry at the HG and Neolithic level and generally lower drift per edge:

http://i.imgur.com/87TN6YM.png

Still a few interesting stats it doesn't fix, from worst fitting Yoruba outgroups

D(Yor, Cau, Wes, Ibe), Model: 0.002721, Actual: -0.000034
D(Yor, Cau, Bar, Ibe), Model: -0.000862, Actual: -0.002036
D(Yor, Wes, Cau, Eas), Model: 0.032653, Actual: 0.029609
D(Yor, Ust, And, Eas), Model: 0, Actual: -0.003486

(More connection between CHG and WHG / Barcin than in the model; some slight Ust_Ishim->Andamanese edge)

The absolute worst fitting non-outgroup stats still involve (>2, <3) all still involve MA-1 or CHG, so this doesn't quite fit them.

But I think it's good enough as a scaffold to use to try adding recent South Asian or European populations to.

On that, if possible, a simple model to add Lithuanian, Sardinian and Basque on to the graph and see how the model can accommodate them together: https://pastebin.com/yH2g9wB3

Thanks

Anthro Survey said...

@Rob

Yeah, few human-mediated processes in the real world lack their bumps and dips. To imagine rigid linearity in such matters is to engage in silly romanticism. I myself wouldn't be surprised if many steppe-enriched beaker groups in W-C Europe spoke non-IE languages.

That being said, there is still a good correlation coefficient between increasing steppe admixture in western europe(starting in the beaker period) and subsequent IE-zation.

Anthro Survey said...

@Sam Andrews

I fundamentally agree with your points, but keep in mind that in post-IronAge societies---characterized by sedentary lifestyles, high pop densities and structures of relatively durable, non-perishible materials---lingustic change need not always require demic shifts. As the clause implies, the time period of interest regarding IE-zation of Europe does not fit the bill, but don't forget the following shifts w/out corresponding mass migrations:

Latinization(or de-Celticization) of Europe----Only pan-Italian dreamers actually conveive of a few thousand legionaires impregnating millions of continental Gauls, Iberians, Illyrians and apparently half the Near East. Apparently, they were unprofessional, had too much time on their hands and, well, conventiently hailed from South Italy, which often happens to be the dreamers' home.

Persianization of Khorasan---No mass migrations from Fars took place there. New Persian evolved in the region from an aristocratic form of Middle Persian spoken by Sassanid governors. Later on, New Persian "back-migrated" and came to be spoken by folks in Western Iran. Again, no corresponding mass migration of Khorasanis.

Arabization of large tracts of North Africa and the Near East---certain areas were demographically impacted by Bedouin settlers, but many more show continuity from pre-Islamic times.

Nonetheless, all of my examples demonstrate a shift WITHIN a language family: Celtic to Latin, Sogdian to Persian, and Berber, Syriac, etc. to Arabic. INTER-familial shifts were all usually mediated by considerable demic changes even after the iron age and under sedentary, "built up" circumstances.

Gioiello said...

@ Anthro Survey
"Only pan-Italian dreamers actually conveive of a few thousand legionaires impregnating millions of continental Gauls, Iberians, Illyrians and apparently half the Near East. Apparently, they were unprofessional, had too much time on their hands and, well, conventiently hailed from South Italy, which often happens to be the dreamers' home".

Of course you are wrong. People from Southern Italy are the least Italian nationalists, but convinced to be all except Italians. I, the theorician of an Italian Refugium of at least hg. R1b1 and subclades (also of the few L23 found in Eastern Europe), come from Tuscany, from at least 1300 AD. About what is "Italian" or "Roman" in modern Europeans will know next, through aDNA, when your friends will decide to test also aDNA in Italy, beyond the 6 samples tested so far.

Davidski said...

@Matt

fatalx:
> 2 edges at B

batman said...

Davidski,

"And metallurgy is not a Proto-Indo-European thing. It was introduced into Indo-European societies by foreigners."

There's ample indications of a relationship between the origin of the (pre-neolithic) ceramcis of northern Eurasia and the metalurgy of northern Eurasia.

Furthermore, the later evolution of pottery and metallurgy - both - have obvious relations with the spread anevolution of agriculture.

Finally we may see that the evolution of all three fields of competence and capabilities are related to the mesolithic routes of travel and trade - basically following the waterways along coasts and major rivers.

During the last two decades there's been several discoveries of major networks - of civil travel and organized trade - across most of Europe and northern Asia already during late-mesolithic Europe. That includes the Atlantic facade and the Mediterranean waters in the west, as well as the Baltic, the Black and the Caspian Seas in central Eurasia. The latter connected via the first, organized river-trade, no later than 8.000 BP.

As the traces of a "basal Eurasian" trading-culture goes back to the mesolithic time, there's no more room for a "unknown, nostratic origin" of the I-E expressions, as recognized in languages, forms and formats, symbols and structures.

Today we may even link the forms and formats of late-Paleolithic symbolism and the well-known world of the Corded, Combed and Pit-Wares. Today we may even follow a tradition of decorative symbolism from the very late-Paleolithic Mezin - via the pottery of the North Eurasia's Neolithic, as well as the art of Eurasia's Bronze Age and Iron Age...

http://donsmaps.com/wolfcamp.html

Per consequence there's no factual reasons left to claim that metalurgy - as little as pottery and agriculture - was explored and developed into (common) craftmanship - in isolation from the Indo-European culture. Unless you're in need to claim that the cultural expressions of the first I-E speakers originated south of the 40th paralell...

Grey said...

@Ric Hern

"Some people want to make too much of exceptions and therefore confuse themselves and others."

and if Basques have EHG but not CHG (iirc) then some people want to pretend Basques don't disprove the idea that "steppe" = exclusively EHG+CHG combination.

either that EHG component was in the west originally, moved to the steppe, mixed with the CHG and then bounced back again as EHG+CHG

or

there were two waves from the east, one just EHG and the second a mixed EHG+CHG

Matt said...

@ Davidski: Fixed the trifurcation there: https://pastebin.com/WLQBihfh

Vara said...

@Davidski

"And metallurgy is not a Proto-Indo-European thing. It was introduced into Indo-European societies by foreigners."

How did you reach this conclusion?

Proto-Indo-Europeans held metal work in high regards, and all of the known IE people had a smith god or a smith hero, Hittites included.

Ryan said...

@Sam:

Modern English really do have Anglo Saxon ancestry, modern Russians really do have proto-Slavic ancestry, modern Icelanders really do have Medieval Scandinavian ancestry. Why would Iron age Celtic speakers have been any different?

They wouldn't be and I didn't say they were. It's a question of where the majority of the gene pool came from. We know that the bulk of most of western Europe's Y-chromes and close to half of its aDNA derives from the central European Bell Beakers. Are you claiming that those are the proto-Celts? If so, do you also claim that the Iberian Bell Beakers were proto-Celtics? And if so, how did they become IE without any steppe ancestry?

@Chad:

Iberian Beakers did not L51. There is no demic movement from Iberia. You should read the papers.

I read the papers Chad. We have 7 Iberian Y-chromosomes. 2 are R1b. 1 of those two is L51 negative. The other we don't know if it was L51 or not. Either way a sample size of 1 isn't much to draw a solid conclusion from. We have a grant total of 0 R1b L51 samples that pre-date Central European Bell Beakers.

And yes there was demic movement from Iberia - very limited demic movement. Maybe you should follow your own advice and read the papers yourself.

And that kind of proves my point doesn't it? The Iberian Bell Beakers' spread their culture with only a modest genetic impact. Apparently it's impossible for Celts to do the same? Come on.

@Anthro:

I fundamentally agree with your points, but keep in mind that in post-IronAge societies---characterized by sedentary lifestyles, high pop densities and structures of relatively durable, non-perishible materials---lingustic change need not always require demic shifts. As the clause implies, the time period of interest regarding IE-zation of Europe does not fit the bill, but don't forget the following shifts w/out corresponding mass migrations:

Anthro - the Bell Beaker paper contradicts what you just said here. The Iberian Bell Beaker culture spread to central Europe with only a small

Nonetheless, all of my examples demonstrate a shift WITHIN a language family: Celtic to Latin, Sogdian to Persian, and Berber, Syriac, etc. to Arabic.

Just FYI, "language family" doesn't have a robust definition, and ancient peoples certainly did not know what corresponds to a language family and what doesn't. They didn't stand up and say NOPE WILL NOT LEARN THAT LANGUAGE SORRY. Please don't pretend this is somehow absolute.

I would suggest to you that you may be on to something if you apply that to the Iberian Bell Beakers' spread though. This occurred either just after or during the WHG resurgence that was occurring throughout Europe, and Bell Beakers often have an excess of WHG ancestry. Maybe part of what aided their initial spread is they were tapping in to an existing WHG-derived cultural network.

@Gio: R1b were influenced Culturally but not Genetically by the Bell Beakers if I understand the Paper correctly.

Mostly yah, though IIRC it wasn't 0 genetic influence, just relatively low. 2/7 Iberian Bell Beakers were R1b though.

Hector said...

"Proto-Indo-Europeans held metal work in high regards, and all of the known IE people had a smith god or a smith hero, Hittites included."

So if all the Tungusic peoples have thunder gods, they must have had the ability to control the weather or something.

André de Vasconcelos said...

@Ryan

There are also the BA SW Iberian samples from Martiniano 2017, two of them are R1b1a2a1a2, and a third one R1b1a2. They supposedly have steppe ancestry, but none of them scored any CHG, though

Vara said...

"So if all the Tungusic peoples have thunder gods, they must have had the ability to control the weather or something."

Because thunder = smithing, right?

Ryan said...

And on the who speaks a different language here than their parents - a lot of people where I'm from. Immigrants don't always pass on their language to their kids, and a lot of language shifts from Gaelic to English or from indigenous American languages to English had a pretty clean break (unfortunately).

Anyways, I'd appreciate if the nay-sayers here could clarify their position, as you are either saying:

The bulk of western Europe's Y-DNA pool post-dates the Bell Beakers and originated primarily with iron age Celts.

Or the Bell Beakers were Celtic

Or the Bell Beakers changed languages but not culture when they reached central Europe.

Chad Rohlfsen said...

Ryan,

Those 2 R1b in Iberia are both negative for m269. There were about a half dozen other Beakers from Iberia in a previous paper. None of them were R1b. There was no demic movement from Iberia. If you actually read the paper, you'd know that. qpGraph and Rarecoal both say no. So, that's that. There's no more arguing for it. Finished.

Davidski said...

@Vara

Proto-Indo-European has very limited vocab related to metallurgy. This doesn't look good for the idea that is commonly argued in the comments here that the Proto-Indo-European culture was spread on the back of metallurgy.

More on the topic of metal and Indo-Europeans at the Encyclopedia of Indo-European Culture.

https://books.google.com.au/books?id=tzU3RIV2BWIC&pg=PA379&lpg=PA379&dq=Encyclopedia+of+Indo-European+Culture+metal&source=bl&ots=wWs5-64bdH&sig=1H0DMRytFFEp6_lPu_bLK0AxEZE&hl=en&sa=X&ved=0ahUKEwjynqrbrPjUAhUEI5QKHfG2AukQ6AEIKjAB&authuser=1#v=onepage&q=Encyclopedia%20of%20Indo-European%20Culture%20metal&f=false

Davidski said...

@Matt

I think this model is too complex. I've never seen this before...

fatalx:
looping

Ryan said...

@Andre - Thanks, I can't believe I missed that. If I'm reading that right it doesn't necessarily say it's EHG ancestry over something like KO-1 (who had some ANE but less than EHG did), no? From my comments above:

I would suggest to you that you may be on to something if you apply that to the Iberian Bell Beakers' spread though. This occurred either just after or during the WHG resurgence that was occurring throughout Europe, and Bell Beakers often have an excess of WHG ancestry. Maybe part of what aided their initial spread is they were tapping in to an existing WHG-derived cultural network.

If there are Bell Beakers that lack CHG then they can't be descended from Yamnaya. A somewhat steppe admixed WHG source around the Carpathians would make sense as a donor into both Yamnaya and Bell Beakers though. Still a bit skeptical about totally hijacking a culture like that though, but R1b does seem to radiate out from the Danube.

Ryan said...

@Chad - Check again. One is negative. The other they just couldn't test any additional markers.

Chad Rohlfsen said...

Plenty of Copper Age Iberians are out there. I think about 30. None are L51. Rarecoal and qpGraph are the nail in the coffin. There will also be CHG in the Poruguese samples. Basing anything on Admixture output would be a mistake.

Ric Hern said...

Did the Globular Amphora samples show any CHG admixture ?

Matt said...

@Davidski, thanks for trying. I've simplified it down again:

Without Basques: https://pastebin.com/k56GDaQV
With Basques: https://pastebin.com/H39wAqLg

Previous model had a few different divergent WHG, AN, Steppe_EMBA related edges to accommodate different drift, so removed those in favour of using the same ones as Iberia_CA; also removed Basques in favour of just having the two most distant populations.

If that still doesn't work, it may just be that this model is too complex to try and stick Bronze Age European admixture under.

(Probably would've guessed the looping error is from the model not being able to converge...)

Ric Hern said...

Did Steppe R1b that was found at Csepel enter Hungary through Southeastern Poland and Slovakia or through Romania ?

Did R1b maybe split in two directions with one group migrating along the Northern side of the Carpathians and the other hugging the Black Sea Coastline towards Bulgaria ?

Davidski said...

@Matt

Still looping.

Simon_W said...

Gioiello said: "Don't forget that these said scholars tested everywhere but not from Tyrrhenian Italy, Liguria, Southern France through the Rhone-Rhine line and of course Switzerland. It seems that aDNA from Sion is waiting. Let's wait we too... before saying more."

What are you brabbling again?

Figure 1b from Olalde et al. 2017:
Geographic distribution of samples with new genome-wide data

https://jpst.it/118FD

I see samples in Southern France, Savoy east of the Rhone, Parma, which is close to Liguria, and in Alsace on the Rhine, close to Switzerland.

Simon_W said...

@Ryan

Olalde et al. 2017, page 6: "Conversely, the Neolithic farmer-related ancestry in Beaker Complex individuals outside Iberia was most closely related to central and northern European Neolithic populations with relatively high hunter-gatherer admixture (e.g. Globular_Amphora_LN, P = 0.14; TRB_Sweden_MN, P = 0.29), and we could significantly exclude Iberian sources (P < 3.18E-08) (Fig. 2c). These results support largely different origins for Beaker Complex individuals, with no discernible Iberia-related ancestry outside Iberia."

Simon_W said...

@Ric Hern

Judging from the ADMIXTURE runs in Mathieson et al. 2017, the GAC had neither CHG nor Steppe admixture, they were simple MN Farmers with particularly strong WHG.

Simon_W said...

Over 80% of Bell Beaker samples outside of Iberia had R1b, especially R1b-P312, and the large majority of them had substantial Steppe admixture, unlike all local cultures prior to the Bell Beaker era, except Corded Ware of course. I don't want to jump to conclusions, but what this appears to suggest is a significant eastern ancestry in these first spreaders of R1b-P312, not a significant Italian ancestry.

Simon_W said...

Strangely, R1b is very rare among Iberian Bell Beaker samples and the large majority of them had zero steppe admixture. How odd, it looks as though R1b in Bell Beakers was correlated with Steppe admixture?! How is this possible?

Matt said...

@ Davidski, cheers. Do the following loop as well?

Bell Beaker Germany and Hungary BA: https://pastebin.com/UthHtnzn
Same topology, Lithuanian only: https://pastebin.com/BJeubFS8
Same topology, Sardinian only: https://pastebin.com/PQ7xzBxr
Same topology, Basque only: https://pastebin.com/Uaefg1Sa

(Wondering if there's some issue with using the modern here, or an issue with fitting the populations together, or if it's something more basic to do with the pNorthCaucasian edge into the Anatolian population).

How long do these take you to run btw? I don't really want to be spamming you with models that are taking a long time to run.

Ric Hern said...

Thanks Simon.

Davidski said...

@Matt

They're all looping.

These models usually take a few minutes to run when they run correctly. The looping ones stall very quickly.

Grey said...

Davidski

"This doesn't look good for the idea that is commonly argued in the comments here that the Proto-Indo-European culture was spread on the back of metallurgy."

I don't know if anyone argues that.

I think PIE had an early (pre-smelting) metallurgy which led to small groups spreading as artisans/traders as far as Iberia and Nepal and being a minority adopting the language and possibly the religion/culture of the host population - hence the ydna but *not* the culture/language etc.

The ydna spreading *without* the culture is why it might explain the Basques.

Matt said...

Cheers. Any joy with this: https://pastebin.com/F35zhNf9? I think it should have a worse fit, but eliminates the admixture edge I'd be most dubious about contributing to the looping.

Gioiello said...

@ Simon_W

"Gioiello said: "Don't forget that these said scholars tested everywhere but not from Tyrrhenian Italy, Liguria, Southern France through the Rhone-Rhine line and of course Switzerland. It seems that aDNA from Sion is waiting. Let's wait we too... before saying more."

What are you brabbling again?

Figure 1b from Olalde et al. 2017:
Geographic distribution of samples with new genome-wide data

https://jpst.it/118FD"

In fact they decided that these R1b1a2 from France, only 1 from Italy, from 2566 to 1930 BC were introgressed from Germany and before from Yamnaya, and that R1b weren't before there, having tested some I2a1b from France (and also the I2 from Remedello), etc etc
But ask yourself, and also to your said scholars, why
1) they published their data so late
2) they haven't yet published the autosome and all the data
3) they are disproved from the comtemporaneous paper of Mathieson with all the WHG introgression in Balkans and eastern Europe, with massive R-V88 (perhaps) which would be only the "isolate" Villabruna and nothing more
4) so far no R-L51 from eastern Europe
And say them that Gioiello Tognoni thinks that also the R-L23 (like his) found at Samara came from Western Europe, and that it is clear now that R1 (and also mt beyond U5 as H, K1 etc) were the haplogroups of WHGs....

Rob said...

Yeah there was early/ 'basal' clades of R1 and R1b as far as France in the Late Upper Palaeolithic. But it does seem that by the Mesolithic the way all the lineages had sorted themselves was that P297 derived clades were in eastern Europe, V88 Mediterranean Europe.
The range of probability for M269 expansion is the late Neolithic period from Germany to Ukraine, most likely the eastern end.

Gioiello said...

@ Rob

"Yeah there was early/ 'basal' clades of R1 and R1b as far as France in the Late Upper Palaeolithic. But it does seem that by the Mesolithic the way all the lineages had sorted themselves was that P297 derived clades were in eastern Europe, V88 Mediterranean Europe.
The range of probability for M269 expansion is the late Neolithic period from Germany to Ukraine, most likely the eastern end".

Read my posts, if you find them, now that pretty all the sites where I wrote are out (why?), and you will understand that very likely it wasn't so.

Davidski said...

@Matt

Seems like it's stuck. Keeps running and running.

Matt said...

@ Davidski, thanks for trying. Weird, model seems relatively simple -

Anatolian=Basal Eurasian+UHG;
European_Farmer=Anatolian+WHG;
Caucasian=Basal+EHG;
Steppe=Caucasian+EHG

Seems like then having any kind of European_Farmer+Steppe in these models just wrecks qpGraph. I think that exhausts it :(.

Chad Rohlfsen said...

@ all,

Look at this craziness for Yamnaya. No East Asians though.

worst f-stat: CHG Bic Yam EHG 0.019592 0.020210 0.000619 0.001716 0.361

https://drive.google.com/file/d/0B962TtPkX1YnaVVFWkVWVjg4Q1E/view?usp=sharing

Jen Gittings-Dalton said...

@ Ryan

"This basically tells me that most genetic admixture R1b Men underwent was in Central Europe with Neolithic Farmers.
The Irish Paper also shows more affinity towards Central Europe rather than Southwestern Europe. So this also tells me that Bell Beakers did not have a significant genetic influence..."

So after following this blog for years but never posting, I am asking for a favor: can someone here with the gift of translation into archaeological and lay language, like
Gray, Rob or Batman, attempt a summary of the genetic history of the Irish, and the Welsh? I know they are genetically different. I am fascinated that the ripple of Yamnaya sloshed from Hibernia to the Himalayas. I have read the paper on Rathlin etc. But I need a "deep history" summary of dna/geographic migrations, back to the mesolithic if possible. Even earlier if one is willing to guess. For example, are the Gravettians linked in any way with the early Holocene expansions down the Dnieper, Dniester, and Volga rivers which eventually contributed to PIE feeder populations? Is it coincidence that the Irish remain, and the Scythians and Sarmations were, red-haired? The Welsh are not. I would be grateful for anyone risking a post on this for a layperson, and will sink back into the shadows here where I know I belong. Thanks for this blog!

Anthro Survey said...

@Ryan

Adopting an entire language is not tantamount to acquiring certain elements of a material culture and the skillsets required to reproduce them.

"ancient peoples certainly did not know what corresponds to a language family and what doesn't."

Not on a deep, analytical level, no. I don't see how that changes things? I'm pretty sure Parthians intuitively grasped how similar their language was to Bactrian. It's also fair to say Berbers had an easier time adopting Arabic due to their many parallels. Or how, say, Ukranians can understand limited Polish speech and vice versa.

Chad Rohlfsen said...

@ all,

This one is cleaned up even more. Looks pretty good.

worst f-stat: Din Bar Bic Yam -0.001387 -0.001266 0.000120 0.001120 0.108

https://drive.google.com/file/d/0B962TtPkX1YnaV9GTHdNY3lCVk0/view?usp=sharing

Davidski said...

@Chad

Those Yamnaya trees look similar to models for Yamnaya using the Basal-rich K7 and nMonte, with the big chunk of Anatolian/EEF-related stuff.

Chad Rohlfsen said...

The last one is about as perfect as it can get.

Chad Rohlfsen said...

Adding Bell Beaker. Yamnaya goes back to Ukraine_HG like, with a lot of the same in Beaker.

worst f-stat: CHG Bic Bar Yam -0.001889 -0.002314 -0.000426 0.001376 -0.310

https://drive.google.com/file/d/0B962TtPkX1YnUXZiU1J0VEZYb0U/view?usp=sharing

Seinundzeit said...

David,

For what it's worth, I think if one is looking at modelling more "recent" ancestry proportions (like Neolithic or Bronze Age ancestry proportions), nMonte with PCA or the Basal-rich K7 is probably best.

On the other hand, if you want to examine really deep relationships (with Upper Paleolithic Eurasian genomes or something), qpGraph is probably best.

I think David Reich and company have come across a similar insight, which explains the new interest in haplotypes.

So, I think that your Global_10 PCA datasheet/Basal-rich K7 will prove to be very important tools, for a very long time to come. It'll be very exciting to start experimenting with those again, once the new aDNA data is added.

Anyway, I was wondering, could you eventually try this topology for the Kalash:

https://drive.google.com/file/d/0B8XSV9HEoqpFLUNIVkVNeU55Yms/view?usp=sharing

But with only one change; Srubnaya_outlier in place of Yamnaya.

As always, only once you find the time, and only once you have the inclination. No rush.

I'd really appreciate this. Thanks in advance.

Rob said...

Chad
Interesting
Because that is what I was getting with nMonte PCA for BB- ANF , CHG and WHG; little EHG. (~6%)
Yamnaya is somewhat similar but with a lot less ANF
Can you do one for Mako I1502 ?

Davidski said...

@Sein

I generally try to avoid using singleton very low coverage, or even non-UDG treated, samples in qpGraph, but...

https://drive.google.com/file/d/0B8XSV9HEoqpFZmNsR2xPczNrM1U/view?usp=sharing

https://drive.google.com/file/d/0B8XSV9HEoqpFV3hrT1Fzb0t5RlU/view?usp=sharing

Matt said...

@Chad, nice graph. The Basal Eurasian is pretty consistent with most measures in Yamnaya (Anatolia and CHG very slightly lower than in models with Onge?), the HG is pretty different to the norm.

Bell Beaker is still fairly close to the expected values of what would be 50% Yamnaya, 50% MN (= 37.5% Anatolian, 12.5% WHG) in your graph... but BB is slightly lower in Basal Eurasian and UHG from Anatolia and EE_Meso from Yamnaya; higher in WHG instead.

How do AG-3/MA-1 and European_MN samples fit on the graph?

Seinundzeit said...

David,

Thanks! I very much appreciate this.

Although, it is a shame that the Srubnaya_outlier isn't a high-coverage sample. Doesn't seem to improve the fit.

Matt said...

@ Chad, I guess another thing: if the models where Loschbour = Bichon+EHG hold, then no separate EHG edge required into pBeaker necessarily?

@Davidski, couple of graphs trying to hybridise the close fitting working model Chad has produced with the top layer branching structure from earlier topology I was running.

Probably will be a worse fit than anything Chad would get by just carefully adding in new populations to his graph mostly just to see if these still get "looping"!:

Hybrid A (fewer samples): https://pastebin.com/SxU8vH3L

Hybrid B (more samples): https://pastebin.com/UZK79dUM

Davidski said...

@Matt

The first one worked, but some of the labels were off, like MA-1 instead of MA1 and Onge instead of Andamanese_Onge.

https://drive.google.com/file/d/0B8XSV9HEoqpFUkVKdmtWY3NFczA/view?usp=sharing

https://drive.google.com/file/d/0B8XSV9HEoqpFLWttZjFtWnJmWGM/view?usp=sharing

The second one appears to be stuck.

Grey said...

Jen Gittings-Dalton said...

"attempt a summary of the genetic history of the Irish, and the Welsh? I know they are genetically different"

my opinion - which may be a minority of one - is the key to the (neolithic onwards) difference is the mostly ignored Atlantic Megalith culture and a more dramatic population turnover from Atlantic Megalith population to Yamnaya related population in Ireland than Wales cos Wales has more mountains.

Southern Portugal is part of the Mediterranean climate region so was (imo) the furthest point an east med. farmer population could reach (either by the southern or northern med. coast route) without needing to change their farming package - and they reached it quite early imo.

Once there I think they developed into a neolithic version of Venice, with the large settlement(s) restricted to Southern Portugal (cos farming package) but with trading and resource extraction colonies along the Atlantic coast, north and maybe south too - hence the concentration in places like Brittany, Ireland, Wales and SW England which had a lot of soft metals i.e. i don't think they were densely populated across the whole territory (cos unsuitable climate for their farming package)(thus possibly reliant on fish?) but were concentrated around certain resource extraction sites (nb similar to one of the theories about Zimbabwe).

Then the more Yamnaya/Corded Ware populations arrived in Britain from around the Rhinemouth and/or North Sea and either squished the Atlantic Megalith colonies, with more survivors in regions with the best refuges (like Wales) *or* the incomers simply took over unused land that wasn't suitable for the neolithic package but was suitable for their dairy farming package and simply swamped the Atlantic Megalith colonies by weight of numbers (less so in regions less suited to cattle, like Wales).

(or a bit of both)

(or even the dairy herders were invited to provide food for the mining colonies nb Zimbabwe again)

So (over simplifying) say Britain and Ireland both started out as 100% AM (Atlantic Megalith) but ended up with different proportions of AM and CW (Corded Ware like) say
- England, Ireland, Scotland : low AM, high CW
- Wales: low-medium AM, high CW
so the difference (imo) is in the proportions - Wales (and certain other more mountainous/swampy regions like Cumbria in England) having a (still small but) higher proportion of the AM component.

I think that's the physical difference the Romans noticed.

#

"Is it coincidence that the Irish remain, and the Scythians and Sarmations were, red-haired? The Welsh are not"

personally i think the red hair thing is a major clue

#

but i'm a big guesser so take all with a pinch of salt or two

Matt said...

@ Davidski, cheers; I was hoping it would combine the better features of both, but looks like adopting the mix of the high (more ancient) level from my graphs and the lower (more recent) levels from Chad's just leads to qpGraph fitting less well / producing more bad stats.

With the second graph halting, I guess at least that points to the problem being either in the high levels or the sheer number of layers or populations in the graph, rather than anything going on in the low levels.

ak2014b said...

"And there is aDNA from South Asia. I said that in the blog post."

Being disappointed again by the lack of a paper must have prevented me from reading down to the "It's not exactly the best kept secret out there that ancient samples from India and Pakistan are now ready, and...oops I probably can't say more than that for now. Pity."

However, if the paper or genomes aren't published yet, knowing they're ready is still no more useful than before. Unless you already have the genomes to inspect.

Are you allowed to hint at when any South Asia or South Central Asia paper is coming out at last? And the Mycenaean one, that's also been dangled about for so long?


"Limited aDNA from Iron Age Poland is interesting, but not relevant to the Indo-European debate."

Even if they were Goths rather than Slavs at that time, they'd still be Indo-European speakers too, and in that sense ought to be closer to PIE than anyone alive.

And once ascertained that we're dealing with either Gothic or Slavic speakers, then we at least know for a fact that they spoke Indo-European, whereas it's a fair assumption that Yamnaya would have spoken Indo-European, but we still don't know so, because there's no actual record of what Yamnaya or Corded Ware spoke.

I think both reasons make Iron Age aDNA from Poland relevant, despite the limitations of the sample set.

Davidski said...

How are you going to ascertain what language they spoke?

Linguistic data relevant to metallurgy suggest the survival of non-Indo-European metallurgical traditions in Central Europe as late as the Iron Age.

So what if these people were very late non-Indo-Europeans, like those that provided the non-Indo-European substrata in Germanic?

Rob said...

@ Dave

"Linguistic data relevant to metallurgy suggest the survival of non-Indo-European metallurgical traditions in Central Europe as late as the Iron Age"

theres no such data do be had because such a pattern didn't exist


Anyhow, there are runic of inscriptions in Wielbark contexts

Matt said...

@ Grey: So (over simplifying) say Britain and Ireland both started out as 100% AM (Atlantic Megalith) but ended up with different proportions of AM and CW (Corded Ware like) say

- England, Ireland, Scotland : low AM, high CW
- Wales: low-medium AM, high CW


The slightly exasperating thing about this is we've discussed this before and to the extent we have samples of Welsh people - and we do have thousands(!) with the POBI and UK10G -it indicates that the Welsh seem have fairly comparable Yamnaya / Corded Ware ancestry to the Scottish, if anything slightly less than the Irish, and more than the English - https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5097941/.

All that seems to happen is some specific local genetic drift.
So I don't really get why this idea persists for you? The evidence points against it pretty clearly?

epoch2013 said...

@Chad

The last one is interesting: EHG = 70% WHG + 30% EE_Meso. Is that an offshoot of ANE? Also, the consistent popping up of an unknown HG, sister clade of WHG is interesting. Could Goyet116 or El Miron take that place? Or could WHG even be UHG + Vestonice + El Miron (+tad ANE), which is my pet theorie. If you swap Barcin for Natufians, what happens. Natufians show he similar affinity to WHG as Anatolian, albeit a tad smaller, in D-stats.

AJ said...

@ak2014b

He doesn't know anything. Neolithic India paper will be out in late-September/early October, according to person who is part of the project.

It will come out with Biomolecular Prehistory of South Asia Project (Reich et al.) and Reconstructing the human population history of the Indian subcontinent using ancient population genomics. (Rai, Thangaraj et al.)

Chad Rohlfsen said...

EE_Meso is off the same branch as EHG. Look it over again.

Davidski said...

@Rob

I was just briefly entertaining ak2014b's nonsensical musings that the Iron Age Polish results have anything at all to do with the Indo-European debate. Although what I said was correct in so far that there is speculation about non-Indo-European metallurgical traditions surviving into the Iron Age in Central Europe.

@AJ

Very presumptuous of you to suggest that I don't know anything in this context. I do, for instance, know that the data support Bronze Age migrations from the steppe into India.

Karl_K said...

@AJ

"It will come out with Biomolecular Prehistory of South Asia Project (Reich et al.) and Reconstructing the human population history of the Indian subcontinent using ancient population genomics. (Rai, Thangaraj et al.)"

And this is why Reich and the others have been so against the steppe theory in their latest talks.

Al Bundy said...

Agreed.Actually SimonW wrote that in a previous post but it was about all LPIE coming from Yamnaya and IndoIranian is LPIE.

Al Bundy said...

We agree the steppe was at the least a spread zone for IE languages but because the steppe people's were subject to a lot of influences from the Caucasus and surrounds the possibility exists they took IE from there.But that's not reflected in the genome evidence, quite the opposite.

Rob said...

So what do you call the shift from 0% CHG to 55% ChG (all in the space of a few hundred years) ?
A mirage ?

Rob said...

And why do ancient saami plot next to steppe eneolithic ?

Al Bundy said...

It's huge no doubt, but it was mostly female-mediated.e

Davidski said...

@Rob

Ancient Saami don't really plot next to Steppe Eneolithic.

It's not possible to accurately analyze most Saami in PCA of West Eurasian variation because they have 15-20% East Asian admixture, which the other samples lack.

This East Asian admixture pushes them artificially east and south, so that they end up next to ancient steppe samples even though they have very different genetic structure.

Rob said...

Al don't worry I'm just clowning
#slowdays

AJ said...

@Al Bundy

"all LPIE coming from Yamnaya and Indo-Iranian is LPIE."

On his Eurogenes Yamnaya K6 spreadsheet, Armenians on average have 7%-20% Yamnaya admixture, and Iranians have 20% Yamnaya admixture. Are you claiming their steppe admixture is Indo-European? Because such small admixture can't lead to language shift.

PIE may as well have originated somewhere in Caucasus and influenced steppes. Steppe-admixture in Armenia and Iran may represent later intrusions, instead of Proto-Armenian or Proto-Indo-Iranian.

"if it wasn't from the steppe, where did it come from?"

I'm not suggesting there was no steppe migration to S.A, but attributing entire "steppe" component to Indo-Europeans without aDNA from Paleolithic and Mesolithic Central Asia is unrealistic. It's unrealistic to model Tajiks 60%-70% steppe and say 'all' this steppe admixture represents Indo-European admixture. There are also various Non-Indo-Europeans from Urals to Himalayas who score very high steppe admixture.

We don't know what Paleolithic and Mesolithic population were like between Urals and Himalayas due to lack of aDNA from that period and if there is something (non-basal) inflating steppe admixture.

If there were movements in Bronze Age, there will be similar movements in Paleolithic and Mesolithic along such trail, ANE in EHG is example of that, we won't know without aDNA.

Davidski said...

You're ignoring the Y-chromosome and linguistic evidence.

You have to consider all of the evidence at once, not try to debunk each set in a vacuum as if the others didn't exist.

There's a very good reason why at least the LPIE homeland is being put on the steppe after the latest rounds of ancient DNA. You're in la la land pretending as if this isn't happening. Well it is.

AWood said...

@Grey

Modern Welsh have some of the highest frequencies of red hair in the world! What are you talking about man?

Perhaps if the neolithic signal that is seen in Welsh today is a remnant from Spain and perhaps those E-V13 guys who were unlikely to be light skinned or red haired for that matter.

Ric Hern said...

@ AJ

There are many scenarios that can prove that a 20% Admixture is enough to bring about Language change....

Samuel Andrews said...

What ancient Saami PCA?

epoch2013 said...

@Chad

O, OK, so that's how the model is set up? Doesn't matter, my main curiosity is whether Natufian/Levant would get similar admixture of that UHG as Anatolian.

Chad Rohlfsen said...

Fairly similar, but Anatolia also has a good chunk of ANE on top of that. Natufians are much more BEu. I am actually playing around with the deep ancestry of ancient farmers.

Chad Rohlfsen said...

This wasn't set up with that focus. It would need too many admixture edges.

Seinundzeit said...

I have a tendency to do this; this is somewhat random/off-topic, but still somewhat relevant...

Since haplotype-based analyses are apparently the way of the future (or, at the very least, that's the impression I've gotten), I really wanted to see some ancestry proportions based on said data (for South Central Asians).

So, I found this for Pashtuns and Pamiris, in the WC1 paper.

Pathan

78.45% Iran_N (WC1)

12.85% WHG (Loschbour)

6.11% ENA (Han)


Tajik_Pomiri

66.68% Iran_N (WC1)

24.16% WHG (Loschbour)

6.7% ENA (Han)


Compared to "High Caste" North Indians:

Gujarati_A

72% Iran_N (WC1)

18.95% Ust-Ishim

8.49% WHG (Loschbour)


Compared to "Scheduled Caste" South Indians:

Mala

47.9% Iran_N (WC1)

47.87% Ust-Ishim


Very sensible, but they do suffer from a lack of ancient references.

For example, I really wish they could have included a high-coverage genome from the ancient steppes.

And now that we have Iran_Chal, it would be very interesting to see how those samples work for Tajiks/Afghans/northern + western Pakistanis...

epoch2013 said...

@Chad

"Fairly similar, but Anatolia also has a good chunk of ANE on top of that. Natufians are much more BEu. I am actually playing around with the deep ancestry of ancient farmers."

Would be interesting to see if these can be modeled as something pre-K14 + something pre-ANE + BEu. That way we have three basis groups with which to roughly explain West-Eurasians.

An out of Africa scenario leads to a group moving to the Middle-East/Iran, split in K14-ish and ANE-ish. Some groups move north, remainers admix with basal Eurasian groups. One question would be: Where were the Basal Eurasians? Why don't they appear in ANE/Western HG's? Another would be: What is the place of Oase 1 in this?

ak2014b said...

"How are you going to ascertain what language they spoke?"

It would be the archaeologists and historians who'd be trying to ascertain such things, if and where possible, in attempting to identify the people and language. They'd probably be doing so based on the archaeological contexts of the burials and what may be known about the region at that time, such as which people were to have inhabited various areas. The periods of the samples being far nearer to the present than the Bronze age, these things would possibly be more confidently traceable.

The geneticist authors of the paper on Polish Iron Age to Medieval genomes themselves choose to consider a connection between the I1 and "Germanic tribes and Vikings":

We successfully assigned haplogroups to sixteen individuals. Eight belonged to haplogroup I1 (I-M253). Three of them belonged to the sub-branch I1a3a1a1a (I-L1237) and one to I1a2a (I-Z59). I1 is the most common haplogroup in present day Scandinavia, and it is found in all places invaded by ancient Germanic tribes and Vikings. Four samples belonged to haplogroup G2a (G-P15) which is spread uniformly throughout Europe. Other individuals were assigned to I2a2 (I-M436), R1a (R-M420), R1bl (R-L278), E1b1 (E-P2). The next portion of samples is under investigation.

Davidski said...

I see, so it's nothing that hasn't been done before, like linking the relationship between R1a-Z282 and R1a-Z93 with Balto-Slavic and Indo-Iranian languages, and inferring that the early Indo-European ancestors of Balto-Slavs and Indo-Iranians came from the same place during the Bronze Age and carried R1a-Z645?

In which case, what does it matter if (a few?) groups in Iron Age Poland carried high frequencies of I1 and maybe spoke Germanic or some other Indo-European language, since no Indo-European speakers in Asia carry I1 and thus are not recently paternally related to them, and thus it's highly unlikely that any early Indo-European groups carried a lot of I1?

Maybe you just feel like discussing Iron Age Polish DNA? If so, why bring up the Indo-European angle at all?

epoch2013 said...

@ak2014b

Quite so. And how did that came about? Because we have a number of ways to consider Germanic presence in that area. Roman factual sources. Germanic mythical sources. The Wielbark culture conveniently fitting what can be distilled from these sources. And now DNA.


And archeologists have long considered Corded Ware Culture the source for IE in Europe. Linguists have a number of reasons to consider the North-Pontic as the PIE Urheimat. We find R1a in Corded Ware and in India way after these theories were formed.

Paleogenetics, how utterly revolutionary it may be, is a archeological tool to validate existing theories. It is in that respect just like C14 dating, which was just as revolutionary in its early days. And so far paleogenetics validates the steppe theory very well.

Al Bundy said...

Corded Ware is the source for Greek? It probably came from the steppe,we'll see what the Mycenean paper says but haven't heard that one.

Grey said...

Matt

"So I don't really get why this idea persists for you? The evidence points against it pretty clearly?"

I think it's regional
- Strathclyde (SW Scotland)
- Rheged (Cumbria) (NW England)
- Gwynned (North wales)
so it won't necessarily show up in national scale comparisons.

http://www.robertsewell.ca/britain802.jpg

Grey said...

AWood

"Modern Welsh have some of the highest frequencies of red hair in the world! What are you talking about man?"

I didn't say none. The point is there's always been a thing in the UK about the "Celtic Fringe" where Wales, Ireland and partly Scotland are grouped together but they look different, which is interesting imo.

#

"Perhaps if the neolithic signal that is seen in Welsh today is a remnant from Spain and perhaps those E-V13 guys who were unlikely to be light skinned or red haired for that matter."

Yes - the valleys in north Wales with lots of ydna E might provide a clue one way or another.

Matt said...

Grey: I think it's regional

Things like the POBI have thousands of samples with regional structure over large geographical scales. !national scale. So worth checking out if your regions are in there. If these are large regions, relatively representative of their country, then they should have good presence in UK Biobank and POBI.

IRC the POBI sample in a region overlapping the Gwynedd region. This is where the N Wales come from in POBI, which are used in Galinsky 2016 (Population Structure of UK Biobank...) - http://www.well.ox.ac.uk/_asset/image/pobi-map-jpg.jpeg. Majority of the N Wales cluster they find.

Really, check out - https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5097941/, esp Figure 2 (POBI projected on to Biobank) and Figure 3 (ancients projected on) and Table 2 (direct statistical comparison between clusters in affinity to ancients).

It doesn't seem like people from Gwynedd (as majority of N Wales cluster) have more early farmer ancestry than the English and then the South Welsh have more steppe ancestry than the English and that balances out for Wales as a whole or anything like that.

If we're talking about microregions, like really down to the scale of individual valleys and such, not really representative of their country then.... Firstly, you would need quite a strong degree of regional inbreeding for that to be true, and all the Fst and clustering haplotype data as in POBI indicates that doesn't really happen in the modern UK - North Wales is not Friuli-Venezia Giulia. Secondly, they wouldn't really tell you much about "the Welsh" or "the Scottish" as a whole, so that could be a bit misleading to describe them as telling anyone much about "the Welsh", "the Scottish", etc.

epoch2013 said...

@Al Bundy

Make that the source for a large part of IE in Europe. My point is that DNA results aren't the base of the idea of the steppe as PIE Urheimat, they confirm it.

Al Bundy said...

I agree to some extent but what explains the early deep split of Anatolian? That's before Yamnaya.The steppe theory is that all IE comes from the steppe not just some.It's the most likely theory but we need DNA from crucial areas like Greece,Anatolia,and India to confirm things.It's on the way.

vvv666 said...

Gyaneshwar Chaubey and K.Thangaraj write:

http://www.thehindu.com/opinion/op-ed/too-early-to-settle-the-aryan-migration-debate/article19265947.ece

"
Moreover, there is evidence which is consistent with the early presence of several R1a branches in India (our unpublished data).
"

Davidski said...

@vvv666

Moreover, there is evidence which is consistent with the early presence of several R1a branches in India (our unpublished data).

They don't have any such data. It's called confirmation bias.

Grey said...

@Matt

sorry for slow reply

"Things like the POBI have thousands of samples with regional structure over large geographical scales..."

I agree with everything you say however they look different* (at least to me). Now maybe i am imagining it but if not there has to be an explanation and if the Romans noticed it then it was probably pre-Roman (imo).

(*by which i mean there is an uncommon phenotype throughout the UK but it pops up mostly in people from north Wales or NW England/SW Scotland)

(it was especially noticeable in the UK military 30 years ago with the regional regiments)

So if my premises are
1) they look different (at least imo)
and
2) accepting that the dna differences are tiny
then to my level of knowledge that just implies the dna involved in the phenotype must be very *divergent* so it only takes a little to make an impact.

oddly enough the phenotype i am talking about: pale but black often curly hair is becoming a trend on the TV: Game of Thrones, The Expanse and the coptic kid from that show about hacking.

maybe i'm wrong but until someone explains the phenotype difference i'll believe there is (or was until recently) small pools of very divergent dna tucked away in remote parts of the UK and the most recent (pre-Roman) explanation would be Atlantic Megalith culture (and by extension that the AM culture had links with NW Africa or Egypt).

Matt said...

@Grey: Seems simpler to me that, if the phenotype is as we say, we'd simply be talking about some selective difference with some post-Bell Beaker population that is fairly unremarkable in its continuity, and not some phenotypically ultra divergent and ultra potent, yet hithero undetectable Neolithic contribution that doesn't show up in the measures of increased Neolithic affinity we have. But I don't think I'll work any power of suasion over you here. :)

I would only ask that when people ask about differences, we point people to the existing works on genetic structure before speculation which we have no genetic evidence for yet.

Jen Gittings-Dalton said...

I am interested in the possibility that remnant dna from the AM culture sub-layer may remain more phenotypically evident in the N. Wales mountains and other isolated places, in a sort of geographic "tidal pool" analogy. I wish POBI or other fine-screen data could could be explored further for patterns in a micro-regional "deep dive" in N. Wales, Cumbria, etc. Also thanks to those who responded to my query in this forum. I respect the level of composite learning here, since I don't have the genetics training to comment usefully. Still, hints and patterns which emerge through similarities in folklore, linguistics, art, ancient lit., archaeology, etc. are interesting though conjectural. I suspect your blog has many readers like myself, lurkers but appreciators.

Grey said...

Jen Gittings-Dalton said...
"I am interested in the possibility that remnant dna from the AM culture sub-layer may remain more phenotypically evident in the N. Wales mountains and other isolated places, in a sort of geographic "tidal pool" analogy."

yes - there's a valley in Wales with 30% ydna E for example which was being studied but i haven't heard anything about it since (or i missed it).

https://en.wikipedia.org/wiki/Haplogroup_E-V68#Ancient_Britain

"Significant frequencies of E-V13 have also been observed in towns in Wales, around Chester (ancient Deva Victrix) in England, and Scotland. The old trading town of Abergele on the northern coast of Wales in particular showed 7 out of 18 local people tested were in this lineage (approximately 40%), as reported in Weale et al. (2002)."

Grey said...

Matt

"I would only ask that when people ask about differences, we point people to the existing works on genetic structure before speculation which we have no genetic evidence for yet."

there is *some* admittedly very limited evidence relating to small pools of ydna E

https://en.wikipedia.org/wiki/Haplogroup_E-V68#Ancient_Britain

#

"Seems simpler to me that, if the phenotype is as we say, we'd simply be talking about some selective difference with some post-Bell Beaker population that is fairly unremarkable in its continuity,"

agreed that's the simplest explanation which is what professionals should go by but as an amateur i balance Occam's razor with Grey's razor which is what is the most interesting/entertaining theory.

#

"and not some phenotypically ultra divergent and ultra potent"

A quibble but i never said ultra potent as ultra divergent doesn't need to be.

As you know if there's a population with two bits of phenotype related dna where A is 90% and B is 10% then
- 81% will be AA
- 18% will be AB
- 1% will be BB
and if A and B are very divergent i.e. originally from two very distant regions then the AA people and BB people will look noticeably different.

And if you have three nearby populations with slightly varying proportions say
- 95% A and 5% B
- 90% A and 10% B
- 80% A and 20% B
then they will have 0.25% BB, 1% BB and 4% BB phenotype respectively.

anyway i readily admit this is a personal hobby horse and of very minor importance in the general scheme of things but there is *some* evidence (apart from my eyes) of unusual "tidal pool" dna in some bits of the UK.