search this blog

Tuesday, June 26, 2018

Genetic borders are usually linguistic borders too


Note the awesome correlation between the two maps below. The first map is mine. I posted it on this blog almost a year ago (see here). The second map is from the recent Wang et al. preprint (see here). Also note that the Steppe and Caucasus clusters as defined by Wang et al. are rich in Y-haplogroups R1 and J, respectively (see here).



Very cool indeed. But I'm still scratching my head and wondering why Wang et al. entertained the possibility in their conclusion that Indo-European languages diffused into the steppe from south of the Caucasus? That's because, as a rule, human genetic borders also represent linguistic borders, and major linguistic families are strongly associated with Y-haplogroups (for instance, see here).

See also...

Matters of geography

Likely Yamnaya incursion(s) into Northwestern Iran

Yamnaya isn't from Iran just like R1a isn't from India

63 comments:

Ric Hern said...

Maybe there was an isolated Steppe Population in the Mugan Steppe who admixed in that area during the Mesolithic and migrated to the Steppe during the Mesolithic ? That is the only thing I can think of. Maybe they have some Mesolithic Samples that we do not know about ?

Davidski said...

@Ric Hern

I suspect that there foragers related to CHG living in Ciscaucasia during the Mesolithic, and they may have already at that time mixed with EHG nearby to form populations like Eneolithic Steppe and Yamnaya and Sredny Stog II (but obviously minus the European farmer ancestry).

I had a feeling that this was an option a long time ago, when I saw how stable the Yamnaya mixture was, right from the earliest samples, but back then it seemed like a pretty far out proposition, so I didn't push it too hard in the comments here.

Ric Hern said...

@ Davidski

Yes that seems to be the picture. It certainly looks like the Caucasus region had plenty of food (Bison, Moose,Mouflon etc.) to offer Hunter Gatherers and they did not have to migrate far. In this case issolation of populations could have been common in and around that area during the Palaeolithic and Mesolithic...

Ric Hern said...

Maybe a Mesolithic CHG like population who adapted to plains living and hunting in the Shirvan Steppe ? Horse hunting probably would have been one of the primary activities in that specific area. Maybe this activity broke the ice between EHG and CHG like Steppe dwellers....

Le skipper de Pytheas said...

Hoping to find some supply of R1b-L23 in the Caucasus ?

Davidski said...

Hoping to find some supply of R1b-L23 in the Caucasus?

Pretty sure there won't be any in the ancient Caucasus, unless it's with steppe admixture.

Davidski said...

To be honest, I'm kind of shocked how much of my, and our, ancestry comes from this narrow strip of steppe land between the Black and Caspian Seas. Hehe.

Home away from home

MajorTom said...

Hey Davidski, long time fan. I am quite amateur and do not have much to contribute to the topic per se, but a curiosity; does this y Haplo/linguistic correlation have its occasional weird exceptions? I'm in particular thinking of the Basques, so rich in R1b but a people not exactly jam packed with steppe ancestry. Whats the consensus on them these days? Thanks

Tesmos said...

@Davidski,

The big question is: How much Steppe admixture do Northern Europeans actually have? Is there an update?

Unknown128 said...

I have a question for you people. There is an opinion, mostly expressed by Robert Drews, (most fully in his "Militarism and the Indo-Europeanizing of Europe"), that the "steppe migrations" of 3000-2000BC into western Europe have nothing to do with the Indo Europeans, who were a separate group, that spread extremely fast all around Eurasia with the use of chariots, but since its rule was mostly elite dominance, it did not leave much (if any) genetic trace. The wide difference, between various IE languages is explained by the fact that before languages are written down, they supposedly change much faster.
Drews seems to be a respected scholar, so I would like to ask you about your opinion on this topic.

epoch said...

It was quite a surprise to see that conclusion in this paper. The usual suspects claim that David Reich knows more - "has access to many samples" - but if they did, why not include them in this paper?

bellbeakerblogger said...

@Davidski
"during the Mesolithic, and they may have already at that time mixed"

I deleted my previous comment since I realized I'm getting wrapped around the periodization axle. I suppose by Mesolithic you mean what the Russians call the Pottery Neolithic (pre-farming)?

Like 5500-4500, I'll buy that.

Davidski said...

@MajorTom

The dominance of R1b-M269 in Basques can be explained by a relatively recent founder effect. Also, Basques do show a significant level of steppe ancestry, comparable to most Indo-European-speaking Iberians, so it seems that they're just a linguistic anomaly.

@Unknown128

I'm not really familiar with Drews' work, but if what you say is accurate, then he was wrong, because the only way to link European and South Asian Indo-European-speakers genetically is via gene flow from Bronze Age steppe pastoralists.

We can't link them via ancestry from Neolithic farming populations, because the early farmers who moved into Europe and South Asia were very different and realistically couldn't have spoken such closely related Indo-European languages as the precursors of Slavic and Hindi.

Keep in mind also that the most common Y-haplogroups in Europe today are R1b-M269 and R1a-M417, which are both originally from the steppe, and R1a-M417 is very common in South Asians, especially those who speak Indo-European languages. There's also a lot of autosomal/genome-wide ancestry from the Bronze Age steppe in Europeans and South Asians, again, especially in Indo-European speakers.

So it looks like Drews needs to update himself.

Davidski said...

@Unknown128

Ah, wait, I may have misunderstood what you said.

So Drews thinks that Indo-European languages spread into Western Europe with the horse and chariot complex, rather than with the earlier steppe migrations?

I guess this is possible, but the horse/chariot complex originated in the Sintashta culture in the steppes south of the Urals, and there's no trace of these people in Bronze Age Western Europe, even as elites.

But they do pop up in the Bronze Age Balkans and Central Asia at the right time to be the source of Indo-European languages there.

Arza said...

rozenfeld @ AG said...
So, the schedule for ISBA 2018 is now available, so one can see the titles of presentations and posters: https://www.isba8.de/programme/scientific-programme/ (click on session planner).

Panel: Population structure and migration I
10:15–10:30 O–PSM–06 tba
David Reich (Boston, MA/US)

Hmm...

Chad said...

https://programm.conventus.de/index.php?no_cache=1&id=isba2018&tx_coprogramm_programm%5Bday%5D=1&tx_coprogramm_programm%5Bfachgesellschaft%5D=0&tx_coprogramm_programm%5Baction%5D=list&tx_coprogramm_programm%5Bcontroller%5D=Source

Davidski said...

Can't see the abstracts at that link. Are they available or does that website only list the titles?

andrew said...

"The dominance of R1b-M269 in Basques can be explained by a relatively recent founder effect. Also, Basques do show a significant level of steppe ancestry, comparable to most Indo-European-speaking Iberians, so it seems that they're just a linguistic anomaly."

There is a pretty distinct geographic divide on the Steppe between R1b populations (generally in the South) and R1a (generally further North) although it isn't quite complete. All of the really definitively established linguistically Indo-European populations are R1a. And, as the title of this post notes, geographic divides often correspond to linguistic divides.

The case for the Yamnaya R1b people as a source of Indo-European people is far less conclusive. While R1b is common in parts of Europe that subsequently became Celtic or Germanic linguistically, the time depth of the Celtic language family is almost surely Iron Age and not early Bronze Age as it would have to be is the Celtic languages arrived with the R1b migrants from the Southern Steppe (prior to their replacement in the same area by R1a people). So, it is entirely possible that the Yamnaya R1b people and the mostly non-Iberian Bell Beaker people who are genetically related by NRY-DNA shared a language and that this language was not an Indo-European one, which was later replaced (mostly by elite dominance rather than by demic migration) with the Indo-European languages more than a thousand years later, except in the Basque area that held out against the Indo-European language.

Among the factors that support this hypothesis are the purported Vasconic substrate in the Bell Beaker area of toponyms, the fact that a lot of the words borrowed by Basque from Indo-European are from Proto-Indo-European (presumably spoken by the R1a people adjacent to the Yamnaya) rather than from Celtic or italic or Germanic or Aegean IE languages, the existence of some possible borrowing in Basque from Afro-Asiatic languages, and the historical fact that Basque definitely had a greater geographic range in the earliest attested historic period.

None of this is really definitive, and the possibility that the languages spoken by the R1b people of the Steppe and the R1a people of the Steppe were at least in the same language family even if "PIE" was mostly an R1a language is really definitively. For example, it is certainly possible that while the Celtic languages have a common origin in the Iron Age, that they replaced another IE language family associated with the Yamnaya and the non-Iberian Bell Beaker people. In that scenario, you have to imagine that the proto-Basque men adopted the language of their local wives contrary to the pattern seen almost everywhere else that a male dominated group with a shared patrilineage largely replaces local men.

We may never have sufficient information to adequately discriminate between the two possibilities (or the true answer may be something else entirely).

Davidski said...

@andrew

There's both R1a and R1b in pre-Yamnaya Khvalynsk graves in Samara. Also, on average, Corded Ware can be modeled as around 70% Yamnaya, and this is a signal reflected in shared genome-wide and mtDNA haplotypes.

Early Corded Ware graves also contain hammer headed pins which are found in Yamnaya and Catacomb graves north of the Caucasus.

So it's difficult to argue that there was a genetic barrier between these populations, especially since we haven't yet seen any Y-chromosome data from Yamnaya remains from north of the Black Sea.

Unknown128 said...

Davidski thank you very much for the answer!

Well if a chariot riding elite conquered western Europe around 1500 BC, there must be serious archeological evidence for that.

"But they do pop up in the Bronze Age Balkans and Central Asia at the right time to be the source of Indo-European languages there."

Does that mean that steppe ancestry only appears in Greece and Anatolia around 1500 BC? If not, were there several migrations out of the steppe, that created different cultures?
Also why is there so little indication of horse use or riding in the corded ware culture?

Davidski said...

@Unknown128

I don't think chariot riding elites conquered Western Europe. There's no evidence for this anywhere, unless I missed it, and it's in Drews' book. But there's nothing like that in the ancient DNA to date. It seems to me that chariots just spread to Western Europe probably because they were a very useful technology.

And yes, there were many migrations from the steppes during the Bronze Age in all directions, and they created many different cultures.

But no, I don't think that the Corded Ware people used horses for riding. They probably just used them for pulling wagons and for food. Horse riding only became a widespread phenomenon during the Iron Age.

Samuel Andrews said...

Andrew, the young age of Celtic does not refute the ie status of R1b Beaker folk. BTW I think beaker may have been Celtic. Even if they weren't Celtic is a decendant of beaker speech which replaced it's sister ie languages also defended from beaker speech.

Look at BaltoSlavic. It is not much older than Celtic. Corded ware didn't speak BaltoSlavic. But we can be pretty sure they spoke ie and that Baltoslavic is their only defendant left.

R1a Corded ware is considered IE. East/west Slavs carry lots of r1a. Slavic emerged thousands of years after Corded ware. It is younger than celtic. This doesn't refute the

Unknown128 said...

"But no, I don't think that the Corded Ware people used horses for riding. They probably just used them for pulling wagons and for food. Horse riding only became a widespread phenomenon during the Iron Age."

So the Yamnaya didnt either? Which means that Anthony is wrong about that?

Can an IE expansion model be maintained without some rudimentary horse-riding?
Anthonys model is basically that the IE didnt use them to ride into battle, but rather for dragoon like tactics. Use them to move fast, then fight on foot and retreat on horseback.

If the proto IE only used horses to pull wagons, why does the horse play such a key role in all IE myths? Also why does Yamnaya ritual seem so horse centered?

Davidski said...

@Unknown128

So the Yamnaya didnt either? Which means that Anthony is wrong about that?

Not sure on both counts. But even if Yamnaya people didn't use horses for riding or even for mobility of any kind, they sure got around, which is what really counts.

Can an IE expansion model be maintained without some rudimentary horse-riding?

I don't see why not? IE seems to have expanded thanks to the mobility of its speakers, and horse riding isn't crucial for that.

If the proto IE only used horses to pull wagons, why does the horse play such a key role in all IE myths? Also why does Yamnaya ritual seem so horse centered?

Probably because horses were very useful, even if they weren't ridden.

The awesome thing about ancient DNA is that it shows that Yamnaya people and their close relatives expanded out of the steppe in a big way. There's no doubt that they did. So now we have to work out how they managed to do it.

Grey said...

Unknown128 said...
"Can an IE expansion model be maintained without some rudimentary horse-riding?"

depends which part - due to the flat terrain an initial IE expansion over the steppe might only have needed sleds or travois

http://www.native-languages.org/travois.htm

Unknown128 said...

Are there other indications, that the Corded ware are closely related to the Yamnaya?
I mean besides the hammer headed pins?
I did read that the Neolithic hunter gatherers were rather peaceful, sedentary and matriarchal/egalitarian. While the Corded ware seem to have been more mobile, hierarchical, patriarchal and warlike? Such a radical change would strongly confirm the thesis of a steppe migration.

Samuel Andrews said...

Dun Dun Dun. Pretty sure these means EEF decends from Anatolian hunter gatherers.

The first epipaleolithic genome from Anatolia suggests a limited role of demic diffusion in the development of farming in Anatolia
Michal Feldman (Jena/DE)

Davidski said...

@Unknown128

Are there other indications, that the Corded ware are closely related to the Yamnaya?
I mean besides the hammer headed pins?


Genome-wide and mtDNA haplotypes.

Apart from that, just wait for the Yamnaya Y-DNA from Ukraine.

epoch said...

@David

Massively offtopic, but could you do:

Mbuti Natufian GoyetQ116-1 Kostenki14?

Richard Rocca said...

Davidski said... But no, I don't think that the Corded Ware people used horses for riding. They probably just used them for pulling wagons and for food. Horse riding only became a widespread phenomenon during the Iron Age.

David, this is from a detailed paper on Bell Beaker sample I6581 from Kornice, Poland which was R-P312 > U152 > L2:

I6581/HB0031, feature 1561/13: 2456-2146 calBCE (3825±35 BP, Poz-66185). The
burial contained remains of an adult male (30-35 years old at the death). The deceased
was positioned on his back, with legs bent at the knees at a sharp angle and strongly
bent arms with hands placed on the shoulders. The grave goods comprised three vessels
(an ornamented four-footed bowl decorated on the rim and two cups). The bowl
contained poorly preserved animal bones, most likely the remains of food offerings.
Multiple palaeopathologies were identified on these skeletal remains. Some lesions may
be evidence of episodes of violence or other circumstances resulting in head injury. The
high degree of teeth wear can be caused by frequent clenching and “grinding” affected
by using them in a tool-like manner or bruxism resulted by chronic stress. Other traits
identified on lower limb bones indicate that the individual most frequently assumed in
sitting position, with his thighs and shanks in one/almost one plane. Poirier’s facet,
often observed in horse riders, is evident.


If Bell Beaker already had horse riders (even in small numbers), you can bet that Corded Ware did as well. The same sample above had bone formation suggestive of archery use.

Wastrel said...

Yes, all else being equal languages tend to match genes. But there are also a great many exceptions to that guideline, so it cannot respectably be used to rule out theories.

[Even just in Europe alone, we've got cases where the language spread but the genes didn't, with I in Scandinavia and the Balkans, J in Turkey and parts of Greece, E in parts of Greece and the Balkans, the various Is and Rs of the Sami (Uralic-speaking), and the R and I of the Hungarians. We've also got what's probably a much rarer example of genes spreading without language, in the Basques - and some parts of the Baltic (genetically but not linguistically Uralicised). And that's before we get down into subgroups, like Germanic England being dominated by Italo-Celtic R1b.]

K33 said...

"I have a question for you people. There is an opinion, mostly expressed by Robert Drews, (most fully in his "Militarism and the Indo-Europeanizing of Europe"), that the "steppe migrations" of 3000-2000BC into western Europe have nothing to do with the Indo Europeans who were a separate group, that spread extremely fast all around Eurasia with the use of chariots, but since its rule was mostly elite dominance, it did not leave much (if any) genetic trace. "

@Unknown128:

It should be noted that Drews in his earlier seminal work (1988's "The Coming of the Greeks") declared himself as maybe the most well-known/cited advocate of the "Armenian Urheimat" for proto-Indo-European, arguing for an elite dispersal led by LBA chariot riders and not demic diffusion.

In the recent (2017) book you mention, I understand he recants from the Armenian Urheimat theory in an apparent attempt to roll with the genetic data. He now admits that chariotry began on the steppe (which he doubted in Coming of the Greeks, believing the BA Armenian chariot finds to be the oldest). Now, however, Drews has proposed an "Indo-Hittite" Urheimat somewhere in Anatolia "at least to the beginning of the Neolithic period" (!), involving autochthonous development of Anatolian languages, with the nuclear branches of PIE just being one particularly successful "late" offshoot of Indo-Hittite to the steppe. He maintains that nuclear PIE expanded from the steppe only in the Late Bronze Age with elite dominance/chariot warfare.

He also states, incredibly, that the population of Old Europe yielded to the Bronze Age steppe nomads without even giving battle, arguing that urban farming peoples were unfamiliar with warfare as we know it (!!!), and that pitched battles of any sort only really began with chariot warfare (!). His "evidence" for this comes from correspondences in Hammurabi's time indicating that the [urban] Babylonians seemed to rely exclusively on semi-nomadic or nomadic barbarians like Amorites or Gutians as troops, and these warriors were irregular skirmishers. One could just as easily cite late Imperial Rome's reliance on Germanic tribesmen, or the late Caliphate's reliance on the Turks, to "prove" that these empires were unfamiliar with war. Obviously, all it "proves" is that late-stage empires grow decrepit and suffer from failing combat power.

Drews is trying to save face with these recent modifications to his theory; as a matter of fact his old theory of elite diffusion of PIE from the Armenian Urheimat is more defensible than this "Anatolian Neolithic" theory he has put forth. And frankly, his concept of pitched battle as some novel invention of pastoralist charioteers borders on ludicrous.

Synome said...

With regards to figuring out why the Yamnaya were so successful in their expansion, we have a few tantalizing clues and theories.

For one, a large part of their expansion may have had to do with their ability to exploit environments that were previously uninhabitable by Neolithic peoples. The Eurasian steppe is a huge area. Whoever figured out how to live on the vast grasslands in high densities first would probably have seen a massive population expansion until nearly the whole Eurasian steppe reached carrying capacity. And that's indeed what we saw with Yamnaya and their descendant cultures. Nomadic herding with horse drawn wagon homes is likely what made this possible. So they may have achieved a large population surplus in a comparatively short time period.

Another advantage they may have had over neighboring Neolithic populations is hinted at in the recent paper on patrilineal kin networks and neolithic warfare. The Yamnaya may have developed a cultural system, aided by their new pastoralist lifestyle, which allowed them to recruit very large numbers of related men into these patrilineal networks. This may have given them a crucial advantage in competing with other male patrilineal groups who couldn't organize on the same scale.

These are speculations, nothing is settled yet. But I suspect these two factors played a big role in the Yamnaya expansion.

epoch said...

@Synome

Also, inheritances work differently in agricultural societies than in highly tribal pastoral societies. The former, if a dad has only daughters, the son-in-law inherits the plots of lands. The latter, a dad with only daughters would marry them off in the nephews from fathers side, just as Arab bedouins do.

These differences will spark completely different demographic consequences.

Unknown128 said...

Thanks a lot for the answers, especially to you Lenny Dykstra!

I know that the Corded ware and the Yanmaya share many genes. My question was poorly phrased and was directed at the possibility of any similarities discovered by archeology.

Drews theory seems...strange to put it mildly.

Unknown128 said...

"He also states, incredibly, that the population of Old Europe yielded to the Bronze Age steppe nomads without even giving battle, arguing that urban farming peoples were unfamiliar with warfare as we know it"

Problem is that the Corded ware and bell beaker europeans were far from peaceful farmers.

K33 said...

It's almost like a game of rock-paper-scissors: when cultural frontiers clash, farmers almost always beat hunters because they can feed so many more mouths. But pastoralists always beat farmers, because they don't face quite as crippling of a numerical advantage, and can field as much if not more warriors. Every single man in a pastoralist society is natural horse rider and a marksman with the bow; this translates directly to warfare. The Indo-Europeans and the Turks are obvious examples, but so are the Comanche.

In the late-17th century, the Comanche were a relatively minor tribe living in the scrub land around Wyoming. Then they acquired horses after the Pueblo Revolt. The Pueblo only ate horses. Other tribes learned the value of riding them. But only the Comanche totally adopted the horse-centric way of life comparable to Scythians or Turks; they became master breeders and trainers of horses, used them as primary currency etc.

Within 150 years, the Comanche had expanded to become the most powerful tribe in all North America... and it was all due to their expert use of the horse. And they didn't even have the additional advantages of dairying and livestock...

Matt said...

@Sam, that's quite an interesting title you identified "Limited demic diffusion", esp. in context of this thread.

It looks like there are obviously some populations who, despite probably not being the earliest Neolithic pioneers, managed to adopt the innovations without a lot of admixture and then spread heavily.

Like the NW Anatolians if we're interpreting that title correctly, and, if the thinking here in this post more generally is correct, the populations of the Ciscaucasian steppe (or more specifically the North Caucasian piedmont steppe, depending on what the future results show), who were much more Euro HG heavy.

It looks a bit like there's a good amount of survival in places fairly close to the Near Eastern hearth of agriculture, and then a lesser but still good amount of survival quite far away from there, around the Baltic particularly, but intermediate populations maybe were cut out of the loop as expansion of peoples outpaced how easy the technology was to transfer?

(Also the population at Botai, who seem to have been able to independently domestic horses, but I would guess, because they lacked the whole package of innovations allowing both wagon using Steppe EMBA and the later more horseriding focused cultures to exploit the steppe, did not expand as much or leave much of a trace.)

Samuel Andrews said...

@Matt,

Yes, can't wait to see the results. Anatolia might be a home away from home for southern Europeans if 50-60% of their ancestors lived there for 4,000+ years before arriving in Europe. Still waiting for DNA evidence Yamnaya's southern ancestors were farmers. Understandble, we assumed they were at the beginning. Now it appears their southern ancestors were hunter gatherers.

Synome said...

@Matt

I imagine that the people in the Middle eastern hearth were in close enough contact with each other to share innovations as they developed gradually without encountering fully fledged neolithic colonizers as more distant peoples did. The more distant peoples who did adopt agriculture without much admixture often had geo-climatic barriers that allowed them to adopt farming through cultural exchange rather than assimilation.

I think the latter condition applies to both the Baltic (cold, heavily forested) and North Caucasus Piedmont (Mountains, Steppe)

Davidski said...

@epoch

Mbuti Natufian GoyetQ116-1 Kostenki14 -0.0100 -1.541 386565

Matt said...

@Synome, yeah quite likely ideas it seems. Though I was thinking it might be something more to do with population density, where farming waves run into a dense population, the innovations spread quicker than the population growth by colonists and the demic advance stops. Presuming a particular high density of HG population in an area. Though a challenging climate for domestic crops or animals should help with that.

Like maybe the Villabruna related HG population was generally too thin on the ground, and the Anatolian colonists could go around them / avoid them generally. Thinking about Lepenski Vir and the dense population of Euro HG there and then that caused quite a lot of hybridisation that was eventually diluted - if that had been along a "choke point" on the route of colonization, maybe Europe today would be much richer in Euro HG ancestry, if you see what I mean.

There are some ideas at the moment that dense HG populations preceded agriculture and caused adoption of agriculture (experimentation) so poss before agriculture spread Middle Eastern HG populations and HG populations in Ciscaucasia had already shared culture that led to larger population density from changes to HG strategies, and so were in a better position to not be overwhelmed demographically by earliest agriculturalists.

Davidski said...

@All

Please note that I won't publish comments here by people who choose to use IDs such as "Anonymous" or "Unknown" without any other identifiers.

Arza said...

Archaeology, open access
https://www.nature.com/articles/s41467-018-04773-w

Early Neolithic executions indicated by clustered cranial trauma in the mass grave of Halberstadt

Abstract

The later phase of the Central European Early Neolithic witnessed a rise in collective lethal violence to a level undocumented up to this date. This is evidenced by repeated massacres of settled communities of the Linearbandkeramik (ca. 5600–4900 cal bc), the first full farming culture in this area. Skeletal remains of several dozen victims of this prehistoric warfare are known from different sites in Germany and Austria. Here we show that the mass grave of Halberstadt, Germany, a new mass fatality site from the same period, reveals further and so far unknown facets of Early Neolithic collective lethal violence. A highly selected, almost exclusively adult male and non-local population sample was killed by targeted blows to the back of the head, indicating a practice of systematic execution under largely controlled conditions followed by careless disposal of the bodies. This discovery significantly increases current knowledge about warfare-related violent behaviour in Early Neolithic Central Europe.

Richard Rocca said...

Now published: The first horse herders and the impact of early Bronze Age steppe expansions into Asia

http://science.sciencemag.org/content/360/6396/eaar7711/tab-figures-data

Ric Hern said...

@ Richard Rocca

Very interesting. So it looks like there were at least some EHG in Anatolia during the Chalcolithic, Late Bronze Age and Iron Age. The interesting bit is the Yellow in Anatolia during the Iron Age. Was this a migration of Botai descendants maybe ? Or when Central Asian R1a groups entered Anatolia ?

If they were Botai related could it be that Botai people spoke a Sister Branch Language of Indo-European ? Could it be that they made an early home in the Shirvan Steppe and from there later spread into Anatolia ?

Ric Hern said...

Horses are less phlegmatic than cattle. I can not see how horses could be herded without the use of a horse especially in a plains environment. Once tried it with the help of some of my cousins. It is bloody frustrating even within barbed wire enclosures.

Hunter Gatherers usually have a special place in their hearts for the type of animal that they hunt the most. For example the Khoi-San people and the Eland. Mesolithic people and the Deer or even Fisherman of the Mesolithic Danube River.

We see within the bonepiles of the Steppe a lot of Horses. Horses can cut grass very short due to the presence of frontal teeth. They most probably survived longer on the Steppe than most other large game. That is why we see Horse Dairies and not much Dairy Cattle in a Steppe like environment today...

So horses did not need to be ridden to establish a special place in Indo-Europeans hearts. Hehe.

Dmitry said...

This is "Unknown". SO I wanted to ask what the archeological similarities between Yamnaya and Corded ware were? If the one descends from the other, there must be many such similarities me thinks.

Folker said...

To be back to the original question.

Why they are trying to keep open an origin South of Caucasus? Because some colleagues are defending it? Whatever, Wang et al. are using different results to back their hypothesis (or the "possibility"):
- a shared Iran_N/CHG autosomal admixture between the Steppe Cluster and the Caucasus Cluster. Yes, both have some Iran_N/CHG related admixture. But if you look more closely at their results, the population source is different. Not only Caucasus Eneolithic could be modelized as an one-stream admixture (with qpWave) which is not very coherent with a recent admixture, but the CHG-like component in Steppe Eneolithic has diverged from Kotias thousands of years before, which it doesn't seem the case for Caucasus Eneolithic. And the admixture in the Steppe Cluster can't be from Caucasus Eneolithic, since no Anatolian_N related admixture is found.
- a similar distribution of mitochondrial haplogroups between the Steppe Cluster and the Caucasus Cluster. This not really the case, even at first glance, but if you look more precisely, and if we look only at an early stage (not after EBA), the Caucasus samples are: 1 H6, 1 HV1, 1 K1, 1 K3, 1 I5, 5 R1a, 2 T1, 2 T2 (1 a 1 c), 2 U1b, 6 U4 (3 U4c1 3 U4a2), 1 U5, 2 X2
(in which Dolmen are H6, T1, U2). In their Steppe Cluster, Maykop Steppe are: 1 T2e, 1 H2a1, 1 X1'2'3?, 2 U7b, 1 I5b; and Yamnaya Caucasus and Eneolithic Steppe are: 1 I3a, 1 H2, 2 T2a1, 3 U5a1. As we have other samples available from their supplement (with Khvalynsk and Samara HG): 3 U4a, 7 U5 , 1 T1, 3 T2, 1 R0a1, 1 W3, 1 W6, 1 H1, 2 H2, 1 H6. So now, what are the similarities? The Yamnaya R0a1 is the Ozera sample, probably the daughter of a Caucasian female migrant. The highest numbers in the Caucasus cluster are R1a and U4, Yamnaya are U5 and T2. Note that the U4a subclade found in Yamnaya is the same in Khvalynsk and Ukraine Neolithic. Overall, it's easy to see that similarities are limited, and are probably explained by the CHG-like ancestry in the Steppe, and to a lesser level, to the shared Anatolian_N ancestry.
- they have also agglomered Maykop Steppe and Yamnaya in the same cluster, which can be logical if you look at autosomal DNA or Y DNA, but not so for mitochondrial DNA, as it is highlighting potential similarities with the Caucasus due to specificities of Maykop Steppe.

All in all, their results don't show a clear admixture event from the Caucasus between the mid-Vth millenium and EBA. Quite the contrary.



Unknown said...

Re: the Bell Beaker language question -Italo-Celtic or Vasconic. Celtic proper does look like Iron Age given the degree of relatedness between branches, so going with BB as Italo-Celtic, that gives us 1700 odd years to get from them to the Iron Age dispersion of Celtic proper.

Any linguists know if that's a plausible amount of time for: Italo-Celtic > Proto-Celtic > Celtic proper ? Scenario being that other Italo-Celtic type languages in Britain, Ireland got replaced (cultural, not migration) by a similar new language coming from Central Europe.

Just think that BB being Proto-Celtic is one or two stages too early.

Onur Dincer said...

@Ric Hern

So it looks like there were at least some EHG in Anatolia during the Chalcolithic, Late Bronze Age and Iron Age.

Where do you see EHG component in the pre-Iron Age Anatolian samples above statistically insignificant levels?

The interesting bit is the Yellow in Anatolia during the Iron Age. Was this a migration of Botai descendants maybe ? Or when Central Asian R1a groups entered Anatolia ?

If they were Botai related could it be that Botai people spoke a Sister Branch Language of Indo-European ? Could it be that they made an early home in the Shirvan Steppe and from there later spread into Anatolia ?


More likely, that sample is from a Turkish-era skeleton that collapsed into the Iron Age layer. Carbon-14 dating of the tested skeleton would settle the matter.

Davidski said...

@Dmitry

So I wanted to ask what the archaeological similarities between Yamnaya and Corded ware were? If the one descends from the other, there must be many such similarities me thinks.

Diet and Mobility in the Corded Ware of Central Europe

Davidski said...

@Folker

All in all, their results don't show a clear admixture event from the Caucasus between the mid-Vth millenium and EBA. Quite the contrary.

Good post.

Any normal person would have to suspend their disbelief to read the conclusion in Wang et al. and say, yeah, makes perfect sense.

Twilight Zone material.

Ric Hern said...

@ Onur Dinner

I didn't say significant but nonetheless EHG was there...I would like to know where all the samples were taken from. My guess is more to the South and less to the North and Northeast....

Chad said...

Called that years ago. Really no cultural shifts between 17-10kya.

Onur Dincer said...

@Ric Hern

I didn't say significant but nonetheless EHG was there...

When formal test results indicate that they have no EHG input you should not put much weight on those noise-level ADMIXTURE component values.

I would like to know where all the samples were taken from. My guess is more to the South and less to the North and Northeast....

All those newly-sequenced Anatolian samples are from central Anatolia, specifically from Kırşehir and Nevşehir. Check out the section S1.5 in the supplementary materials for Damgaard et al. 2018 for info on their provenance.

Ric Hern said...

@ Onur Dincer

I still wonder why not anything near Hattusa which is more to the North of those places.

Onur Dincer said...

@Ric Hern

I still wonder why not anything near Hattusa which is more to the North of those places.

Fair enough. It is not clear that the Hittite-era Anatolian samples are from IE speakers. I guess when Hittite-era Anatolian elite burials are sampled we will find some individuals with EHG hence steppe input.

Wastrel said...

Brett: I'm not an academic linguist, just an enthusiast, but no, BB could not have spoken Proto-Italo-Celtic. It's just much too early, and Celtic is much too late - and the changes between PIC and PC are not all that extensive (assuming PIC was a thing at all).

Realistically, BBs (at least late, non-Iberian BBs) would have spoken a language that was either a dialect of a generally intelligible common Indo-European, or a daughter language only slightly differentiated (maybe if they spoke the equivalent of "Spanish", their neighbours spoke the equivalent of "Portuguese" - but even that might be too distinct; maybe we're talking French vs Walloon or Picard).

Was that language ancestral to Celtic? There's just no way to tell. One idea is that the BB language might have been "Northwestern" - the ancestor of Italo-Celtic, Germanic, and possibly Balto-Slavic*. Northwestern not only shares internal structural features (which could easily be shared retentions or areal innovations) but more important also has a significant amount of shared borrowed vocabulary. This suggests that either the whole Northwestern area was previous home to another large language family (the LBK languages) or else the words were borrowed in one location into Northwestern as a family (or at least a very tight areal grouping) that then expanded, and the BB expansions would seem to fit the bill.

It's also possible that it wasn't Northwestern, but a dialect group that yielded Celtic eventually. Or, perhaps more probably, a dialect group that just died out - unfortunately, we have no evidence whatsoever of any of the IE languages spoken west of Celtic before the Celtic expansion. [with the possible exceptions of Pictish and Lusitanian, about which little is known, and which may be Celtic or very close to it]. So there's absolutely no way to really say.



*BS is clearly a transitional dialect group - either northwestern with southeastern influences, or southeastern with northwestern influences, or an intermediate form from the beginning.

Ric Hern said...

I wonder if that Ogam inscription was written by a Gaulish Druid ? I can not help to think about how Latin was used as a written language all over most probably because of trade advantages accompanied by it.

Were people in the Celtic world bilingual for the same purpose...? We certainly see trade over huge distances since the Bronze Age...and yes that doesn't necessarily mean that there was a common trade language all over but it also doesn't exclude the possibility that there could have been...

In such a scenario it can be difficult to pinpoint the origin of a Linguistic group since all we have are inscriptions of the Common trade language or variants of it...

Dmitry said...

Sorry to bother again.

I got told that this article here http://languagelog.ldc.upenn.edu/nll/?p=980
argues that the IE languages spread into Europe long after 3000 BC and that this confirms drews theory. I honestly didnt see it in the text. I got the impression that Don Ringe does say the opposite, arguing against the anatolian hypothesis in favor of a steppe one, that started around 3000 BC (That of Anthony). Maybe I overlooked something? If languages split after at most 1000 years, what does that say for the date of the IE expansion?

At the start of roman conquest there were still non IE languages in Spain (Iberian). Do ancient Iberians show any steppe ancestry?

Ric Hern said...

@ Dmitry

Have a look.

http://umap.openstreetmap.fr/en/map/ancient-human-dna_41837#4/50.33/7.76

Moe Dawry said...

There's only one way to make the new Bell Beaker and older Corded Ware data showing early, substantial steppic ancestry in Europe fit with any "late PIE entry" model linked with the war chariot. And that is to claim that the Yamnaya descendants in the new BB and CW cultures they founded stopped speaking PIE and instead adopted the language(s) of their Neolithic "hosts". When they migrated, then, it was these languages that spread, rather than PIE. Then only later - after 2000BC - does PIE become dominant, when war charioteers from the steppe take over Europe very quickly in a strict elite dominance model, leaving little genetic legacy but a massive linguistic one.

But honestly this is far fetched. It makes much more sense just to concede that PIE was established in the heartland of Europe with the Yamna, and then spread further with their descendants in the BB and CW cultures. There is perhaps room for the later charioteers to introduce different branches of the IE family. Or more interestingly, the arrival of chariot technology opened up opportunities for the established European branches to reorganise themselves through internal conquest, till we get a picture of the European linguistic distributions similar to that of the early historical period.

I think charioteers might have a much greater,even primary, role in the Indian and Iranian contexts.