search this blog

Sunday, March 14, 2021

A comedy of errors


A couple of years ago, the authors of a paper about a group of Iron Age nomads from the site of Shirenzigou, in the eastern Tian Shan, made a mistake. They wrongly assigned two of these nomads to Y-haplogroup R1b-M269.

This faux pas made them believe that the Shirenzigou nomads were closely related to the M269-rich population associated with the Afanasievo culture.

Indeed, since the Afanasievo culture was often credited with the spread of Tocharian languages to the Tarim Basin, these authors, led by Chao Ning, also concluded that the Shirenzigou nomads were potentially the missing link between the Afanasievo culture and the Tocharians (see here).

Moreover, Ning et al. used formal statistics to argue that the Shirenzegou nomads harbored Afanasievo-related genome-wide ancestry, rather than Sintashta-related genome-wide ancestry, despite the fact that the latter ancestry was widespread in the Tian Shan and surrounds during the Bronze and Iron ages. Soon after, another group of authors, led by Chuan-Chao Wang, also went out of their way to link the Shirenzigou nomads to the Afanasievo people with genome-wide DNA using formal statistics (see here).

Interestingly, one of the Shirenzigou nomads belongs to Y-haplogroup R1a-Z93, which is an obvious Sintashta-related lineage. Both Ning et al. and Wang et al. missed this important fact.

They also missed the key fact that the R1b lineage found in the Shirenzigou nomads actually belongs to an Inner Asian subclade, which is only very distantly related to the originally Eastern European R1b-M269.

Now, formal stats are a very useful tool for studying genome-wide ancestry. But they're not infallible, and that's actually something of an understatement. Indeed, if you don't run sanity checks when using formal stats, you're likely to come to some unusual, even arse about face, conclusions. Uniparental markers, like Y-chromosome haplogroups, can provide a robust sanity check when running formal stats on genome-wide data.

One problem with formal stats is that Sintashta-related ancestry often looks very much like Afanasievo-related ancestry when it's mixed with indigenous Central Asian ancestry. Basically, the reason why this happens is that the Central Asian ancestry dampens the Early European Farmer (EEF) signal in the Sintashta-related ancestry.

This is an artifact that once caused scientists at Harvard to believe that Central Asian Scythians and present-day South Asians lacked Sintashta-related ancestry.

Unfortunately, since the publication of the Ning et al. paper, a consensus has emerged in academia that the Shirenzigou nomads are indeed the missing link between the Afanasievo culture and the Tocharians. But, let's be objective and honest here, it's a consensus based on nothing more than a comedy of errors.

On the other hand, me and most of the commentators at this blog have formed opinions about the Shirenzigou nomads that are totally at odds with the academic consensus, that:

- they're a complex mixture of Sintashta-related, indigenous Central Asian and Tibetan-related ancestries, with no clear, unambiguous signal of Afanasievo-related ancestry

- they weren't the speakers of Proto-Tocharian or even related in any specific way to the Tocharians

- they were probably the speakers of a now extinct Indo-Iranian language, and, at least based on geographic proximity, possibly related to the Yuezhi.

Feel free to make up your own mind. But for me, the question of how Tocharian languages ended up in the Tarim Basin remains wide open. I admit though, I'm currently quite partial to the idea floated here by commentator Copper Axe that the Chemurchek culture may have had something to do with it.

See also...

Don't believe everything you read in peer reviewed papers

103 comments:

Rob said...

Aside from a genetic sanity check; there should also be historical ones
It would be odd to have a lack of Andronovo ancestry in the Dzhungar nomads when the region is replete with Andronovo sites; whilst Afansievo sites are missing

I’m not sold in the Chemurchek theory either ; that is basically just a modified Afansievo theory. It requires too many indirect transmissions and doses of faith for language transfer & adoption

Davidski said...

@Rob

Aside from a genetic sanity check; there should also be historical ones

It would be odd to have a lack of Andronovo ancestry in the Dzhungar nomads when the region is replete with Andronovo sites; whilst Afansievo sites are missing


Absolutely.

epoch said...

I think part of the attractiveness of the Shirenzigou theory is that it would link the Yuezhi with proto-Tocharians which was one of the origin theories floating around.

Mallory considers Chemurchek the first step to link Afanasievo with Thocharians. While he also considers a Yuezhi connection but makes it very clear this is disputable.

http://sino-platonic.org/complete/spp259_tocharian_origins.pdf

From what I understand of the history of the area's tribal coalitions were rife and volatile so maybe a number of Indo-Iranian R1a's in Tocharian graves wouldn't be impossible.

Davidski said...

Well, we've got an unambiguous link between the Shirenzigou nomads and Sintashta (Z93). But where's the unambiguous link between the Shirenzigou nomads and Afanasievo?

Some dodgy qpAdm runs don't really cut it for me.

Rob said...

Not to underestimate the importance of what Chemerchuk does show- the heterogeneity of ancestral streams (afansievo; Okunevo, East Asian and BMAC derived) all sharing some cultural features
But it’s a higher weight of evidence to show that the language of Afansievo was paramount. I don’t think there is either a weight of numbers no cultural supremacy or economic revolution to go on; and then this in turn transmitted its language further into the Tarim basin

The problem with tocharians IMO is the view in traditional IE literature that Andronovo is pan-indo-Iranian, and I don’t think that is necessarily the case

Copper Axe said...

Hey Dave thanks for the shout!

Funny you should make this topic because I was just rambling about the Shirenzigou nomads on Anthrogenica.

Sample F004 might have Tocharian ancestry in my opinion. I'm still a noob when it comes to modeling and making ghosts and such but when I removed all the ancestry that is "iron age steppe" related, I more or less ended up with a ghost profile that was 50% Steppe_EMBA and the other 50% being a mixture of Tian Shan pastoralists and bronze age Altaic populations.

F004 takes up 50% ancestry from this ghost population.

If the conventional theory of Afanasievo being the origin of Tocharian languages, a people with 50% steppe_EMBA and 50%, local ancestry sounds like a great candidate.

Plus it is very similar in ancestry to the Chemurchek samples.

https://anthrogenica.com/showthread.php?23274-Early-Tocharian-split-without-European-farmer-admixture

Ric Hern said...

How to connect the Tocharian Language/s with Indo-Iranian to the exclusion of Western Indo-European branches and their similarities to Tocharian I think is the big question...

Ric Hern said...

Was there an Indo-Iranian Language so far East closer to Western Indo-European branches ? How did this come about ? Was Proto-Indo-European then indirectly closer to Western Indo-European branches eg. Germanic and Celtic, than previously thought ?

Ric Hern said...

Or was there a much later migration from Central Europe to Western China ?

Ric Hern said...

Looking at the possibility of how fast Uralic took a foothold in parts of Europe it could be that a very late migration from West to East was responsible for the Tocharian Language.

epoch said...

@Davidski


The Xiaohe graves were predominantly R1a1 if I recall correctly. But they could also be Indo-Aryans and the connection of these Bronze Age graves to later Tocharian speakers isn't at all certain, Mallory clearly states.

Didn't we have an unambiguous Y-DNA link between Afanasievo and Chimurchek?

Davidski said...

@Copper Axe

It's possible that F004 has some Afanasievo-related ancestry.

But no one can show this unambiguously. Not with qpAdm nor even with IBD hits.

The idea that qpAdm can provide such unambiguous evidence in fine scale models for samples with such complex ancestry is very strange and extremely naive.

G25 can't do it either, unfortunately.

Ric Hern said...

@ Davidski


"They also missed the key fact that the R1b lineage found in the Shirenzigou nomads actually belongs a native Central Asian subclade, which is only very distantly related to the originally Eastern European R1b-M269."

How distantly related are these R1b and did their distant R1b Ancestors come from Europe ?

Davidski said...

@epoch

Didn't we have an unambiguous Y-DNA link between Afanasievo and Chimurchek?

Yes, but there's no real link between the Shirenzigou nomads and the Chemurchek people.

So the fact that the Chemurchek people do have Afanasievo-related ancestry doesn't save the Shirenzigou nomads = Tocharians theory.

Copper Axe said...

To clarify I do think that some of these samples, F004 in particular may have ancestry from Tocharians but the Shirenzigou site has fuck all to do with the Tocharian city states. A presence of Tocharian ancestry in some samples doesn't make a Tocharian.

They all probably do have significant ancestry from the Tarim and Dzunghar basins based on the high amount of Aigyrzhal and WSHG ancestry, but Indo-Iranians lived here too and would've absorbed the same substrates.

I wish the dating was a bit more clear because the age estimates cover the period of Yuezhi presence in the region and their subsequent migrations due to Xiongnu and Wusun conflicts. They could also be the Xiao Yuezhi which remained after the Da Yuezhi migrated.

Rob said...

@ Ric

“ Or was there a much later migration from Central Europe to Western China ?”

That hasn’t been ruled out. Which is why I previously joked that tocharians were west european Buddhist pilgrims

Rob said...

But jokes aside; I think tocharians derive from Andronovo

Davidski said...

I don't think it's such an unusual thing to say that Tocharian may have been one of the Andronovo languages.

The Andronovo horizon was massive, and this idea actually fits well with the linguistic theory floated recently in peer reviewed literature that Tocharian was an Indo-European language heavily influenced by Uralic.

Copper Axe said...

It's definitely not that strange, it seems like more people on the interwebs prefer that option than Tocharian being Afanasievo derived.

https://brill.com/view/journals/ieul/7/1/article-p72_3.xml?language=en

It was particularly Samoyedic which influenced pre-Proto-Tocharian according to these authors, with Yeniseian being in the mix as well.

If we're being realistic, this should've occurred in the vicinity of the Altai region. This could fit the northeastern extensions of the Andronovo horizon, a presence from around 2000 bc onwards, but it could also fit Afanasievo and their potential successors.

I'm not sure how to reconcile the Andronovo pathway with the linguistic data reflecting Indo-Iranian loanwords in Tocharian. Uralic languages show a lot of it but when taking away historical loans from Indo-Aryans and Iranics in places such as Khotan, there isn't much of an Indo-Iranian sub/adstrate in Tocharian.

That isn't to say that there isn't anything from ancient Indo-Iranians present in Tocharian languages:

https://www.academia.edu/37724756/Peyrot_2018_Tocharian_B_etswe_mule_and_Eastern_East_Iranian

Norfern-Ostrobothnian said...

Would the substantial WSHG ancestry of Chemurchek imply that they influenced Tocharian linguistically?

epoch said...

@David

This paper suggests that there is no link between Tocharians and the Yuezhi:

https://www.researchgate.net/publication/318897882_The_separate_origins_of_the_Tocharians_and_the_Yuezhi_Results_from_recent_advances_in_archaeology_and_genetics

It has some summary of genetic evidence, mostly mtDNA. It states however, like mallory does, that Xiaohe graves show some possible cultural ties to Chemurchek.

Anonymous said...

In fact, Shirenzigou may have both Yuezhi and Xiongnu who fought there. No one excludes the presence of Toharians there either. Archaeology will not help there, because it cannot really distinguish the Xiongnu, Yuezhi and Tokharian from each other there. The only criterion for belonging to Tokharians is burials from the 5th-8th century AD in the area where their script was found, and until they are tested all discussions about who the Tokharians were are anti-scientific, it is shameful that all sorts of scientists are engaged in these fantasies.

It feels like they actively want to forget about the Tarim mummies and are looking for evidence that the Tarim mummies from Xiaohe were not Tocharian, although it is unequivocally proved that they were neither Indo-Iranians nor belonged to the Andronovo culture, certainly they did not speak the Tocharian language the in Andronovo culture. The curious thing is that they no one tests by autosomal analysis.

Chemurchek is an unknown what, it is quite possible that it is just a late type of Afanasievo burials, transitional to Okunevo. So far, there is no understanding in this matter.

Romulus the I2a L233+ Proto Balto-Slav, layer of Corded Ware Women said...

Great post I agree. I think this image from Jeong 2020 concisely demonstrates no continuity from Afansievo and the later appearnce of Sintashta.

https://ars.els-cdn.com/content/image/1-s2.0-S0092867420313210-gr3.jpg

Tocharian is more of an outlier than a Centum language since it has features of both. The best explanation is that it is Sintashta derived.

epoch said...

There is a very interesting remark in that paper by Mallory:

"While the two languages belonged to the same branch, they were mutually unintelligible, at least as much as modern Germanic or Romance languages, and possessed considerable differences in even their most basic vocabulary (Lane 1966, 222–223)."

Mallory states that this means the languages must have been separated quite a while ago, longer than mostly indicated, and cites Douglas Adams - Not the hitchhiker guy - that we have examples of languages that were isolated for more than 500 years that were mutually intelligible. So Mallory deduces from this that either the Tocharians were in the Basin for a very long time, long enough for these languages to separate in Tocharian A, B and maybe C, or in case that the different languages reflect different settlement events, the origin of Tocharians must be very close to the Tarim Basin.

A Shirenzigou (200 - 100 y BC) origin of Tocharians would therefore not solve the problem of the origin of the Tocharians.

Andrzejewski said...

The question is - what is the link between Tocharians and Tarim Basin Mummies?

Anonymous said...

But I personally still believe that it is the Western Xiongnu that are buried in Shirenzigou, all autosomal and uniparental data point to just that. Autosomally these people are close to the early western Xiongnu and Mongolian Sagly culture, haplogroup-wise they look like Xiongnu and the preceding Mongolian Sagsai culture. While we have no data on Yuezhi this is the most likely result, for Iranians they have little R1a-Z93.

Rob said...

The Andronovo period is the time when the Tarim basin Bronze Age horizon comes into being , Chemerchek is a just an interesting epi-phenomenon but overall inconsequential

Linguistically; Andronovo would have been pre-Satem IE; with Tocharian & other forms of IE developing from it in east inner Asia as they became isolated from the main body of Andronovo tribes.This accounts for the pseudo-NW IE loans in Uralic
Indo-Iranian aubsequently evolved amongst those groups west of the Tian Shan; with later migrant groups bringing indo-Iranian to the Tarim basin like Wusun Yuezhi or whatever the philologist-scrollers come up with as they dust the cobwebs of ancient mentions

Andrzejewski said...

@Norfern “ Would the substantial WSHG ancestry of Chemurchek imply that they influenced Tocharian linguistically?”

Interesting question. I used to think until very lately that Yenisseyan speakers are WSHG until I was corrected there Kett at most score 25% under the best models, and that they are most likely Tianyuan derived (it’s also intriguing that Ainu were though of as something in between Europeans and Asians but they turned out to be overwhelmingly East Eurasian).

So I infer that Botai, Okunevo, Xiongnu, Kelteminar and other WSHG rich populations had NOTHING to do with Native Americans, Yeniseians, Beringians, Nivkh/Ainu or Devil Gate, and instead they were much more Europoid/Caucasian in their genotype, physical appearance, culture, lifestyle and language than we give them credit for.

Therefore bottom line: The possible WSHG substrate in Tocharian is NOT Yenisseyan related.

Anonymous said...

"Linguistically; Andronovo would have been pre-Satem IE; with Tocharian & other forms of IE"

Andronovo is not pre-Satem, neither Balts nor Slavs nor Thracians nor Albanians nor Armenians are from Andronovo, but they are all Satem. Andronovo is just Indo-Iranians, in any case has nothing to do with Tokharians.

Copper Axe said...

@Andrzejewski

The Ket dont need to derive majority of their ancestry from WSHGs to have their language derived from them. Languages dont work that way.

They also have quite a bit of Uralic related ancestry and steppe_mlba even btw.

Yeniseian hydronyms more or less correlate with the Q carrying populations that lived in the Altai-Sayan region during the bronze and I think the most logical solution would to link Yeniseian to populations from there like Glazkovo, Okunevo, Bolshemysh etc. Not those material cultures per se but from populations related to them.

Those Q populations had WSHG-like ancestry in varying rates but they also had significant ancestry from the earlier Neolithic populations in those regions.

Its a debate on its own to say whether these populations spoke the languages of their WSHG side (predominant Y-dna) or the earlier East Asian populations that lived their before and they overtake.

Yeniseians were very patriarchal peoples but there according to linguists such as Edward Vajda there might be a hint of an earlier matriarchy reflected in their mythology.

Not sure what to make of that because all the Q populations seemed like patrilineal and patriarchal peoples. The Glazkovo had a small shift in autosomal ancestry but a big one in Y-dna when compared to the Neolithic populations.

So ultimately I dont think your earlier idea of Ket having a WSHG relation is all that strange actually. Its definitely connected to the bronze and iron age populations there and the Yeniseians show sufficient autosomal ancestry and Y-dna connections to such populations to be positioned as their (final) linguistic descendants.


My suspicion for a while now is that those Altai_MLBA/Deer Stone Khirigsuur samples from Jeong et al have something to do with the linguistic presence of Yeniseian languages in the Xiongnu confederacy and empire. Perhaps not directly them but a closely related population.

Onno Hovers said...

"Satem" versus "Centum" is a bit of a myth. There is no big divide between "Satem" branches that merged the labiovelars and back velars, while palatizing the front velars and "Centum" branches that merged the front and back velars. Anatolian maintained all three velar series. Hittite is "Centum"-like but Luwian is more "Satem"-like. Latin (Italic) is not strictly "Centum", since it has *ḱe -> ce, *ke -> ca and *kʷe -> que. Armenian and Albanian are not strictly "Satem" since they also have remnants of all three series.

Anonymous said...

@Rob
"How can one come up with such absurd claims when the Tarim basin horizon, all the way to northern China, is derived from Andronovo migrants"

This is an anti-scientific lie and nonsense. The Tarim mummies have nothing to do with Andronovo neither archaeologically nor anthropologically nor genetically. They have nothing to do with the migration from Andronovo is a well-known scientific fact. Do not use words you do not understand the meaning of.

Copper Axe said...

@Everyone

Where can I find a sheet with the updated haplogroups of the Shirenzigou samples?

Thanks in advance.

Genos Historia said...

No matter how big Andronovo's range was, genetically R1a Z93 folk were all closely related. It is safe to assume based on that they all spoke the version of Indo European.

So if Tocherian is from Andronovo, it should be a sister branch to Indo Iranian. Not a basal branch of Indo European as lingustits say.

Either Lingustits are wrong about Tocherian being basal. Or we have a hard time explaining how two different IE languages coexisted in R1a Z93 folk from 2600 to 1500 BC.

Andrzejewski said...

@Archi “
This is an anti-scientific lie and nonsense. The Tarim mummies have nothing to do with Andronovo neither archaeologically nor anthropologically nor genetically. They have nothing to do with the migration from Andronovo is a well-known scientific fact. Do not use words you do not understand the meaning of.”

Not according to Hollard 2018.

Ric Hern said...

Maybe Poltavka Culture swept up within the Eastward movement ?

Rob said...

@ Genos


''. Or we have a hard time explaining how two different IE languages coexisted in R1a Z93 folk from 2600 to 1500 BC.''


That's not necessarily the case
The attested version of Tocharian is after the Turn of the Era, that's some 2000 years after its divergence from ''late eastern I.E.'', which would be the parent language of pre-proto-Tocharian & proto-Indo-Iranian + whichever other extinct & unattested Central Asian I.E. once existed.

Tocharian & I-I then diverged sharply because they developed in completely different environments, the former in eastern Inner Asia, the latter over a broad area between Turan & eastern Europe. The 2 idioms met again in the Era of Great (Asian) Migrations, in the wake of Hun, Wusun, Yuezhi dislocations, etc



Genos Historia said...

@Rob,

If that is the case, then linguists are wrong about Tocharian being a basal branch. Isn't what they say, that Tocherian broke away from LPIE right after Hittite did?

I don't know about linguistics. But the fact Tocherian is extinct and the writings in it are not extensive, I am skeptical of any attempt to place it in the IE tree.

Rob said...

@ Genos

Well at the moment, nobody is wrong because we dont have any (peer reviewed) aDNA from the Tarim to help guide our understanding of the interaction between people, culture & language. So any reasonable scenario remains probable

But I am drawn to a 2000 BC for the simple fact that the Tarim basin horizon appears at the same time as the Andronovo culture does so in the neighbouring Altai & Dhungar regions. Also at this time, Eurasian bronzes reach China via the Hexi corridor. Recent excavations have shown that the Andronovo cutlure sites in Xinjiang were settled by mobile people who were often moving off to 'somewhere else' . We know enough about aDNA studies in the past, not to mention the collateral evidence from Mongolia, to assume that these events are all linked.

If this is correct, it doesnt mean ''linguists were wrong', at least not all. Most linguists are very cautious about committing to absolute dates, it is only a handful who do so, and a smaller subset still how make definitive models. Possible shortfalls were (i) their uncritical acceptance of the more popularly circulated archaeological models of IE dispersal, with a resulting assumption circularity between Afansievo must be Tocharian-related because Tocharian must be older (ii) The strict use of family-tree models which sometimes pump out funny results because they compare very different volumes of evidence (extant langauges, extinct but well represented langauegs, and fragmentary languages); & dont consider secondary convergence, intereference, & other linguistic phenomena. So the traditional family tree model works well for obvious nodes (e.g. within Germanic, within Celtic, within Balto-Slavic, but less so for more controversial & hypothetical higher order 'families').
But then again, not all linguists espoused to the the view that Tocharian is the 2nd oldest langauge (e.g. The second one to branch off? The Tocharian lexicon revisited. M Malzahn)

So I think it'll be holitically understood paleoanthropological models which will ultimately polish the IE family tree, byt understanding how their speakers melded into what they are.




Anonymous said...

@Rob
" So if Tocherian is from Andronovo, it should be a sister branch to Indo Iranian. Not a basal branch of Indo European as lingustits say.
Either Lingustits are wrong about Tocherian being basal. Or we have a hard time explaining how two different IE languages coexisted in R1a Z93 folk from 2600 to 1500 BC."

There are no Tocherians, there are Tocharians. The Tarim mummies are not R1a-Z93. So your fantasies about pseudo Tocharians being R1a-Z93 are just your fantasies.

@Andrzejewski
"Not according to Hollard 2018."

You write nonsense and always do.

Anonymous said...

^^@Genos Historia

@Onno Hovers
""Satem" versus "Centum" is a bit of a myth."

It's not a myth. It is a real separation, supported not only by the Satem Law, but also by the RUKI-law, which is found only in Satem IE languages and is found in no other language in the world. This law cannot be borrowed or developed in parallel because it is unique and cannot be explained in any way. The Satem group of languages can be called the RUKI-group of languages.

Of course, no Centum languages are found in India, it is just a myth started by someone.

"There is no big divide between "Satem" branches that merged the labiovelars and back velars, while palatizing the front velars and "Centum" branches that merged the front and back velars. Anatolian maintained all three velar series. Hittite is "Centum"-like but Luwian is more "Satem"-like."

These are palatalizations, not the Centum-Satem division, none of them are the same as Centum or Satem, none of them contain RUKI-law.

"Latin (Italic) is not strictly "Centum", since it has *ḱe -> ce, *ke -> ca and *kʷe -> que."

This medieval palatalization has nothing to do with Satem. You wrote it untrue - Not *ḱe -> ce, *ke -> ca, but *ke = ce -> tse, *ka = ca -> ka.

"Armenian and Albanian are not strictly "Satem" since they also have remnants of all three series."

It is not true it is a mistake.

Desdichado said...

@Rob "But then again, not all linguists espoused to the the view that Tocharian is the 2nd oldest langauge (e.g. The second one to branch off? The Tocharian lexicon revisited. M Malzahn)"

But most have done so. That's clearly the mainstream opinion that most linguists default to, and there are several studies in print that corroborate it, and few that attempt to refute it. The alternatives from the distant past of linguistic study have Tocharians migrating from Western Europe across the entire face of Eurasia without picking up notable linguistic influence from the Indo-Iranian languages that they're crossing through.

The fact that this is the mainstream, commonly accepted, and widely accepted model, and that alternatives are linguistically unlikely and unsatisfying is exactly why the search to link the Tocharians to the Afanasevo has been such a popular pastime for the last several decades.

DragonHermit said...

This Archi person has no idea what they're talking about. It's long been accepted that Balto-Slavic/Germanic are closely related, yet belong to Satem/Centum. Greek for example has been much more closely linked to Armenian, than say Celtic. Centum/Satem does not automatically imply close kinship.

All Centum/Satem is is a classification of how certain consonants are treated, and yes they can develop entirely independently. There are no split branches of IE languages called "Centum" or "Satem". Balto-Slavic languages and Indo-Aryan languages had contact early on, so similar features are no surprise. That does not mean everything split into 2 groups from the get go.

It's long been accepted by scholars that an original Centum/Satem split of IE was a false assumption. Balto-Slavic/Indo-Aryan areal contact does not change that.

https://www.jstor.org/stable/40849463?seq=1

Genos Historia said...

If David Reich-Harvard, makes short sighted mistakes like this we can excuse people at this blog who do the same.

It is short sighted of Reich to not know the Shirenigzou nomad R1b is not R1b M269. Then repeat several times publicly that it is.

Ancient DNA is tricky when the data is limited.

Anonymous said...

@Desdichado
There is a noticeable Indo-Iranian influence on the Tocharian languages, but its interpretation is difficult because of the nature of the Buddhist texts and the complexity of the graphics.

----

And in general it is striking that numerous Tarim mummies from the Iron Age burials of the 5th-3rd centuries BC of Subeisi, Yanghai-3 which are close to the Tocharian settlement area a thousand years later are not yet tested, although they are very much influenced by the Pazyryk culture, so it could be Yuezhi.

Very interesting are the burial grounds of the Iron Age Chauhu culture, which is difficult to attribute to the Iranian and which is accurately located in the Tocharian area.


@DragonHermit

"This Archi person has no idea what they're talking about. It's long been accepted that Balto-Slavic/Germanic are closely related, yet belong to Satem/Centum."


I know perfectly well what I am writing, but you are completely ignorant. There is no Balto-Slavic-Germanic group of languages, linguistics has categorically rejected it, and all the common innovations for these languages are associated with a later areal contact that took place quite late in the Bronze Age as Porzig discovered in the Urnfield culture.

"Balto-Slavic languages and Indo-Aryan languages had contact early on, so similar features are no surprise. That does not mean everything split into 2 groups from the get go."

That is what all modern linguists think of language relatedness. And that is what absolutely everyone thinks.

https://i.ibb.co/3NQh1wX/IE-with-TMRCA.png trees

"All Centum/Satem is is a classification of how certain consonants are treated, and yes they can develop entirely independently. There are no split branches of IE languages called "Centum" or "Satem"."

No, it's all untrue from first to last word.

"It's long been accepted by scholars that an original Centum/Satem split of IE was a false assumption."

This is a disproved lie. All those who thought so are recognized by linguists as deceivers. Centum/Satem split of IE is main split.

Onno Hovers said...

@Archie: The Centum/Satem theory is that *ḱ -> *k was one of the first things to happen to the Centum language, while *kʷ -> *k and *ḱ -> *c was one of the first things to happen in Satem languages. Some like Meillet and Kortlandt even speculated that PIE really only had two velar series.

The significant part in Italic is *ḱe -> ke vs. *ke -> ka, which is described in Schrijver's "The reflexes of the Proto-Indo-European laryngeals in Latin". This means that the "Centum" merger of ḱ and k happened after this sound change. Albanian has *ḱ -> θ; *k -> k and *kʷ[e/i] -> s, but otherwise *kʷ -> k (Vladimir Orel, 2000, "A concise historical grammar of the Albanian language"). Armenian has *g[e/i] -> tʃ but *gʷ[e/i] -> k (Reinhard Stempel, 1994, "Zur Vertretung der drei indogermanischen Gutturalreihen im Armenischen"). Luwian even has *ḱ -> *ts, k -> k and kʷ -> kʷ.

So the sound changes associated with Satem/Centum happened relatively late or not at all in some branches.

Note that RUKI is also not done the same way in all branches. Indo-Iranian applies RUKI also to instances of r that came from an earlier *l. But other branches never applied RUKI to *l (that's why it's not called RULKI) so RUKI must have happened in Indo-Iranian after the change IE *l > II *r. Nuristani doesn't do RUKI when there is a laryngeal in between, unlike the rest of Indo-Iranian. Slavic doesn't do RUKI before a consonant. So there is no special common ancestor to these languages. All RUKI tells you is that these languages were spoken in the same area at one time.

Genos Historia said...

Rob, brings up a good sanity check.

Tocherian writings come from 2,000 years after Andronovo existed. That might be long enough to make its affiliation to Indo-Iranian disappear.

Genos Historia said...

@Davidski,
"This is an artifact that once caused scientists at Harvard to believe that Central Asian Scythians and present-day South Asians lacked Sintashta-related ancestry."

Yes, it is an annoying artifact, which made all of us think Kalash have more Steppe ancestry than they do.

Turns out in total, they only have as much "Yamnaya" ancestry as do modern Greeks.

Leron said...

I'm sure the people who we call Tocharians existed, but can we be completely sure the remnants of the Tocharian language weren't a 20th century hoax?

The field is ripe with frauds and hoaxes; especially started in earnest during the Middle Ages but perfected in the modern era. There's also a lot of political and personal agendas, I don't need to tell you. Back then there was a big controversy regarding the IE centum and satem split. Imagine an incredible archeological discovery in the Tarim Basin, but due to various reasons (primarily the location), the sites were highly restricted but to a few. It wouldn't be impossible for erudite opportunists to create a centum language from scratch, as knowledge of IE was relatively advanced by then. Seed the sites with many pieces of parchment from this "ancient Esperanto" that translated already well known Buddhist liturgy. Then make the case that someone's Western European forefather once thrived as far as China.

The Tocharians, language aside, would simply be an Iranic people. An extension of the Bactrians and Sogdians.

Rob said...

@ Archie

'' The Tarim mummies are not R1a-Z93. So your fantasies about pseudo Tocharians being R1a-Z93 are just your fantasies.''

I did not reference whether they were Z93 or not. It doesn’t even matter a huge deal if it is R1a-Z93 , some other R1a clade, R1b, Q1a etc, what matters is the big picture, which is that this horizon appears in 2000 BC, which synchronizes with the post-Sintashta-Andronovo period appearing in inner Asia.
Moreover, those x Z93 results haven’t been confirmed, and given the issues with Uighurs China has, who knows when or if it’ll happen
So again the issue is you neither understand the big picture nor the fine print


@ Desdichado

None of that actually matters
What matters is the data shows that there is little if any surviving afansievo ancestry in subsequent periods, whilst there is plenty of eastern variant -Andronovo ancestry in the Dzhungar basin

https://imgur.com/ZkKbeT7


Rob said...


@ Genos

If we consider that Sintashta dispersed from 2,200 and Tocharian is attested from ~ 400 CE.
That's 2,600 years.
The notion that Sintashta is not necessarily Indo-Aryan is a bit too complex from some token commentators like ''Cold Mountains'' to understand, but that's irrelevant to the real world.

Anonymous said...

@Onno Hovers

" The significant part in Italic is *ḱe -> ke vs. *ke -> ka, which is described in Schrijver's "The reflexes of the Proto-Indo-European laryngeals in Latin". This means that the "Centum" merger of ḱ and k happened after this sound change."

This is pseudoscientific nonsense. There is no such thing. Schrijver has no counterposition of *ḱe -> ke vs. *ke -> ka, Schrijver examines the laryngals that come after vowels, and these laryngals that you have maliciously ignored give *Ke -> ke & *Keh2 -> ka no matter what K=*k or *ḱ, this is what Schrijver has written.
Further, he gives several examples where other languages have (s)ka- or (s)kR- (zero grade) or (s)koRh2- which give the Latin ca-, none of these cases has ke- in other languages. He gives numerous examples in *ke -> ke and in *ḱe -> ka, so he has no counterposition of *ḱe -> ke vs. *ke -> ka. All these evolutions are of great antiquity.

"Albanian has *ḱ -> θ; *k -> k and *kʷ[e/i] -> s, but otherwise *kʷ -> k (Vladimir Orel, 2000, "A concise historical grammar of the Albanian language")."

About Albanian, Orel was wrong, he has no *ke reflexes, he did not cite them in the book, and they are the same as *kÊ·e, giving the same palatalization before the front vowels *k=*kÊ·. Unfortunately, this book came out before this mistake was pointed out to him.

"Luwian even has *ḱ -> *ts, k -> k and kʷ -> kʷ."

Hittite and Luvian separated from the PIE even before the Centum/Satem division arose. And in Hittite there are Satem reflexes before *ḱu > su, palatalization is not an unusual phenomenon, in Luvian there are different reflexes of all three series of velars.

"Note that RUKI is also not done the same way in all branches. Indo-Iranian applies RUKI also to instances of r that came from an earlier *l. But other branches never applied RUKI to *l (that's why it's not called RULKI) so RUKI must have happened in Indo-Iranian after the change IE *l > II *r. Nuristani doesn't do RUKI when there is a laryngeal in between, unlike the rest of Indo-Iranian. Slavic doesn't do RUKI before a consonant. So there is no special common ancestor to these languages."

You don't understand, RUKI is an automatic reflex. Even in Sanskrit it was a living reflex in the 7th-8th century BC, that is, in any combination of sounds there was a RUKI sound no matter where it came from + s, then s automatically became sh, this was valid even in the outer sandhi between words u s > u sh, and for borrowed words too. Therefore, it naturally worked for the Indo-Iranian transition *l > *r. It is not an instantaneous transition, this state lasted for at least 1500 thousand years, the Indo-Aryan simply could not physically pronounce rs, he just physically always pronounced rsh. It's like how a Russian physically can't pronounce [te], he will always pronounce only [t'e], he just can't pronounce it otherwise, and this state has lasted for over a thousand years.
Therefore, if there is any change in the action of the RUKI law in some language, it instantly applies to absolutely all words where this combination of sounds stands.

@Leron
"It wouldn't be impossible for erudite opportunists to create a centum language from scratch, as knowledge of IE was relatively advanced by then. "

Ridiculous. No one would have the imagination, it's not just one language, but three. There's a huge number of things you just can't make up, from graphics to conjugation. No one would have the imagination to come up with such a fancy language and stick strictly to such perversions.

@Rob "...Moreover, those x Z93 results haven’t been confirmed, and given the issues with Uighurs China has, who knows when or if it’ll happen
So again the issue is you neither understand the big picture nor the fine print"

Rob, don't be ridiculous. It's all confirmed and you're the one who can't read.

Copper Axe said...

@Leron

Are you seriously suggesting that 20th century Uyghur market merchants made up an entire new, quite basal, Indo-European language branch to sell forged manuscripts to European archaeologists visiting these region?

And not a single linguist in the century afterwards picked up on the language being a forgery?

"But the similar set of token don't want to accept this reality due to their own issues & would rather censure the truth in their gay little forum .."

What is it with you and your insults all the time?

Copper Axe said...

There is Z283+ in Srubnaya and one of those Kangju-like Xiongnu "Sarmatians" from Jeong et al 2020.

So even if the Xiaohe samples didnt have Z93+ that doesn't mean they do not have a connection to the Z93+ steppe_mlba Indo-Iranian populations considering not all Indo-Iranian R1a-M417 is under Z93.

Anonymous said...

@Onno Hovers

Once again, the existence of Satem and Centum languages has been confirmed by computational linguistics and confirmed by genetics.

See https://i.ibb.co/3NQh1wX/IE-with-TMRCA.png trees

"The Centum/Satem theory is that *ḱ -> *k was one of the first things to happen to the Centum language, while *kʷ -> *k and *ḱ -> *c was one of the first things to happen in Satem languages."

And this is not quite correct, there was not a fusion of rows of sounds, but a fusion of rows of phonemes. For a very long time, the distinction between palatovelar and labiovelar sounds front to the syllabic sonorities persisted across languages. This is because syllabic sonants were pronounced with a vowel vocation that had the same quality as the velar consonant, that is, pronounced KÊ·R̥C KʸR̥C with an extra short u or i before the sonant. This is due to the general mechanics of the pronunciation of syllabic sonants, which accordingly gave them different reflexions in the languages. For example, in Hittite KÊ· and Ku are not distinguished at all, everything is Ku, which, by the way, has given grounds to some scholars to claim that there were no KÊ· and that it appeared in PIE after the separation of Hittite.

Michalis Moriopoulos said...

I can attest that it's not easy to untangle components in Central Asia and Siberia. Vahaduo cannot always easily distinguish ANE-rich components from each other. Trying to figure out which samples have normal ANE versus WSHG versus actual steppe ancestry is a headache, especially when those ancestries are strongly diluted with East Asian ancestry. The Central Steppe EMBA samples like Kumsay and the Kazakh Steppe BA are a good example of complex ancestry sources that might be at play. Look at Okunevo. Do they have Afanasievo ancestry as one might expect for their location? Or are they actually mixed with the WSHG-heavy Central Steppe EMBA type people from further west? It's not always easy to tell what's going on, especially since all of these groups have ANE ancestry anyway.

Desdichado said...

@Rob "None of that actually matters
What matters is the data shows that there is little if any surviving afansievo ancestry in subsequent periods, whilst there is plenty of eastern variant -Andronovo ancestry in the Dzhungar basin"

It matters in the sense that if there is a linguistic entity east of the eastern Indo-Iranian spread, that shows only very late influence from Indo-Iranian but more closely resembles languages far to the West of it in many respects, then it can't have come from the same horizon as the Indo-Iranian radiation, and there's no archaeological reason to suspect a later arrival across the face of the Indo-Iranian radiation. It's not about where the genetics of the Tarim mummies came from, or anybody else in the Medieval period, it's about where did the languages come from, regardless of who spoke them or wrote them down later. The Tocharian languages might not have any genetic signature at all from the time period that the texts are known from, but from a linguistic standpoint, it doesn't work well at all to bury them in the Andronovo horizon or its later variants. There has to have been an earlier spread into the region somehow.

It's possible that it's kind of a just-so story that we're trying to corroborate, but from a linguistic standpoint, there really aren't any good options other than hoping to find a connection between Afanasevo and the Tarim Basin where Tocharian appears. Unproven as the link may be, no other option explains the linguistic situation, given the state of our knowledge right now.

Leron said...

@Archi

Frauds happen. Experts are not beyond distorting facts or making outright hoaxes. In this blog alone there has been many complaints about renowned scientists engaging in manipulation to reach inexplicable conclusions that disagree with actual data.

Potential "Tocharian C" material was made up
https://languagelog.ldc.upenn.edu/nll/?p=44503

Faked Luwian writings by legendary archeologist
https://www.livescience.com/61989-famed-archaeologist-created-fakes.html

Anonymous said...


This https://i.ibb.co/3NQh1wX/IE-with-TMRCA.png shows the genetic connection between the Balto-Slavic and Germanic language groups, which Ringe-Warnow-Taylor showed in 2002 https://c.radikal.ru/c25/1908/56/b82d2c6369ef.png, and Porzig showed in 1953. Гn doing so they clearly showed that the Germanic language belongs to the Germanic-Illiro-Italo-Celtic language group, but in time Tumulus Urnfield Cultures (maybe even a little earlier) got a huge impact from the Balto-Slavic group, but has no connection to the Indo-Iranian language group. Here you can see perfectly how linguistics and genetics converge.

Onno Hovers said...

@Archi: "...Schrijver has no counterposition of *ḱe -> ke vs. *ke -> ka, Schrijver examines the laryngals that come after vowels, and these laryngals that you have maliciously ignored..."

You may want to read the materials I quoted before childishly writing that it is all a lie.

See section "VII. LATIN A OF NON-LARYNGEAL ORIGIN: B. LATIN A AFTER PURE VELARS" (p. 425-435) e.g.
calidus <*kel-
candēre <*kend-
carpere <*kerp-
carrere <*kers-
scabere <*skebh-
scandere <*skend-

CleverPrimate said...

Leron
Are you saying that Tocharian is the linguistic equivalent of Piltdown Man?

Rob said...

RE : Germanic & Balto-Slavic

Archaeogenetics can also clarify this relationship

Population sharing at a basal level: Z282-related CWC, with divergence of Z284-related lineages partaking in Proto-Germanic

End-point differences include
- U106, I2a2a1b2, I1 in Germanic
- P312 related populations via Celtic contact in Germanic
- I2a1b in proto-Slavic

More data from MLBA central Europe will clarify any additional convergences which occurred in the middle periods, over & above from what ensued in late stages (Goths & Slavs, etc)

Anonymous said...

@Onno Hovers
"You may want to read the materials I quoted before childishly writing that it is all a lie.

See section "VII. LATIN A OF NON-LARYNGEAL ORIGIN: B. LATIN A AFTER PURE VELARS" (p. 425-435) e.g.
calidus <*kel-
candēre <*kend-
carpere <*kerp-
carrere <*kers-
scabere <*skebh-"

I described this section to you, learn to read.

"Archi said...
Further, he gives several examples where other languages have (s)ka- or (s)kR̥С- (zero grade) or (s)koRh2- which give the Latin ca-, none of these cases has ke- in other languages." Therefore, the fact that there was *ke is only his assumption. "He gives numerous examples in *ke -> ke and in *ḱe -> ka,
March 17, 2021 at 4:24 AM"

You deliberately omitted the word "probable" before this list and he emphasizes in the text that this is only a hypothesis.

It is you childishly speculating that there is some connection with the process of Satemization, he just does not write anything like that, he just points out that this process is very ancient. It is quite possible that it dates back 5,000 years.

Real reconstructions
calidus = Lith. kalybas, Snskrt. karkii, Gr. keelas < *kolh2-
candēre = W./MBret. cann Gr. kandaros < *kand-
carpere = Lith. kirpti, Snskrt. krpaana, Ir. cirrim, Gr. kroopion < *kr(h3)p-
carrere = Lith. karshti, Lat. kaarsti, OCS. krasta < *kar(H)s-
scabere = Lith. skabiu, skoobti, Goth. skaban, ORuss. skoblu, Gr. skaptoo, skafos < *skabh-

Vladimir said...

David, why did you assign sample RUS_Khaptsagai_IA:KPT005 to IA? Reich Lab dataset anno file, and the calibrated dates for KPT005 are 2192-2030 BC, Y-hap Q1a2a and mtDNA D4j?

Rob said...

@ desdichado

“ Indo-Iranian spread, that shows only very late influence from Indo-Iranian but more closely resembles languages far to the West of it in many respects, then it can't have come from the same horizon as the Indo-Iranian radiation, and there's no archaeological reason to suspect a later arrival across the face of the Indo-Iranian radiation. “

We know that. This could suggest that Indo-Iranian speakers came into contact with the “ tocharians” after they had been separated for 3000 years from the initial IE founding population in Asia. Indo-Iranian & Tocharian subsequently developed in very different regions of Asia and for a while remained isolated from each other


“ but more closely resembles languages far to the West of it in many respects,”

As explained above; this is merely apparent; because centum-like dialects were the original form. Some features which define I-A , Balto-Slavic etc are later developments; which have tended to be over-emphasised in relevance. See Onno’s comments for greater detail


“ It's possible that it's kind of a just-so story that we're trying to corroborate, but from a linguistic standpoint”

no, it's reconciling the evidence as it currently stands instead of conjuring ghost populations

Anonymous said...

@Rob
" As explained above; this is merely apparent; because centum-like dialects were the original form."

Rob said anti-science nonsense as usual. There is no primordiality of Centum, science categorically denies it. The Centum/Satem division happened at the same time, and before this division there were no Centum and Satem, there were three rows of velar consonants in the language. The Hittite-Luwian languages separated from the PIE before this division occurred show reflexes of exactly three rows, because the palatovelar row positionally gives Satem-like reflexes as well.

vAsiSTha said...

The connections of the dzhungar plains and across into china with south central Asia run very deep. Michael frachetti and his student Spengler, who are is most prominent archaeobotanists for the region prove that the oldest wheat grains in China came from south central Asia through the inner asian mountain corridor.

Non local wheat is then found even as offering in the cremated remains found at tasbas Kazakhstan 2600bce, along with Turan related ancestry at Dali. This sort of ancestry is also visible in the chemurchek samples.

The connections between sc asia and across the dzhungar plains into china exist even in the common Era as is clear from the Buddhist literature present written in tokharian. This route is what became the silk road, and on this route somewhere is where you will find the ancestors of the earliest tokharian speakers.

Interestingly, another such randomly placed centum dialect is bangani found in indian Himalayas. So it's between these 2 regions that you need to look for the answer.



Anonymous said...

@vAsiSTha
"Interestingly, another such randomly placed centum dialect is bangani found in indian Himalayas."

That's completely untrue; it is not Centum. This has been proven unequivocally, it is not even clear who started this myth.

Onno Hovers said...

@Archi: You misinterpret the careful writing of scientists. Real scientists don't overstate their case. You don't cite any scientific papers. You're also getting basic IE sound laws wrong: IE CR̥C does not give Latin CaRC. Cl̥C and Cr̥C gives Latin ColC, CorC while Cn̥C and Cm̥C gives Latin CenC, CemC.

I am only explaining the modern scientific consensus here. And that is that there was no such thing as an early Satem/Centum split.

Arza said...

@ Vladimir

You should definitely send an email to the authors. The sample may be contaminated or misdated and they should check it and publish the errata even if the dating is OK, because certainly this is not Afanasievo as they claim in the text:

Distance difference: ( AC - BC ) ↑
A: RUS_Khaptsagai_IA:KPT005
B: RUS_Baikal_BA
C: ↴
-0.24927963 Ukrainian:UKR-1992
-0.24914886 Lithuanian_PA:LTG-1366
-0.24896483 German:German13
-0.24886451 Finnish:HG00350
-0.24883816 Belarusian:Belarusian13
-0.24883260 Lithuanian_PZ:LTG-1079
-0.24879167 VK2020_POL_Sandomierz_VA:VK494
-0.24878539 VK2020_POL_Bodzia_VA:VK156
-0.24871213 HUN_Avar_Szolad:Av2
-0.24869270 Polish:Polish40

Sending an email to Reich is also a good idea. From .anno file:

Method for Determining Date (including any warnings)
Direct (WARNING MISSING LAB CODE, NEED TO CHANGE TO INTCAL20 AFTER REAPPLYING FWRE CORRECTION): IntCal13 with FWRE correction

Desdichado said...

@Rob "As explained above; this is merely apparent; because centum-like dialects were the original form. Some features which define I-A , Balto-Slavic etc are later developments; which have tended to be over-emphasised in relevance. See Onno’s comments for greater detail"

Sure, I'm aware of that, and that's been the model since... oh, I dunno, since at least Mallory's 1989 book, if not considerably earlier. But my point is that there has to have been another source of the language then. If the linguistic changes that make up satemization et al are strongly associated with eastern Corded Ware variants, and more specifically the Andronovo horizon, when it comes to spreading into the east anyway, then the languages that have older, archaic features, like "centumization" had to either cross across the face of the Andronovo horizon from further west, or have already been there before Andronovo-derived cultures moved into the area.

I guess you can call that a ghost population, given that we don't know exactly where they came from, and to date, our ability to connect them via genetics to the Afansievo population has proven fruitless, but it's the best alternative, and the lack of ability to connect them is much more likely related to the general lack of sampling still.

I have to admit, I'm not sure if I'm amused or intrigued by the idea that Tocharian is a fraud language group, but if that were to turn out to be a viable hypothesis, then it would certainly cause us to stop looking for another connection other than Andronovo-derived cultures for the European genetics in western China.

Onno Hovers said...

@Archie about Bangani: "That's completely untrue; it is not Centum. This has been proven unequivocally, it is not even clear who started this myth."

http://www-personal.umich.edu/~pehook/bangani.abbi2.html

vAsiSTha said...

@archi

Yes, unequivocally untrue indeed.

http://www-personal.umich.edu/~pehook/bangani.hock.html

From: "Hans H. Hock" hhhock@ux1.cso.uiuc.edu
Subject: Bangani
To anyone concerned with the question of Bangani,

"Van Driem and Sharma claim that their fieldwork suggests that Zoller's forms are spurious, that some are based on misidentification and others are simply non-existent. In a recent summary of arguments pro and con, Dr. Kevin Tuitte further suggests that Zoller may have fallen victim to fieldwork consultants' tendency to provide evidence that they think may please the investigator. Even a priori, however, the latter suggestion is dubious, since it would be hard to imagine how illiterate villagers would be able to know that words like _dOkrO, lOktO, gOsti_ would please an investigator (to have that knowledge would require more than a superficial understanding of comparative Indo-European linguistics)."

Anonymous said...

@Onno Hovers said...
"You misinterpret the careful writing of scientists. Real scientists don't overstate their case. You don't cite any scientific papers."

That's not true. You give no reference, your reference to Latin is false, there are no such statements that you attribute to it and that you are fantasizing to us here.

I just gave you pictures from scientific sources that claim what I claim, you have not given anything.

https://i.ibb.co/3NQh1wX/IE-with-TMRCA.png
https://c.radikal.ru/c25/1908/56/b82d2c6369ef.png,

Your words are meaningless, first of all, there ka- in all cognates from parallel languages, so Schrijver's unscientific fantasy that there was *ke- in Latin is trashed in a categorical way, it is not accepted by anyone. Further, the syllabic sonant next to the laryngal gives exactly -a-, in that single example the syllabic sonant next to the larigal. But Schrijver is mistaken here in the choice of cognate forms altogether. because Lat. carpere = Gr. karpos, Eng. harvest < *karp-. That is, *ka- here as well, so Schrijver's unscientific fantasy that in Latin there was *ke- is categorically trashed, it is not accepted by anyone.

"You're also getting basic IE sound laws wrong"
Don't teach me, you understand nothing about linguistics. you understand nothing about science.you don't understand that many people publish their thoughts, but many of them are wrong and Schrijver was wrong too.
You accept Schrijver's fantasy, but you do not know that the laryngal in Latin gives -a-, and that the syllabic sonant next to the laryngal also gives a combination with -a-.

"I am only explaining the modern scientific consensus here. And that is that there was no such thing as an early Satem/Centum split."

That's not true. There is only one linguistic consensus, the Centum/Satem division is most ancient and it is common. I know this for a fact. That you pass off your marginal opinion as some kind of consensus, you are writing just a hoax.

Onno Hovers said...

The western ancestry in RUS_Khaptsagai_IA (But ~2000BCE is definitely not Iron Age!) reminds me a bit of the western ancestry in Khanty, Mansi and Nenets. It's low in CHG, but has lots of EHG, EEF and here it even has some WHG. It is not as extreme as those Eastern Uralic peoples and some of the ancestry in North-Central Europe comes close:

Target: RUS_Khaptsagai_IA:KPT005
Distance: 3.9595% / 0.03959541
43.2 RUS_Ust_Belaya
30.0 Baltic_LTU_BA
17.8 POL_EBA
8.0 RUS_Devils_Gate_Cave_N
1.0 RUS_Krasnoyarsk_BA

It's just a single low coverage sample, but it does look a lot like the Andronovans and Afansievans were not the only westerners moving into East-Asia between 3500-2000 BCE.

Rob said...

@ Desdichado

''As explained above; this is merely apparent; because centum-like dialects were the original form. Some features which define I-A , Balto-Slavic etc are later developments; which have tended to be over-emphasised in relevance. See Onno’s comments for greater detail"

Sure, I'm aware of that, and that's been the model since... oh, I dunno, since at least Mallory's 1989 book, if not considerably earlier. But my point is that there has to have been another source of the language then. ''


That is not Mallory's et al model, that is 'my' brainstorm. His model is what you & most other people are familiar with - 'Tocharian' derives from an earlier steppe Migration (Afansievo) because Sintashta-Andronovo must be Indo-Aryan and Tocharian must be older / more divergent and is related to west European languages.

The link between Afansievo in the Yenesi-Minusinsk basin and the Tarim basin is via Chermurchek sites in the southern Altai fringe & Dzungaria.
So the sequence went Afansievo-Okunevo-Chemurchek

As late as 2010, the assumption was of a relatively linear continuity from Afanseivo to Chermurchek & Okunevo (e.g. Chia & Betts)
e.g. ''The Qiemu’erqieke sites have also been linked with the eastward migration of Indo-European speaking peoples, specifically those who brought the ancestral version of the Tocharian language into the Tarim Basin'

But we now know this is not quite the case. Okunevo represents a ~ 90% population replacment in the Minusinsk basin by local Siberian-related groups (Russian physical anthropolgoists had already noted a significant shift). So they are not from the west.

Chemurchek in Mongolia & Dzungaria in turn, is more heterogeneous. Some of the Chemurchek individuals do have Afanasievo ancestry, but others are east Asian, some have BMAC. But the overall link amongst them is WSHG ancestry, especially in those in the southern Altai (On the way to the Basins region). This link is therefore not as surmised, therefore everythign needs to be revisited & further data

''I guess you can call that a ghost population,''

A ghost population in genetic terms is we construct samples which otherwise don't exist or have not been sampled.




@ Michaelis

''The Central Steppe EMBA samples like Kumsay and the Kazakh Steppe BA are a good example of complex ancestry sources that might be at play. Look at Okunevo. Do they have Afanasievo ancestry as one might expect for their location? Or are they actually mixed with the WSHG-heavy Central Steppe EMBA type people from further west? It's not always easy to tell what's going on, especially since all of these groups have ANE ancestry anyway''

I didnt think there was any confusion here - the Kazakh MLBA groups rerpesent the Okunevoans pushing further west. Kumsay are essentially the same population as Steppe Majkop. There were several movements going on, especially in the pre-Andronovo period

Ric Hern said...

With the Plague spreading from the Altai to the West at such an early date it is not totally irrational to think that some kind of connection between West and East remained. Why did some Late Indo-Europeans spread so far West only to make a U-turn and go all the way to the Altai ? What was the push or pull factor ?

Copper Axe said...

@Ric Hern

If you look at the geography of Central Asia, it's no surprise that populations ended up 'all the way there'. Natural resources play a big role in determining where people live.


@Davidski

I tried to DM you on AG but your inbox is full. Could you clear it up a bit? I got some burning questions :)

Ric Hern said...

@ Copper Axe

Yes indeed but it looks so extremely purposeful as if they knew precisely where these resources were. So maybe a long existing trade route since the time Afanasevo moved East.

Davidski said...

@copper Axe

This might take some time.

Copper Axe said...

@Ric Hern

It could be trade routes, or it could be the result of explorers/adventurous types exploring lands further east as groups were entering the Central Asian steppes. They then report back home that if you just keep travelling in a straight line for a few weeks you reach the garden of Eden.

I have never taken a journey by carriage but I imagine that with a whole community, livestock and a reasonable pace you can make the journey from Samara to the Altai in a few months.


So if the entire process of them exploring through Central Asia, learning of the Altai region, then convincing folks back home to migrate to that region happened within a few years, the timeframe would be a bit too short to leave a noticeable archaeological imprint of their presence. Particularly because they were wagon-dwellers and most archaeological presence is determined by burial sites.

That being said it is interesting you didn't have much of a presence throughout Northern Kazakhstan and Siberia.

Indo-Iranian presence in the Altai is basically as old as the Indo-Iranian expansions themselves, and in my opinion this is for sure due to pre-existing knowledge of the region due to trade routes.

@Davidski

No worries man. I already appreciate you taking the time do so anyways, I can wait for a bit.

Heyerdahl said...

@Davidski

I would really like to get my G25 coordinates, and so would a few of my relatives. Is eurogenesblog@gmail.com still the go-to contact email?

Davidski said...

@Heyerdahl

Yes, that's the email. Please get in touch and we can discuss this, but I'm currently not running G25 analyses for personal genomics customers.

Simon_W said...

It's the big flood in eastern Australia, for anyone who hasn't understood yet.

Copper Axe said...


Ancient genomic time transect from the Central Asian Steppe unravels the history of the Scythians

https://advances.sciencemag.org/content/7/13/eabe4414

Oh hell yeah!

Andrzejewski said...

Scythians are fabulous ancient warriors and horseback riders.

The word “Ashkenazi” stems from an aberration of their name.

Ossetians are their only modern descendants, and they live in the breakaway republic of South Ossetia, cultivating their old time religious myths and other cultural traits.

Rob said...

Nice study . Getting closer understanding the complexities behind “Scythians”


Key points from what I understood

1. For the eastern “Scythians”
''As an eastern Eurasian proxy, we chose LBA herders from Khovsgol in northern Mongolia based on their geographic and temporal proximity. Other eastern proxies fail the model because of a lack or an excess of affinity toward the Ancient North Eurasian (ANE) lineage (25). However, this two-way admixture model of Khovsgol + steppe_MLBA does not fully explain the genetic compositions of the Scythian gene pools (data file S4). We find that the missing piece matches well with a small contribution from a source related to ancient populations living in the southern regions of the Caucasus/Iran or Turan

As predicted, however ESHGs probably still pertinent for some groups

2. Moreover the individuals associated with the sedentary Sargat culture in the forest-steppe zone north of the Kazakh Steppe (“Sargat_300BCE”) partially overlap with the Tasmola/Pazyryk cluster although forming a cloud in PCA that is shifted toward western Eurasians and toward the uppermost cline of northern Inner Eurasians (PC1 and PC2, respectively; Fig. 2B). In line with PCA, Sargat individuals carry a small proportion of a different type of northeast Asian ancestry not detected in the nomad groups further to the south (Fig. 2D).

i.e. multiple sources of eastern affinities are implicated
However, the study should emphasise that the eastern shift is simply due to increasing admixture rather than 'turnover' of MLBA steppe groups, (instead the real turnover is, as expected, during the Hunnic-Turk period)

3. ''Sarmatians..cluster shifted toward west Eurasians (Fig. 2 and table S2).''
The Black Sea Scythians would be even more western

The section on modern Kazakhs very good too. Sets in reality cf romanticized notions of them being descendants of ancient Scythians

Vladimir said...

David W. Anthony Presentation on Steppe Genetics & Societal Organization (March 2021)

https://youtu.be/AhlzOj8ouaw

Anonymous said...

Sargat culture is not directly related to the steppe world and Indo-Iranians, it is (Finno-)Ugrian.

Andrzejewski said...

@Rob How’s the Xiaobe/Xiongnu type ancestry admixture from around the time of the common era different genetically from previous East Eurasian populations introgressing into LBA Scythian tribes like Khorgol? The article brings up proportions of ANE ratios but it doesn’t go into details.

2. I suspect that the Devil’s Gates deep ancestry which came to be associated with the East Eurasian component in Beringians and Native Americans is also ancestral to “New Siberian” ethnogenesis that gave rise to Altaic peoples: Tungus (Evenks), Turkic, Ruan Ruan, Mongolians and others. Am I off the mark here?

3. Going off on a tangent here but still speaking of “population turnover” and NorthEast Asia - I noticed that the scattered remnants of Old Siberian populations that used to be extant in the past: Chukchi, Ainu, Nivkh, Para-Innuit, Kamachdals, Yukaghir, Yenisseyan etc tend to have lots of WSHG MA1 type ancestry, although their predominant and ore-eminent one is East Asian and Tyuanyuan-descendant, Han-related one. Some linguists like Vajda also think their language families are somewhat related on a post-LGM macro level. Do you think that their languages are actually West Siberian HG/MA-1 related? And where do prehistoric extinct civilizations like Botai and/or Okunevo fit in this picture/puzzle? Do you believe that genotypically and maybe culture or phenotype wise the latter 2 are more similar to us Europeans/“Caucasians”/“West Eurasians” than we know or give them credit for?

Andrzejewski said...

Speaking of BMAC, they were mentioned in aforementioned article as being mostly of Iranic/related ancestry. However, we already know that they were also WSHG as well as Anatolian Barcin type ancestry. Linguists managed to reconstruct some of their vocabulary. If indeed it’s Iranian Chl based, would you say that it’s related to Elamite? (Latter culture’s artifacts have been found at BMAC sites nevertheless).
On top of it, what’s the validity of the so-called “Elamo-Dravidian theory”? If I’m not mistaken then Dravidians are essentially the descendants of Iran Neolithic (Hotu) admixing with indigenous Onge-like tribes, thereby creating an ASI population, whose language could be Iran based but it could be indigenous? It seems also that the Iranic migrants to India were Iran HG, thus being distinct from Iran Chl Elamites, which were heavily diluted and shifted by intermarriage with incoming Anatolians and Levantines?

I’m also curious to find out who exactly the Nihali, Burusho and Kusunda are; they are relics from populations that had obviously preceded the Andronovo, Dravidians and Munda in the Indian subcontinent. As far as I recall, Narasimhan 2019 debunked any (previously held) notion regarding either the BMAC or the Botai leaving any significant incorporation into Sintashta (save for a few outliers), and hence ruled out any large scale signals of them in the current Indian gene pool.

I just wish we could find the missing links.

Ric Hern said...

Tocharians maybe related to the Tagar Culture ? The Tagar Culture basically overlaps with some of the areas where Samoyedic people live/d. And apparently the Baskirs and Western Tatars seems to be the closest to the Tagar people genetically...

mzp1 said...

@Andre

"then Dravidians are essentially the descendants of Iran Neolithic (Hotu) admixing with indigenous Onge-like tribes, "

South Asians (IA and Dravidian) cannot be modelled as a mixture of ASI and West Eurasian (Iran_N, ANE etc) sources. This is due to the genetic variation within South Asians where some groups are closer to both ASI (ONG) and West Eurasians (Iran_N, Steppe etc) compared to others.

In some comparisons affinity to ASI correlates with affinity to West Eurasian, and the main drift vector in NW South Asia actually pulls away from both ASI and West Eurasian.

Rob said...

@ Andrze

you can see table S4 for the ancestry modelling. But the previous paper on Mongolia by Jeong et al goes into the details of Xiongnu in Mongolia (essentially, it followed a period of Slab-Grave expansion westward at the expence of Mongun-Taiga type groups, then counter-expansion of Sagly-uyuk type ancestry, thus establishing two major 'poles' of diversity; followed by additional Han-Chinese admixture in the mature Empire period; also more 'Sarmatian' + BMAC inflow (which could be, at least in part, rationalised together into southern Sakae))

After the Xiongnu collapse, the next major barbarian power were the Xianbei. I would treat these people as separate , because they are Dong-Hu (eastern barbarians), but there is only a few samples for them at present.

For outside Mongolia, the shift from Scytho-Siberians to Huns needs more data & analysis.

Anonymous said...

Interesting from Anthony 2021 about the relationship between Usatovo and Khvalynsk, Sredny Stog and Yamnaya Russia Don circled on PCAs.


https://i.ibb.co/C6XnZyz/Anthony2021-Usanovo-Khvalynsk-SStog-Don-PCAs.png


Matt said...

Re; PCA in Anthony's recent lecture, yes, full of interest. As I noted on Razib's blog, seems interesting to me that the Yamnaya Hungary, Yamnaya Moldova, Yamnaya Serbia samples, who are displaced towards European farmers are *not* intermediate the position we'd expect to find Romanian or maybe even Hungarian farmer populations and main Yamnaya cluster.

Instead they look intermediate Yamnaya and the main European Corded_Ware cluster for the most part (and thus to some extent to Globular Amphora).

A group of the "star" icons labelled either "PreYamnaya Romania" or "Yamnaya Păuleşti Romania"* are also in between Corded Ware and Globular Amphora. There are also some outlier "Yamnaya Romania" samples who are pretty close to GAC.

This may suggest that Steppe cultures who entered Europe were mixing with particular cultures, whether or not they were the main cultures in a region? (Maybe there is some hetereogenity where some cultures wanted to/were forced by circumstances to marry daughters into cultures of steppe?).

(Obligatory imgur link: https://imgur.com/a/4coC36k)

Interesting that the 1x "early" Sredni-Stog and Don Yamnaya are directly "north" of main Yamnaya and later Sredni-Stog. An early Latvian Corded Ware sample (surely I4629?) overlaps this cluster as well (although v. close to the main Yamnaya cluster).

*Paulesti Romania seems potentially this Paulesti Burial Mound - https://research-information.bris.ac.uk/en/publications/pit-graves-yamnaya-and-kurgans-along-the-lower-danube-disentangli ?

Norfern-Ostrobothnian said...

https://edmond.mpdl.mpg.de/imeji/item/Pz_BOM52KA1ZbcTe?q=&fq=&filter=&pos=0#pageTitle
The genotypes are finally here!

Rob said...

@ Ric
Karasuk, Tagar, ..

Rob said...

Matt, those samples make sense because the eastern half of the carpathian basin had high WHG agropastoralists rather than the older ANF rich type . Eg look at GB from Gonzalez-Fortes
But there were also GAC groups there too

Matt said...

Rob, interesting sample to mention, a shame that it is a single sample (though could well have represented a large and stable community there as you suggest).

Observant gnxp commentor Forgetful noted to me before that as well as the high hg balance, this sample seems to have a fairly high degree of "Balto-Slavic" drift/preference, which is unusual for period - https://m.imgur.com/a/R7v9GvK

Various random other thoughts: Anthony also discusses a Yamnaya sample in East Slovakia with close ancestral relationship with samples from Moldova, the Central Steppe and Afanasievo. So unpublished Yamnaya samples go somewhat wider geographically even than shown in that PCA's legend (as far as Serbia).

We will really need to see the dates of these samples to get a full picture I think. You have highlighted before about how the Malopolska Corded Ware materials, from site that shows R1b-L51 (https://www.nature.com/articles/s41598-020-63138-w, samples pcw250, pcw420, pcw430), post 2500 BCE, show Catacomb Culture influences. Furholt made a suggestion in his recent attempt at a review (saying nothing about the overall quality of this review/model.): "Whereas the new single grave burial ritual associated with Corded Ware (Furholt 2019a) started around 2900 BC, there is no temporal priority for Yamnaya kurgan graves in the Carpathian Basin over early Corded Ware (as it was long supposed). Thus, it now seems that there might be a first, pre-Yamnaya movement of people from the east into central Europe who were buried in the earliest single graves with Corded Ware. During the same time, Yamnaya graves are found in the Carpathian Basin, and at a later stage some individuals from that source mixed with the first Corded Ware generations. This second, more closely Yamanya-associated influx of people became more dominant in central European single graves only after 2500 BC, when it was predominantly connected to Bell Beaker materials.. We will see if there is any chronological sense in this proposal of Furholt.

...

Off topic: https://www.pnas.org/content/118/14/e2025739118 - "Deep genetic affinity between coastal Pacific and Amazonian natives evidenced by Australasian ancestry" - more finding of low level Oceanian related ancestry in South Americans.

gamerz_J said...

@Matt

"Off topic: https://www.pnas.org/content/118/14/e2025739118 - "Deep genetic affinity between coastal Pacific and Amazonian natives evidenced by Australasian ancestry" - more finding of low level Oceanian related ancestry in South Americans."

Thanks for linking this, the authors seem to suggest it's old (I remember Fu's group had a paper that said such a signal was found in Tianyuan man) but don't you think it's more likely to be recent? There was that study that found Native American in a few Pacific islands after all.


@mzp1

"In some comparisons affinity to ASI correlates with affinity to West Eurasian, and the main drift vector in NW South Asia actually pulls away from both ASI and West Eurasian."

Could you please expand on this?

Davidski said...

mzp1 talks a lot of shit, so there's no point asking him to expand on it, for obvious reasons.

archlingo said...

Thanks for the many informations. However, they would be even more informative when accompanied by the ID. Thus, if You mean Shirenzigou M4 being an attestation for Z93, who believes that? Imho it is M417-PF6162-Z645 - no Z93 in the actual ISOGG.
Best,
archaeo