search this blog

Thursday, January 30, 2020

The great and the good


Here's a quote from a new paper on the impact of genetics, and especially ancient DNA, on archeology and linguistics co-authored by archeologist James Mallory and geneticist Oleg Balanovsky:

Just as the genetic evidence for a steppe homeland appeared to weaken a popular theory (among archaeologists more than linguists) that the Indo-European languages spread from an Anatolian homeland with the spread of farming and the AF genetic signature, a new complication arose: the steppe signal that is found from Ireland to the Yenisei comprises an admixture of EHG and CHG. Such an admixture would appear to involve two deep sources that should have developed separately over the course of thousands of years; in short, there is no reason to believe that the two components spoke closely related languages or even belonged to the same language families. Such a model suggested that Proto-Indo-European may have originated out of the merger of two very different language families, a theory that had once had been suggested by several linguists but had never attained anything remotely resembling consensus [62]. If one does not accept an “admixture language” then the natural question remains: did Proto-Indo-European evolve out of language spoken by EHG or out of language spoken by CHG? So genetics has pushed the current homeland debate into several camps: those who seek the homeland either in the southern Caucasus or Iran (CHG) and those who locate it in the steppelands north of the Caucasus and Caspian Sea (EHG). DOI: https://doi.org/10.1134/S1022795419120081

Make no mistake, this is, in common parlance, total horsehit. That's because:

- if we go back far enough, every goddamn human population that ever existed is a mixture of genetically highly diverged earlier populations, but this obviously doesn't mean that all languages are creoles

- in fact, the so called CHG/EHG mixture that Balanovsky and Mallory are talking about was already present on the Pontic-Caspian steppe around 4,300 BCE, and probably much earlier, so it's likely that it first emerged there before the existence of anything even resembling an Indo-European language

- come to think of it, I'm not aware of any tradition in historical linguistics that requires language families to be directly traced back to specific Mesolithic hunter-gatherer populations. So, with all due respect to Mallory and Balanovsky, it looks like they pulled that theory out of their hats.

The impression that I've been getting for a while now is that the great and the good at various major academic institutions are having a rather difficult time interpreting the ancient DNA data relevant to the Indo-European homeland debate. Why? I don't have a clue. Someone should e-mail them and ask. Feel free to let me know what they say in the comments below.

See also...

A final note for the year

A note on Steppe Maykop

Did South Caspian hunter-fishers really migrate to Eastern Europe?

233 comments:

1 – 200 of 233   Newer›   Newest»
Pine trees and mountains said...

So, the situation is pretty clear to me personally, but I am still not 100% certain because there is still not enough proof. Nonetheless I think this is the most probable case:

PIE/IE and Kartvelian have very interesting cognates, like the words for some bird species, soil/earth, pigs/tusks, hills, mountains, clover etc. Thing is that some of these could spread with technology spread, like pigs and clover being related to new forms of pastoralism, but that doesn't explain such core similarities like earth, hills, mountains and bird species endemic to the steppe and the Caucasus. Some people speculate the similarities are from contact with Maykop, which according to this theory was either Kartvelian or contributed a big substrate to Kartvelian languages.

I disagree for several reasons. It is much more likely for Maykop to have been a mix of proto-NWC and the Leyla Tepe related migrants. Now, you do oppose this Davidski but it can't be denied that in isolated areas the lexicon can change minimally even on a span of thousands of years, and the Caucasus and Western Caucasus especially was always very isolated until the significant steppe migrations started with the Catacomb-related groups and later their GAC admixed descendants and Turkics and etc.

Why does NWC not have more cognates with PIE than Kartvelian does is what many ask. It is simple. You look at Darkveti-Meshoko and they are J dominated, if I recall correctly all were J1/J2. Meanwhile, modern NWC are dominated by G2a2, this is the overwhelming majority of their G2, an ENF-related clade of G2a. There are no EEF and ENF with G2a1. Meanwhile Georgian and Ossetian G2a is dominated by G2a1 even though Georgians have a significant minority of G2a2. What does this mean? NWC speak a language of Anatolia_Chalcolithic people. Why not the language of Kaskians? Because by the time of Kaskians Hurro-Urartians had already left a huge amount of J2 and J1 haplogroup in Anatolia and especially Eastern Anatolia, so G2a2 wouldn't have dominated NWC as much if it were spread by Kaskians.

NWC are linguistically Anatolia_C+Maykop. Darkveti were linguistically proto-Kartvelian or perhaps related to Kartvelian, early Maykop were Leyla Tepe who eventually admixed with locals. NEC languages are Leyla Tepe, Iran_C related languages. The similarities between NEC and NWC exist precisely due to the Maykop/Leyla-Tepe common element in both peoples (despite modern NWC having only tiny Iran_N admixture).

Kartvelians meanwhile spoke the language most closely related to the language of the CHG who spread North and admixed with EHG. The similarities of Kartvelian and PIE come from this common element. This common element was wiped out from NWC and NEC during the Chalcolithic and early Bronze Age, which is why Kartvelian is more similar to PIE than NC languages are.

Gaska said...

1-Nobody can date a language and the languages ​​do not arise spontaneously, then nobody can know if IE or PIE or Pre-PIE etc.. were spoken in the Mesolithic or in the Paleolithic

2-Relating unipersonal markers with languages ​​is risky but linking autosomal components with languages ​​is surreal among other things because those autosomal markers are not "pure" in genetic terms either (EHG has part of WHG, Yamnaya also has EEF etc...)

3-The linguistic debate has been polluting and complicating the genetic and archaeological debate for years

4-The only way to find out which language was spoken by any European prehistoric culture is to draw a line of evident genetic continuity between that culture and current populations that speak a certain language. Everything else is speculation

5-The above is impossible with regard to female markers and very complicated with respect to male markers because the vast majority of prehistoric cultures from the V Millennium are absolutely heterogeneous (genetically speaking)

6-Then, the dispute regarding the origin of IE is a sterile fight because it will never be resolved satisfactorily

Davidski said...

@Gaska

Quit talking shit.

There's a consensus that Proto-Indo-European dates to around 4,000 BCE.

Archi said...

This article is purely review, it is devoted to the historiography of the issue and review of issues. The article is generally banal, but in one place introduces a correlation between genetics, archeology and anthropology, it's good. In it, not only the problem of the Indo-Europeans is considered, but also the origin of the Altaians, it is proved that the Altaians come from whole over the territory of Eastern Mongolia and Manchuria in the Early Neolithic, it is shown that they were agricultural farmers.

As for the quote, so it continues with the next quote:

"Those who prefer a southern homeland look to proposed contacts between Indo-European and the Kartvelian and Semitic languages [63–65] while those who prefer the steppe hypothesis point to the evidence that Indo-European is most closely related to Uralic, which should pitch its origin nearer the Urals [54, 66, 67]. The matter is complicated enormously by the fact that CHG cannot possibly represent the signature of a single language family as it is found over a broad area from the Caucasus to the southern Zagros that encompassed the territory of a whole series of other language families,e.g., Hurro-Urartian, Elamite, Kartvelian, and one must devise a credible model of how the Indo-European-speaking segment of CHG found its way north of the Caucasus [68]."

Davidski said...

@Archi

It's a review, a review of nonsense.

All of the arguments they discuss are meaningless, because of the three points I mentioned.

Gaska said...

@Davidski

Really? Not everyone thinks like you or the linguists who have rebuilt that language. It is one thing to find out when the common language was spoken and another thing to establish with absolute certainty its origin. So let me be skeptical about establishing precisely the geographical and temporal origin of any language spoken in the world - Regarding Europe I think that the different IE languages ​​have expanded in recent dates (bronze age and iron age)

Archi said...

@Davidski

In reality, they do not affirm anything at all, they only refer to opinions already expressed. Reacting to this article is pointless. There is nothing that they would not have written earlier, the authors do not say anything except the archae-genetic correlation they found, so it could be discussed.

Once again, the autors only list the questions; they do not even express their opinion.

Davidski said...

LOL

Mallory and Balanovsky also think that CHG lived in Iran.

They must be in the same country club as David Anthony or something.

Archi said...

@Davidski

They know that CHG lived in the Caucasus and Iran. They attribute the Iranian variation of this component to CHG, which is generally true.

Davidski said...

@Archi

CHG didn't live in Iran.

Archi said...

@Davidski

Iran variant of CHG is CHG.

Better post a picture with a correlation between genetics, archeology and anthropology in the topic.

https://i.ibb.co/kXrCfks/image.png

This is the only new thing in this paper.

Drigu said...

@All if you haven't read it yet, here's an interesting paper on the homeland of "early PIE" and the hypothetical Indo-Uralic by Alexander Kozintsev, with some references to Davidski:

https://www.academia.edu/41077803/Proto-Indo-Europeans_The_Prologue

Davidski said...

@Archi

Iran variant of CHG is CHG.

No it's not.

They're separated by thousands of years, and the "Iran variant" has some admix from far to the east. CHG proper, on the other hand, is much more western shifted.

This was obvious ages ago...

On the enigmatic early Neolithic farmers from Iran

Archi said...

@Drigu

We have read. Everything that Kozintsev writes has long been refuted and is of no interest.

Archi said...

@Davidski

These nearer impurities related to the Neolithic do not matter, it is still a cluster of CHG.

Matt said...

Notions of a sudden admixture event at the Mesolithic / Early Neolithic / Copper Age giving rise to a mixed EHG+CHG language do seem at least bit dubious yes.

That said, there's probably enough continuous contact that sprachbund / convergence areas and chains of genetic transmission that don't involve notions of a sudden mixed language can probably relate the languages of North Eurasia in lots of complex ways.

Like, you have models where EHG has CHG ancestry, and then West_Siberia_N had EHG ancestry, and then present day North Eurasian / Siberian ancestry groups (Ngananssans and so on) probably have West_Siberia_N related ancestry (although literally most likely really an ancestor of West_Siberia_N).

So you could have a chain of influence where CHG->EHG (Caucasian->early IE ancestor), followed by EHG->West Siberia (IE ancestor->Uralic ancestor), followed by West Siberia ->NeoSiberian (Uralic ancestor->Altaic etc ancestor), for instance.

It seems a bit shadowy as to whether we'll really totally be able to know.

Davidski said...

@Archi

Nope.

https://drive.google.com/file/d/1aRiUb8efQvjfqfE4Yb9AnxkK5iWMWMyL/view?usp=sharing

https://2.bp.blogspot.com/-7cU7R6UfPa8/XgnaBPMwwBI/AAAAAAAAIpo/uRRkZUZpFm0Pu1qdV1WcyzHGYEd0PnA8QCLcBGAsYHQ/s1600/Ancient_pseudo-haploid_PCA1a.png

Davidski said...

@Matt

Well, the assumption that a mixed ancient genetic heritage should equate to a mixed language, or even a creole, is definitely dubious.

But how dubious do you think is the idea that language families should be traced back directly to Mesolithic genetic isolates?

Archi said...

@Davidski

Your pictures do not prove anything. The difference between them is minimal, such that it is not resolved by most tests, you can see this everywhere on the PCA, ADMIXTURES,... This is really one cluster, that's why some are substituting Iran for the steppes and starting to talk about coming from Neolithic Iran to the steppe. I think no need to remind.

In general, they understand by CHG precisely the common cluster of the Caucasus and Iran.

zardos said...

Without having older samples for the Northern M'lefaatian and the early Lower Don, who knows their exact genetic profile? Thousand years earlier or later, completely different people could have dwelled in the same places.
If Iran Neolithic is Eastern shifted, how old is this shift in the region?

However, there is little to know space to circumvent the R1 dominance it seems. This is the clearest and most striking correlation to IE. Even if not all R1 people might have spoken PIE, never, how another people should have imposed their language on the patriarchal hunter-warrior clans needs a lot of factual help.
The core assumption must be that PPIE was spoken by EHG hunter-fisher societies and matured to PIE under CHG and Neolithic influences. This doesnt qualify it as a create language, because there was no equal contribution. Like German and English have Latin influences (primarily special and extended vocabulary), but are clearly Germanic nevertheless.

TLT said...

>those who seek the homeland either in the southern Caucasus or Iran (CHG)

People are STILL insisting on conflating CHG with Iran Pleistocene-early Holocene populations?

Archi said...

@TLT "People are STILL insisting on conflating CHG with Iran Pleistocene-early Holocene populations?"

No, it means that CHG cluster one, which was deferring already later in the place of residence.

Iran and Caucasus are parts one CHG cluster, it is true in general.

Archi said...



People began to misquote this expression of theirs, here is their expression in full:

"Just as the genetic evidence for a steppe homeland appeared to weaken a popular theory (among archaeologists more than linguists) that the Indo-European languages spread from an Anatolian homeland with the spread of farming and the AF genetic signature, a new complication arose: the steppe signal that is found from Ireland to the Yenisei comprises an admixture of EHG and CHG. Such an admixture would appear to involve two deep sources that should have developed separately over the course of thousands of years; in short, there is no reason to believe that the two components spoke closely related languages or even belonged to the same language families. Such a model suggested that Proto-Indo-European may have originated out of the merger of two very different language families, a theory that had once had been suggested by several linguists but had never attained anything remotely resembling consensus [62]. If one does not accept an “admixture language” then the natural question remains: did Proto-Indo-European evolve out of language spoken by EHG or out of language spoken by CHG? So genetics has pushed the current homeland debate into several camps: those who seek the homeland either in the southern Caucasus or Iran (CHG) and those who locate it in the steppelands north of the Caucasus and Caspian Sea (EHG). Those who prefer a southern homeland look to proposed contacts between Indo-European and the Kartvelian and Semitic languages [63–65] while those who prefer the steppe hypothesis point to the evidence that Indo-European is most closely related to Uralic, which should pitch its origin nearer the Urals [54, 66, 67]. The matter is complicated enormously by the fact that CHG cannot possibly represent the signature of a single language family as it is found over a broad area from the Caucasus to the southern Zagros that encompassed the territory of a whole series of other language families,e.g., Hurro-Urartian, Elamite, Kartvelian, and one must devise a credible model of how the Indo-European-speaking segment of CHG found its way north of the Caucasus [68]."


This is a full paragraph that reflects the essence of the paper, you can not cut the end, because it expresses their position.

TLT said...

@archi
The CHG 'cluster'? Like the Norwegian-Syrian cluster?

"sample": "IRN_Ganj_Dareh_N:Average",
"fit": 18.5352,
"GEO_CHG": 100,
"closestDistances": [
"GEO_CHG:KK1: 18.53523"

"sample": "Syrian:Average",
"fit": 18.9546,
"Norwegian": 100,
"closestDistances": [
"Norwegian:NOR106: 19.02440",
"Norwegian:NOR101: 19.21903",
"Norwegian:NOR107: 19.47051",
"Norwegian:NOR108: 20.01137",
"Norwegian:NOR150: 20.08321",
"Norwegian:NOR152: 20.08957",
"Norwegian:NOR109: 20.17621"


The core common ancestry between CHG and Iran populations probably goes back to the LGM which is a fairly long time ago, long before the Zagros pastoralists and mesolithic CHG existed, long before PIE existed.

Archi said...

@TLT

It is not clear what you mean. Here's what they mean in this case.

https://i.ibb.co/dt8ZGLS/Minoans-Mycenaeans.png

Samuel Andrews said...

@TLT,
Thanks for showing that Distance between CHG and IranN. Experts need to stop repeating the lie CHG and IranN are the same population. The lie keeps being repeating again and again while no one checks DNA stats to see if it is actually true.

The lie started with David Recih's team. They emphasized CHG and IranN's similarities too much. Then, every non-DNA expert trusted what David Recih said whole heartidly because they don't have the ability to test his statements themselves.

Archi said...

@Samuel Andrews

LOL. Don't talk nonsense. No distance was shown. This cluster was not invented by anyone, it is obtained by formal tests, it stands out in ADMIXTURE and is perfectly visible on PCA. No one lies except you.

Palacista said...



The biggest problem is the strange idea >

one must devise a credible model of how the Indo-European-speaking segment of CHG found its way north of the Caucasus


It is a fact that there is an CHG component in the steppe population, so why is there a need for a credible model for an uncontested fact. The CHGs did get there and they most certainly spoke a language but there is no reason to assume that it was IE speaking before reaching the steppe, non at all. It is clear that the closest relation to PIE is proto Uralic making the case for a southern origin even weaker.

I still struggle to understand what the attraction is to site PIE south of the Caucasus because it seems to be a wish rather than an evidence based proposition.

Rob said...

The PiE is solved . We’ve got all the data/ samples required.
I’ve been looking at Caucasian languages lately

Arza said...

@ Rob
The PiE is solved . We’ve got all the data/ samples required.

OK, so where does R-M269 come from if we know everything and if we have all the required samples?

Rob said...

@ Arza
Maybe somewhere along the don basin ?
But I don’t understand your question; the 2 are not synonymous (unless you’re Carlos or Zardos)

Archi said...

Rob said...
The PiE is solved . We’ve got all the data/ samples required.

Rob said...
Maybe somewhere ... ?


LOL. LOL. LOL.



Rob said...

@ Archie
Keep laughing, clown

K33 said...

Not to mention all the most recently published qpGraphs are showing Steppe NOT as a simple EHG/CHG hybrid, but as a hybrid of EHG + something BASAL to CHG, ie, something with less WHG-related drift than CHG has.

The fact that an adequate proxy for this deep population has not been found (despite our possession of pre-Neolithic and Neolithic DNA from Iran, Caucasus, and Anatolia) proves ipso facto that the "CHG-like" signature on the steppe must be very old-- likely even pre-Neolithic.

Archi said...

@Rob

Have you ever written something that is not shameful or even a word of truth in your life? clown mammoth hunter.

old europe said...


@K33

You mean that CHG on the northern Caucasus could be a left over of a more Dzuzdzuana like population? While proper CHG is more ANE shifted IIRC.

Seems like proto WHG + Basal Eurasian= Dzudzuana

Dzuzdzuana+ ANE= CHG

In these equations how do you frame the Iran farmers. Maybe they are CHG+ more ANE?
I'm just asking

Arza said...

@ Rob
Equating R-M269 with PIE would be foolish, but its origin is one of the things we need to know to be able to say that we know how different IE languages have emerged.

Unfortunately, no, we don't have any of the required samples.

Matt said...

'Pretty unlikely'

Chain of genetic language transmission all along CHG->EHG->West_Siberian->Altaic unlikely (and sort of an extreme example), genetic transmission along some set of points certainly plausible, while areal+contact influence seems quite likely to me.

I don't want to get into the weeds of your "pIE descends from WHG-like groups of SE Europe via a *very* complicated chain of transmission" theory.

Davidski said...

@Arza

The samples are there, they just haven't been published yet. The spread of M269 went something like this...

Forest > Forest Steppe > Sredny Stog > Yamnaya & Corded Ware

Archi said...

@Matt

All three chains are not possible: CHG->EHG, WHG->EHG, ANF->EHG.

The CHG in the North Caucasus dates back to the final Paleolithic, and the steppe component clearly has links to the Caucasian CHG, but not to the Iranian component.

@Davidski

We have not any evidence about Sredny Stog > Corded Ware for M269. M269 can be though from R1b of Dnieper-Donets culture or others, like Narva, or others.

George said...


Off topic:

Identifying and Interpreting Apparent Neanderthal Ancestry in African Individuals
DOI:https://doi.org/10.1016/j.cell.2020.01.012

Lu Chen , Aaron B. Wolf, Wenqing Fu, Liming Li, Joshua M. Akey Cell Jan 30, 2020

Davidski said...

@Archi

I wasn't commenting on the position of Mallory and Balanovsky, but rather on the quality of their arguments.

They may well believe that PIE could not have originated in CHG, but I don't care about that, because this is what I have problems with:

Such an admixture would appear to involve two deep sources that should have developed separately over the course of thousands of years; in short, there is no reason to believe that the two components spoke closely related languages or even belonged to the same language families. Such a model suggested that Proto-Indo-European may have originated out of the merger of two very different language families, a theory that had once had been suggested by several linguists but had never attained anything remotely resembling consensus [62]. If one does not accept an “admixture language” then the natural question remains: did Proto-Indo-European evolve out of language spoken by EHG or out of language spoken by CHG?

Also, I see that you still don't understand that CHG and Iran_N are very different populations. But you'll have to accept it when it's eventually shown in scientific literature, and it will be.

And when I was talking about M269 in Sredny Stog, I was referring to unpublished data.

Rob said...


@ Arza

In terms of understanding the genesis of PIE grosso modo, all the data has been published

@ Matt
It’s really not that complicated or weedy & above all, it’s the most correct model
Indeed, that’s exactly what S/S was- a complex chain of transmission.

Archi said...

@Davidski

"Also, I see that you still don't understand that CHG and Iran_N are very different populations."

Here they are not talking about a population, but there are there really different populations from at least the late Paleolithic period, but about the genetic component that founded them. They're not talking about a population but about a component, I don't understand how you don't understand that. The population and the genetic component are different concepts.

R1b was in the (forest) Steppe already in the Neolithic Mariupol community.

Davidski said...

@Archi

CHG and Iran_N are different genetic components. Duh!

Archi said...

@Davidski

They have a common part, which is traditionally called CHG after the oldest samples from the Caucasus, which are older than the Iranian samples. This component is without additional impurities EHG, WHG, ANE, ...

Davidski said...

@Archi

You appear to be mentally retarded. But I'll try one last time.

CHG and Iran_N share mostly the same deep components, but CHG is more Dzudzuana/Anatolian-related while Iran_N has more ancestry from East Eurasia.

There's no evidence that CHG and Iran_N largely descend from the same one population. They may well be similar products of parallel processes that affected the Caucasus and the Iranian Plateau at around the same time.

Samuel Andrews said...

@Davidski, Archi is right when he says IranN/CHG create the same component in ADMIXTURE runs. No reason to call him mentally retarded. This is just a fact. But obviously, ADMIXTURE isn't a great tool for measuring deep relationship between ancient pops. So this fact is irrelvant in understanding how IranN and CHG were related.

Samuel Andrews said...

@Davidski,"while Iran_N has more ancestry from East Eurasia."

?????? No East Asian uniparentals in Middle East. What do you mean by East Eurasia?

Archi said...

@@Davidski

"They may well be similar products of parallel processes that affected the Caucasus and the Iranian Plateau at around the same time."

There is no evidence for this. I strongly doubt it. You just listed the impurities to pure CHG. Iran_HG is close to CHG than Iran_N.

Davidski said...

@Samuel Andrews

Well, strictly speaking there aren't any East Eurasian uniparentals in West Siberian foragers either, but I guess you wouldn't argue that they don't have East Eurasian autosomal DNA?

There seems to be an unusually high frequency of R2 in the Zagros Neolithic farmers. This might be a signal of their far eastern ancestry, which may have been a composite of various things from Siberia and Central Asia.

K33 said...

In Global25/Vahaduo, if you place modern Celto-Germanic and East/West Slavic speaking peoples in the "Target" slot, the Eneolithic Source combination of:

Sredny_Stog_II_En + UKR_Dereivka_I_En2 + RUS_Vonyuchka_En

actually produces substantially better fits than substituting those three for the familiar RUS_Yamnaya_Samara (using Globular_Amphora in both models to sop up farmer ancestry)

Romance speakers and South Slavs obviously have additional "southern" ancestry, but also East Slavs and Balts show poor fits across the board in these models.

Neither WHG nor EHG nor Narva really help the fits for East Slavs or Balts. So
we are clearly missing something from the Forest Zone.

@Davidski, can you divulge which archaeological culture the M269 sample came from? Northern reache Dnepier-Donets Culture?

Davidski said...

There's M269 in Volosovo and Sredny Stog.

Rob said...

The 'eastern' in Iran stone age at least partly is due to remnant ancestry from the IUP dispersal which otherwise disappeared further West (e.g. what has often been misunderstood as the east Asian affinity of some European UPs).
Y-DNA R is probably a 'central Eurasian' marker rather than 'eastern' per se.

Samuel Andrews said...

@Davidski,
"Well, strictly speaking there aren't any East Eurasian uniparentals in West Siberian foragers either, but I guess you wouldn't argue that they don't have East Eurasian autosomal DNA?"

That's because there like four samples from West Siberian foragers. East Asian uniparental will pop up if more samples come out.

if IranN had significant East Asian ancestry we should expect strains of East Asian uniparentals in the ancient and modern Middle East.

Samuel Andrews said...

@Davidski, How high coverage is Sredny STog R1b M269? Is it for sure?

Samuel Andrews said...

No one can definitivly say how the Epipaleolithic Middle East pops formed. For example, IranN and CHG are related but also different, but we have almost no way to know what the differences and similarities are till someone gets older ancient DNA.

David Reich speaks too confidently about what the relationship between EPipaleolithic pops was. Lumping together IranN & CHG as the same thing is one part of the problem.

Ariel said...

I think that Satsurblia and Zagros HG are approximately as distant as Pinabarsi and Natufian.

rozenfag said...

Regarding linguistic comments in this review: a lot of them may come not from Balanovsky or Mallory, but from a third author - Anna Dybo(I wonder why her name was not mentioned before). She is a linguist and supports idea of Nostratic languages. She also co-authored several papers with Balanovsky before.

Nick Patterson (Broad) said...

When did the Reich lab think that CHG and Iran_N are the
same genetically? A standard projection
onto W. Eurasian genetics that appears in many
of our papers refutes that.

Iran_N is useful technically because it is related to CHG
(even if very deeply) and we have so few CHG samples.

And a piece of advice: if you want a polite debate go easy
on the insults like "horseshit".

Samuel Andrews said...

@Nick Patterson,

A paper published by the Harvard lab as recently as September 2019 described Steppe-pasotrolists has being a mix of Iranian-related and EHG ancestires. It is a mistake to link Yamnaya's Middle Eastern ancestry to Iran if it is not from Iran but instead from the Paleolithic Caucasus.

https://reich.hms.harvard.edu/sites/reich.hms.harvard.edu/files/inline-files/2019_Science_NarasimhanPatterson_CentralSouthAsia.pdf

". In far eastern Europe at latitudes
spanning the Black and Caspian Seas there was the
Caucasus Cline, consisting of a mixture of Eastern
European hunter-gatherers and Iranian farmer–
related ancestry with additional Anatolian farmer–
related ancestry in some groups"

Samuel Andrews said...

Btw, I do not go a long with calling people names. You guys have given us so much fascinating discoveries that no one should be very upset!

TLT said...

@Archi

ADMIXTURE also fails to distinguish between sub-saharans, neanderthals and apes. It isn't a good argument. It probably detects deep common ancestry in CHG + Iran neolithic and just puts it in one group. There is considerable distance between scales CHG and scaled Iran samples. The reason for this is a combo of large time duration since split + different ghost populations contributing the to minority of the ancestry in these populations. Ultimately this is enough to classify them as 2 separate things.

TLT said...

@old Europe

>In these equations how do you frame the Iran farmers. Maybe they are CHG+ more ANE?
I'm just asking

Well, Iran farmers as in Ganj Dareh pastoralists and Tepe Adbul Hosein farmers had as much or most likely less ANE than CHG does, however, Hotu cave mesolithic seems to have more ANE than CHG and Iran_N. It would make sense since it is clear to central Asia and hence closer to the WSHG/pre-WSHG late ANE range.

Ric Hern said...

Thanks, very interesting.

Ric Hern said...

That Neanderthal paper is very interesting. Apparently Africans have ancestry of an earlier AMH/Neanderthal admixture event occurring around 160 000 years ago.

So when did this back to Africa migration from Eurasia of this population take place ?

What I said a few days ago.

"Interesting how the Climatologists suggest several opportunities for AMH to leave Africa between 250 000 and 65 000 years ago, yet somehow stick to the Pots = People thing... Could it be that some AMH populations adopted Neanderthal Technology to survive in these new environments ? Now I wonder if the Bottlenecks proposed previously do not correlate better with early AMH in Eurasia scenario....?"

vAsiSTha said...

@samuel andrews

I know you won't like it, but ehg + chg does not work for eneolithic steppe. If you have a working qpAdm model for it, please show us.
So in that regard, Narsimhan paper is spot on.

Ric Hern said...

There was an interesting documentary some time ago about two Neighbouring tribes in Papua New Guinea. They spoke completely different languages yet shared a festival yearly apparently for a very long time. They use gestures to communicate yet didn't bother to learn each others Language. Wonder how their DNA compare ? This shows how complicated evidence for language transferring can be...

Davidski said...

@vAsiSTha

Here you go...

Steppe_Eneolithic
GEO_Kotias_HG 0.551+/-0.022
RUS_Samara_HG 0.351+/-0.039
RUS_Tyumen_HG 0.098+/-0.036
chisq 14.736
tail prob 0.141991
Full output

Steppe_Eneolithic
GEO_Kotias_HG 0.889+/-0.553
RUS_Samara_HG 0.407+/-0.054
TKM_Geoksyur_En -0.296+/-0.505
chisq 19.406
tail prob 0.0354001 (infeasible)
Full output

Clearly, all that's happening is that we're missing the proximate hunter-gatherer sources from the southern part of the steppe.

Do you get it now? If not, here's an illustration:

https://drive.google.com/file/d/1c8gUMYnQ2QtCkPKxI-sLez67Il2QwBUA/view?usp=sharing

zardos said...

@Ric: That's because the initial big scale language transfer in most of Papua New Guinea happened a very, very long time ago. The landscape isolated people and they branched a lot of times. There was no paramount, higher social unity. Its the exact opposite to most of the World where there were many, up to recently, large scale expansions which spread new genes and cultural techniques, new languages and ancestry components over wide areas and homogenising them. So differentiation on the local level was either prevented or started anew.

Like with remnants of foragers in the worst places of the world, you can't make rules from exceptional cases for all. Clearly Eurasia was different and it was different for a very long time.
The most which happened on PNG is that one tribal alliance replaced another, but there was no recent sweeping expansion of one dominant group.
They all lived in relative isolation for very long times.

Good to read that widespread Neandertal admixture being now closer to being vindicated. I never believed the zero admixture South of the Sahara claim.

Suevi said...

https://www.academia.edu/41674465/CIEP%C5%81E._AN_ELITE_EARLY_MEDIEVAL_CEMETERY_IN_EASTERN_POMERANIA_ED._S._WADYL

https://i.postimg.cc/fWKCRCtd/20200127-190836.jpg

https://www.schweizerbart.de/papers/homo/detail/prepub/92507/Analysis_of_maternal_lineage_structure_of_individu?af=crossref

http://content1.schweizerbart.de//thumbnail/papers/92507.pdf/1000.jpg

Suevi said...

http://www.iaepan.vot.pl/czasopisma/index.php/przegladarcheologiczny/article/view/347

zardos said...

@David: What is your opinion on the study results and the reliability of the IBDMIX method? Can you apply it?

It seems totally convincing to me and proves, together with the latest ancient DNA from Shum Laka the great back migration I always thought to have happened, introducing fully modern H. s. to Africa, spreading uniparentals like yDNA E and Neandertal introgression.
We're coming closer to earlier assessments by physical anthropology which predicted that scenario.
Next ancient samples might be the final breakthrough.

epoch said...

@Knowledgeable Geneticist

I don't think the steppe was so isolated even in Eneolithic times. There were Khvalynsk kurgans with copper jewels. The copper was analyzed and is supposed to be from the Balkans. That must have meant trade. What could these groups have that was worth trading? I don't know. But whatever it was, you don't find copper on the steppe if these areas were massively isolated.

Davidski said...

@zardos

What's IBDMIX?

Pine trees and mountains said...

@epoch

There were some serious trade networks that traversed the Steppe, it is no surprise. Trade in general was intensive in all of Western Eurasia and Eastern Europe. As an example there were some BMAC artifacts and if I recall correctly specifically jewelry found in Caucasus Maykop remains. Trade doesn't mean however that with materials genes were "traded" as well, or not to the same degree. Just as Yamna had a lot of trade with Caucasus Maykop, but minimal mixing.

zardos said...

IBDmix is the method used in the new Cell paper ob Neandertal and West Eurasian admixture in Africans. It estimates admixture without a modern reference. Could be used to estimate any admixture with just the source, no reference, with more than 10 samples. The advantage is supposed to be the independent estimate, which eliminates reference bias if there was unaccounted gene flow. Like in SSA which are therefore no fitting reference to estimate Neandertal introgression into Eurasians.

Matt said...

Rob: 'It’s really not that complicated or weedy & above all, it’s the most correct model'

Pretty unlikely.

zardos said...

10 (modern) samples for comparison with one ancient source of course. Sounds promising to me.

Davidski said...

Well, it sounds like IBDMIX works great with different sub-species or even species of humans, but is there any evidence that it can be applied successfully in fine scale genetic analyses?

zardos said...

No. Question is whether thats because its just new or not suitable for more modern samples. Idk

old europe said...



If the tie between Sredni Stog and R1b M269 is confirmed then it is game over.

Sredni Stog has already R1a M 417

Sredni Stog is the source population of CWC ( the only difference being the fact that SS absorbed a more CHG shifted steppe population both from the Volga basin and from the northern caucasus)

Cultural ties of SS and balkano-carpathian coltural complex are as evident as day light.

PIE born out in the mating network between farmers and Dneper Donets foragers is the bottom line.

Better Reich Mallory and company update as soon as possible.

Archi said...

@old europe

So far, we don't have any samples from Sredniy Stog culture. Alexandria sample is not Sredniy Stog. The Alexandria sample has an obviously erroneous date.
Knowing how Europe and America make mistakes with Eastern European cultures, we must be very wary of statements about Sredniy Stog. It is necessary to look at what kind of burial ground, for example, the Mariupol burial ground is not Sredniy Stog, at best there are burials of the Novodanilovsky type. It is possible that this is Revova that at north of Odessa on periphery of Sredniy Stog, whose attribution is often done as Post-Stog.

Gaska said...

@Old Europe

Game Over???? Do you mean the language issue, the R1b-M269 issue or both? I've heard that phrase so many times that now it's fun to hear it-I have always said that to believe the fairy tales of the Kurganists we need conclusive evidence. I hope this time is true and not try to confuse the situation anymore

We must be patient and resist the temptation to jump the gun-

I don't understand why people are always drawing hasty conclusions especially after the Yamnaya disaster-The relationship between the Neolithic Balkan cultures and the steppes is evident, but the origin of R1b-M269 has absolutely nothing to do with the steppes

I guess you know that the origin of the Dnieper donets culture is the Swiderian culture-The areas of the upper Dniester in which the Dnieper-Donets culture was situated have mostly Baltic river names. Due to this, and the close relationship between the Dnieper-Donets culture and contemporary cultures of northeast Europe, the Dnieper-Donets culture have been identified with the later Balts-And you know what was in the Baltic countries? tons of R1b-P297-I suppose Archi and others will be able to confirm or deny what I am saying, in any case those cultures are not the source but the sink of R1b-M269


Archi said...

@old europe "Sredni Stog is the source population of CWC"

This is unknown.



Andrzejewski said...

@Knowledgeable Geneticist “NWC are linguistically Anatolia_C+Maykop. Darkveti were linguistically proto-Kartvelian or perhaps related to Kartvelian, early Maykop were Leyla Tepe who eventually admixed with locals. NEC languages are Leyla Tepe, Iran_C related languages. The similarities between NEC and NWC exist precisely due to the Maykop/Leyla-Tepe common element in both peoples (despite modern NWC having only tiny Iran_N admixture).”

So if Leyla Tepe are ancestral to NEC, who are the Elamites then? And what’s the relationship to Kura Araxes?

Archi said...

@Gaska

"I guess you know that the origin of the Dnieper donets culture is the Swiderian culture-The areas of the upper Dniester in which the Dnieper-Donets culture was situated have mostly Baltic river names. Due to this, and the close relationship between the Dnieper-Donets culture and contemporary cultures of northeast Europe, the Dnieper-Donets culture have been identified with the later Balts-And you know what was in the Baltic countries? tons of R1b-P297-I suppose Archi and others will be able to confirm or deny what I am saying,"

I do not confirm your words in the least, the Balts have no even indirect relation to the Dniеpеr-Donets culture. You've got it all mixed up. The names of the rivers have nothing to do with this culture. The origin of the Dniеpеr-Donets culture from Svider is one of countless hypotheses, in any case, unrelated to the Balts.

Andrzejewski said...

@Matt “I don't want to get into the weeds of your "pIE descends from WHG-like groups of SE Europe via a *very* complicated chain of transmission" theory.”

This is what Rob advocates. Besides, the Balkan is the area with the least amount of WHG ancestry in Europe

Archi said...

Andrzejewski said... "This is what Rob advocates."

It's just a laugh here.

epoch said...

@Knowledgeable Geneticist

"Trade doesn't mean however that with materials genes were "traded" as well, or not to the same degree. Just as Yamna had a lot of trade with Caucasus Maykop, but minimal mixing."

True. But surely such contacts are enough of a vector for language influence. The PIE words for donkeys are connected to (Proto-)Semitic ones, just as words for trade. Considering that there were donkey caravans traveling all through neolithic Anatolia I'd say trade brought these words over long distances.

It is striking though to see shared words for birds. Can you provide a list? Could some be onomatopoeic? There is a cognate for cattle in early PIE and Afro-Asiatic: *gwṓu- and Semitic gi, Epytian gw. However, if you ever heard a herd of cows low you know it's simply a representation of the sounds they make.

http://loanwords.prehistoricmap.com/wp-content/uploads/2017/06/Bj%C3%B8rn-2017-Foreign-elements-in-the-Proto-Indo-European-vocabulary.pdf

Andrzejewski said...

@Nick Patterson (Broad) “When did the Reich lab think that CHG and Iran_N are the
same genetically? A standard projection
onto W. Eurasian genetics that appears in many
of our papers refutes that.”

I was fascinated by this debate ever since you colleague Dr. Lazaridis published his essay in 2015, proving that Yamnaya were an almost equal admixture between EHG and CHG. Until then I was influenced by popular description of the Indo-Europeans as the mysterious population who brought horses, metal and fair skin to Europe along with their languages. I wonder where do YOU and the Reich lab think PIE languages came from, what root or trunk

Pine trees and mountains said...

@Andrze
I am not well acquainted with Elamites, but even proto Elamite culture is founded much later than Leyla Tepe, around one thousand years later. I am not aware of similarities existing between Elamites and Leyla Tepe.

Look what the evidence is though. KAC have a pretty high Iran N level, exceeding 20%. KAC comes centuries after Leyla Tepe and is also founded North-West of Leyla Tepe. LT culture itself would have even higher Iran N, I speculate 30%+, and was an extension of Zagros, East Anatolia Chalcolithic people, most likely Shulaveri Shomu, who were probably the ones who brought to Georgians the non-ENF G2a1 haplogroup. Leyla Tepe artifscts are also similar to Chalco Eastern Anatolian and North Ubaid artifacts. It is not out of question this could be due to recent trade, but there is a big chance that is from Shulaveri Shomu culture, which was coming to an end when LT and KAC appeared. Now, KAc samples found in Northern Armenia, who one would think should have been similar to modern Eastern Georgians, are still moderately distinct due to low CHG levels. Eastern Georgia received an input of a high CHG population after the existence of KAC or towards its end, which was certainly Western Georgians from the Colchian culture. This probably happened during the MBA and LBA with the new Shida Kartli culture that appeared in central Georgia and that had a serious weaponry culture and conflicts with an Eastern Georgian Kakheti culture (you won't find much about this in English). Meanwhile the NEC region was part of KAC, but it never received an additional high CHG input, but modern NEC have 8-15% Iran N admixture, similar to extreme Eastern Georgian who mixed the least with West Georgian invaders and retained the most Iran N among Georgians at around 8-10%. Do consider though that even this 8-10% is reduced compared to KAC levels, and even the 8-13% percent in NEC is reduced due to EHG admixture later on.

Hurro-Urartian languages, descendants of the KAC culture, are among modern languages most similar to NEC languages, especially the Lezgic branch, southernmost NEC group.

Basically all the stars are aligning and pointing towards NEC being an Iran C language group. I forgot to mention another thing, modern Vainakh people have a myth that their people have migrated from the South. This could be a myth about Kartvelian invaders expulsing the Hurro Urartians living in Eastern Georgia or it could very well be about KAC settlers, because KAC did cover NEC, all the way to Velikent and the Caspian coast of Dagestan.

Pine trees and mountains said...

@epoch
Of the top of my head I can think of khokhobi and kakabi and also possibly Georgian artsivi, Chechen aerzy, Hurro-Urartian artsiv and Russian orel which is pronounced as aryol being cognates from common CHG roots.

Pine trees and mountains said...

Another one @epoch the isolated Svan language's word for eagle is uerb, also being a cognate with the rest of the list.

epoch said...

khokhobi and kakabi sound like the noises of feasants. The first is a Georgian word, what is the second?

Pine trees and mountains said...

@epoch
Both are Georgian words. Kakabi is partridge in Georgian, kakaba is in PIE.

H'ero is eagle in PIE, Uerb is eagle in Svan, the rest of the cognates I listed above.

epoch said...

@Knowledgeable Geneticist

Kakabi/Kakaba is beyond any doubt a representation of the sounds the bird makes:

https://www.youtube.com/watch?v=fkvjQ7mS4Og

epoch said...

So the idea is that birds are named quite often after the sound they make and thus several unrelated languages could independently develop cognates.

Pine trees and mountains said...

@epoch
That is just what you want to hear, the bird's sounds don't give you a distinct kakaba, just kaka can be interpreted as the sound of the bird, -ba/-bi ending can't be explained by that. Not to mention that there are other animal cognates as well, some arguable but most very evident.

PIE dog - kwon Georgian puppy - lekvi/lekwi
PIE goat - dig-ha Georgian goat - T'kha
PIE lamb - woren/wṛēn Georgian sheep - Tskhwari/Tskhvari
PIE ram - erjos (pronounced eryos) Georgian ram - archvi
PIE small sheep - owika Georgian small sheep - bekeka (possible onomatopoeia)
PIE eagle owl - bughon Georgian owl - bu
PIE pheasant - kukubhos Georgian pheasant - kakabi
PIE cuckoo - kukulos Georgian cuckoo - guguli
PIE raven korwos/wornos Georgian raven - k'orani
PIE hoopoe - opopa Georgian hoopoe - op'op'i
PIE crab - karkros Georgian crab - kirchkhibi

I didn't include some a few like kata which are clearly from Semitic in both languages. Also didn't include a couple aforementioned cognates like khokhobi and uerb. There were also a couple fish specie cognates in PIE and Georgian but I think this suffices to prove my point. I used this video and an online PIE dictionary for the PIE words.
https://www.youtube.com/watch?v=7epZo_BQFqA





old europe said...



The onomatopoeic argument is rather flimsy: reality is different languages use different sound for the same animal.
If you ask an italian which is the sound of a rooster he will reply Chicchirichì

If you ask an english speaker he will reply: cock-a-doodle-doo

The rooster is the same in Italy and England though

Rob said...

@ Andrze

“ This is what Rob advocates. Besides, the Balkan is the area with the least amount of WHG ancestry in Europe”

Not quite. Somewhere north of Danube, maybe even southern Poland
The “amount” of WHG doesn’t really mater .
Anyhow; I really wouldn’t worry about what Matt thinks; he’s clueless about Europe ; and an R1b chauvinist

old europe said...


@archie

AFAIK nobody has put forward the idea that Alexandrya could be later than 4100/3800.

Andrzejewski said...

@Konwlegeable Geneticist “Look what the evidence is though. KAC have a pretty high Iran N level, exceeding 20%. KAC comes centuries after Leyla Tepe and is also founded North-West of Leyla Tepe.”

When you say “Iran_Chl”, do you that population we’re debating how distinct they were from actual CHG, or was it a completely different one?

Pine trees and mountains said...

@Andrze
Not sure what you mean exactly. According to this PCA, despite not being distinct in that PCA, I'd say CHG and Iran N are distinct groups. As long as we don't find individuals who can fill up that large gap between them I wouldn't consider them part of one cluster.

As for Iran C, Iran C already had a lot of ENF and Levant_N admixture so it would be very distinct from CHG. I could guess Iran C are the 5 samples cluster South-East of CHG, but I can't say for sure.

http://eurogenes.blogspot.com/2018/01/another-look-at-genetic-structure-of.html

Archi said...

@old europe
"AFAIK nobody has put forward the idea that Alexandrya could be later than 4100/3800."

You are wrong, there are graves in Alexandria from the Neolithic to the Bronze age, that is, everybody knew it, there are different cultures and even more graves whose culture is undefined. Initially, this burial was tentatively dated to the middle of the fourth Millennium, it does not contain any cultural definitions. I'm sure that radiocarbon date in it was made with errors, this happens. What is absolutely certain, the reservoir effect was not deducted, which can be very large. In any case, even this date does not belong to Sredniy Stog, but to Dereivka culture.

Andrzejewski said...

@KC what I meant was not Iran v. CHG but in regards to Kura Araxes: who were the “Iran” people in KA and NEC?

epoch said...

@Knowledgeable Geneticist

Where did you get that PIE contained a word kakaba for partridge, by the way? I have a hard time finding a proper reference for it.

Pine trees and mountains said...

@epoch
https://indo-european.info/dictionary-translator/translate.inc.php/English/Indo-European/?q=partridge

@Andrze
https://postimg.cc/CdwM0kxX
As I said, Eastern Georgia, Armenia and NEC all spoke languages of one language family, but then Indo-European was established in Armenians and Kartvelian in Eastern Georgia, although due to high CHG in Eastern Georgia I speculate that Eastern Georgia was proto-Kartvelian lands before Iran_C rich people came to the region and the MBA/LBA Kartvelian conquest of Eastern Georgia was a reconquest of the area rather than being the first time Kartvelians steping foot in the region.

Matt said...

@Andrejz: yes, that is the hypothesis of Mammoth Hunter aka Dragos aka Rob. That is why I said it was the hypothesis of Rob aka Dragos aka Mammoth Hunter.

epoch said...

@KC

Yes I saw that. It's the only reference of the existence of a PIE *word* "kakaba" meaning partridge. I can't find it in any other resource.

epoch said...

@KC

So I have a hunch that is not a *proper* resource.

Samuel Andrews said...

@zardos,
"It seems totally convincing to me and proves, together with the latest ancient DNA from Shum Laka the great back migration I always thought to have happened, introducing fully modern H. s. to Africa, spreading uniparentals like yDNA E"

There's no old Eurasian mtDNA in Africa though. Y DNA E in Africa might be of Eurasian origin but...where's the Eurasian mtDNA?

andrew said...

@Archi

"the matter is complicated enormously by the fact that CHG cannot possibly represent the signature of a single language family as it is found over a broad area from the Caucasus to the southern Zagros that encompassed the territory of a whole series of other language families,e.g., Hurro-Urartian, Elamite, Kartvelian, and one must devise a credible model of how the Indo-European-speaking segment of CHG found its way north of the Caucasus [68]."

I have long argued for a West Asian macro-linguistic family, of which one litmus test is ergativity, that includes most or all of these languages. For whatever reason, this corner of the linguistics world is full of splitters and not lumpers, but that has more to do with the sociology of the discipline than the underlying linguistic reality. Those languages aren't well enough documented and attested to rule out a linguistic family relationship, in part because they emerge at the very dawn of genuine written languages. The fact that speakers of these languages share CHG genetics and that CHG genetics remained very segregated and distinct from adjacent Anatolian and Fertile Crescent genetics despite the exchange of Neolithic technologies between the two in the Fertile Crescent Neolithic strongly suggests that there were meaningful cultural divisions between the two groups which frequently are associated with linguistic family divisions.

This doesn't mean that the case of a CHG source of PIE or the associated culture is solid, even though that is a naively plausible scenario before you look hard at detailed evidence. The genetic evidence doesn't support it, and the technology transfers from early Neolithic and Bronze Age farmers to Pontic hunter-gatherers and herders can be supported archaeology to evidence of contacts taking place in the general vicinity of Moldova, even though there are some echoes of similarity in material culture across the steppe-caucuses boundary (the direction of which is indeterminate).

Andrzejewski said...

So according to you, @KG, Kartvelians and Svans speak a CHG language family, NEC came from Iran, whatever their genetic composition implies (including Kura Araxes), NWC spoke Anatolian farmer languages like the LBK or the Raetians?

Andrzejewski said...

So if I can expound on what @KC is saying: who were the Ubaidians, the Sumerians and the Halafians? Semites (Pre-Pottery), Iran Neo/Chal or Anatolian Farmers?

Pine trees and mountains said...

@epoch
"So I have a hunch that is not a *proper* resource."

What the hell are you talking about. These guys are more than sufficiently competent to create a legitimate and trustworthy lexicon.
https://academiaprisca.org/en/modern-indo-european/

Rob said...


@ Matt
Ooh Matt you’re so sassy !. You go girlfriend.

epoch said...

@KC

"What the hell are you talking about. These guys are more than sufficiently competent to create a legitimate and trustworthy lexicon."

Why then can't I find any other reference to the PIE word "kakaba"? And who exactly are "these guys"?

zardos said...

@Samuel: There are two options, either mainly males moved South or the mixed male:female population was more successful, or many mtDNA haplogroups which were considered "African" came ultimately from Eurasia. Remind you, that Mota is close to the Eurasian related branch and Shum Laka have, even though they might have North Eastern African/Eurasian admixture already, showed just L0a2a1 and L1c2a1b.

This new study on African mtDNA highlights ancient Eurasian admixture as well, as they write:
"Most of the non-L low-frequency lineages found in the novel datasets are highly informative about human history. Lineages M1(a) and U6(a) are explained as signals of ancient backflow into North Africa from the Mediterranean area in the Early Upper Paleolithic [[30], [31], [32], [33]]. N1a1a is a low-frequency lineage with a relict distribution likely indicating a Pleistocene dispersal from Arabia [31,32]."

Additionally, despite clearly West Eurasian haplotypes like these, the position of L3 is also not clearly ancient West African, but more East Afrian, probably even Eurasian related. In any case its not at home in West Africa:
"More recent analyses, including Soares et al. (2012) arrive at a more recent date, of roughly 70–60,000 years ago. Soares et al. also suggest that L3 most likely expanded from East Africa into Eurasia sometime around 65–55,000 years ago years ago as part of the recent out-of-Africa event, as well as from East Africa into Central Africa from 60–35,000 years ago."

https://en.wikipedia.org/wiki/Haplogroup_L3_(mtDNA)

From the mtDNA paper:
"The mixed West African sample (n = 145) was composed of 37.9 % L3 haplotypes (in descending frequency L3e, L3f, L3b, L3d, L3k), 37.2 %"

L3 came with the Eurasian backflow to West Africa I'd say, whether directly or rather via East Africa is another debate.

If you consider that even Yoruba showed about 1/3 of the Eurasian Neandertal admixture, we can assume that at least 30 percent of modern Negroid ancestry could be attributed to recent backflow if the paper is right. However, this is the lowest estimate, because the Eurasian backflow itself might have been diluted with Basal Eurasian and East African ancestry, which reduced the actual Neandertal introgression, making an even larger scale replacement feasable imho.

Now compare that with the Shum Laka paper, the branch related to Eurasians in West Africans accounts for the majority of modern SSA ancestry.

I don't know how this can be split up and further differentiated, but to me it is highly likely that:
1. There was a massive, sweeping back migration from Eurasia
2. This back migration was carried on by Eurasians with already reduced Neandertal ancestry ("Basal Eurasian" related)

This means, please point me to logical fallacies, if there are some, if the new paper and IBDmix method is correct, we deal with minimum 40 percent replacement in West Africa, possibly even much higher.

Ric Hern said...

If CHG and EHG mixed during the Mesolithic or earlier, then the innitial CHG/EHG mixture can not be directly linked to the formation of PIE at 4000 BC. If so then the linguistic changes that could have occurred during that long time before the formation of PIE would have left the original EHG and CHG Languages basically unrecognizable compared to PIE...

Rob said...

@ Ric

In fact, the initial EHG/ CHG admixutre is not linked to PIE genesis at all. Again, this should be 100% clear by now. So one can only wonder why some people, both here & in ''academia'' are so hopeless.

Pine trees and mountains said...

@Andrze
"So if I can expound on what @KC is saying: who were the Ubaidians, the Sumerians and the Halafians? Semites (Pre-Pottery), Iran Neo/Chal or Anatolian Farmers?"

Look, I will try making this as short and concise as possible.

Svan and Georgian have quite distinct lexicons but almost identical grammar, despite Svan having diverged from common proto-Kartvelian in 3000 BC, as is thought, so 5000 years ago. As population rises, so does the speed of divergence of languages (and no communication technology was not developed enough during that time to counter the population rise effect). What I think is that Svan and Georgian are more different from each other than proto-Kartvelian in 3000 BC and the language that Satsurblia and Kotia spoke. So I think it is highly probable the grammar of modern Kartvelian languages is minimally changed from Kotia's and maybe even Satsurblia's times.

NEC languages are related to KAC, Maykop, Leyla-Tepe who all spoke similar or one language. Why?
1. Leyla-Tepe reached Maykop lands by traversing all of the Northern slopes of the Caucasus range.
2. KAC had a long presence in NEC
3. NEC currently have the highest amount of Iran_N among native Caucasian groups (by which I mean NC groups and Kartvelians), BUT we have to consider that Kartvelians were not affected by a large Yamna input. Before the Yamna input NEC would have 18-20% Iran_N.
4. Caucasian Albania is a region where NEC-related languages were spoken. It is also the region where Leyla-Tepe was present. The whole spread and habitat area of this cultural triumvirate coincides with the spread of NEC people historically and in the present.
5. They get modeled with Kura-Araxes culture incredibly well (along with Armenians, but Armenians got linguistically Indo-Europeanized with no significant genetically Yamna-related input and this is well-known).

NWC are Anatolia_C language speakers. Why?
1. NWC have BY FAR the highest amount of the ENF subclade in the Caucasus, G2a2.
2. Maykop spoke a NEC-related language, and it left a trace because NWC and NEC have some similarities, but the similarity of NEC and NWC, as you can find out if you Google it, is very vague and ambiguous and the grammar is quite different.
3. Can there be any other explanation for the huge amount of G2a2 in NWC? Not really. If you say it was a founder effect from the small group of G2a2 that Kartvelians had, then NWC would speak something remotely similar to Kartvelian, but they don't. There is no other possible source of G2a2. I even heard some guy claim that it's from Cucuteni-Tripoliye migration all the way to NWC but that's false because NWC would have a trace of extra WHG on top of the WHG they would have from Yamna, but they don't. So no input from the EEF. Maybe it was from Shulaveri-Shomu one might ask. No, SSC did not even spread to NWC and was culturally similar to the preceding Zagros Iran_N cultures which makes me assume that SSC spread G2a1 in the Caucasus (at least the Caucasus part that it covered).

As I said, one fact alone seemingly gives you nothing, but a lot of these facts add up and the "stars align" to give you a clear answer. Only little puzzle here is the Dolmen culture but I don't expect any surprises.

Archi said...

@Knowledgeable Geneticist "What I think is that Svan and Georgian are more different from each other than proto-Kartvelian in 3000 BC and the language that Satsurblia and Kotia spoke. So I think it is highly probable the grammar of modern Kartvelian languages is minimally changed from Kotia's and maybe even Satsurblia's times."

LOL.
Palaeolithic Georgia Satsurblia cave [SATP] 11430-11180 calBCE
Mesolithic Georgia Kotias Klde [KK1] 7940-7600 calBCE

Samuel Andrews said...

@zardos, This is interesting stuff.

"If you consider that even Yoruba showed about 1/3 of the Eurasian Neandertal admixture, we can assume that at least 30 percent of modern Negroid ancestry could be attributed to recent backflow if the paper is right"
"This means, please point me to logical fallacies, if there are some, if the new paper and IBDmix method is correct...."

I just don't believe estimated percentage of Neanderthal ancestry can be used to determine percentage of Eurasian ancestry.

"L3 came with the Eurasian backflow to West Africa I'd say, whether directly or rather via East Africa is another debate."

L3 definitly could be East African originally. You obviously know something about African mtDNA. South Africa was originally only mtDNA L0. Considering this, East Africa could have been L3's home whole South Africa was L0's home. But, even if L3 in West Africa is of East Africa origin, that still means there's basically no Eurasian mtDNA in West Africa.

"There are two options, either mainly males moved South or the mixed male:female population was more successful, or many mtDNA haplogroups which were considered "African" came ultimately from Eurasia."

Both are interesting and upend what was previously thought. The later makes more sense to me. Y DNA E is definitely Eurasian originally. But, the only Eurasian mtDNA seems to be M and N. All other L3 clades seem to be African.

Y DNA E's original mtDNA compadra will probably turn out to be M1. U6 is common in Ibermausian, but originally comes from people related to Paleolithic Europeans who didn't carry anything related to Y DNA E.

Rob said...

@ Sam
U6 in Africa. isn’t necessarily from Europe because it’s found in Anatolia and Caucasus too

Andrzejewski said...

@Ric Hern “
If CHG and EHG mixed during the Mesolithic or earlier, then the innitial CHG/EHG mixture can not be directly linked to the formation of PIE at 4000 BC. If so then the linguistic changes that could have occurred during that long time before the formation of PIE would have left the original EHG and CHG Languages basically unrecognizable compared to PIE...”

Would you consider PIE a language isolate?

zardos said...

@Rob: Let's assume original hunter-fisher groups of EHG spoke a language which ultimately ended up as PPIE after they mixed with CHG from the Caucasus region or even more Southern, developed a new culture and their language was influenced too.
Obviously there would have been different related languages spoken by its descendents after 3.000-4.000 years. So PIE would be just one specific subset, but it would be a descendent nevertheless. That's my opinion as long as you can't prove a broken chain. So far, the chain is not complete, so its speculative, but all in all, it points to that chain of events and ancestors. You always end up with the hunter-warrior clan which fused with CHG people in the Lower Don-Piedmont region. And if you think somebody else spoke PIE, you have to prove how a dominant elite fits into this chain and could turn it around. I see no such proof and the speculative arguments are weaker.

Also, different steppe people of the same EHG-CHG core group could have spoken different languages at later times, simply because of 3-4 thousand years and a lot of influences in between. Even without somebody else on top.

Concerning mtDNA L3: Its being assumed it is from East Africa because of the greater diversity there. But let's not forget how often we heard that argument (R1b and R1a in particular!) and how often it was plain wrong! Unless they have found it, who knows? The Near East in particular was no stable, calm place to exist. For sure not.

But there are different thinkable scenarios.

@Sam: If modern West Africans/Negroids have 1/3 of the European admixture and show by now more recent connections to Eurasians on all levels, how else could that have happened? The spread of E, the spread of agriculture and Niger-Congo languages at the expense of earlier forager people with a very different genetic and cultural profies. This is no coincidence. There are just parts of the puzzle which are still missing.
The extreme physical variation in West Africans always needed an explanation and the explanation of "just variation" and "old diversity" was always crap. Populations homogenise over time in pre-civilisational times and most patterns in this respect have a cause. I think we are close to find the cause, which is the major mixing event happened just fairly recently. Even more recently than I assumed before the Shum Laka revelation. 3.000 BP and the main replacement and admixture event was at best just starting in Cameroon.

Rob said...


@ KG

Yes, the history of Caucasian language speakers are being increasinlgy discernable
However, I would modify some of your suggestions:

1) '' KAC, Maykop, Leyla-Tepe who all spoke similar or one language.''

But Majkop (Leila-Tepe is the southern extension of the Majkop phenomenon) & K-A are very different groups, culturally speaking, even if they are from a ~ common pool of populations.

2) NEC: you need to highlight the large Yamnaya-Catacomb component there.
So, in a nutshell, NEC are a mixture of KAx and Catacomb.

3) NWC: The lineages of NWC are dominanted by ''Anatolian-European'' variety of G2a.
You're right, one cant seem to find a relevant role for European EEFs in playing around with models, but that might relate to the fact that NWC is a bottlenecked population ?
Hard to say anything definitive for NWC. Heavily local CHG, probably derive from Dolmen culture

Rob said...

@ Zardos

“ I see no such proof and the speculative arguments are weaker.”

That’s because you guys are simply blinded by your own biases , and aren’t very attentive to begin with
I mean for heavens sake - look the dominant lineage of Mariupol & Sredniy Stog; look at EBA Bulgaria . One doesn’t need to have a palanatir to figure out what proto-Anatolians and several Balkan IE are going to be (???)

“ You always end up with the hunter-warrior clan which fused with CHG people in the Lower Don-Piedmont region”
Which “warrior clan from the lower Don”?
Please take the time to learn about the origins of the kurgan phenomenon from European scholars , so then you don’t have to make up stories.

zardos said...

"look the dominant lineage of Mariupol & Sredniy Stog; look at EBA Bulgaria . One doesn’t need to have a palanatir to figure out what proto-Anatolians and several Balkan IE are going to be (???)"

Well, what do you have from the Mariupol elite and Sredniy Stog? For sure not sufficient data base.

The core & main body of the PIE/IE is clearly dominated by R1. Corded Ware, Sintashta? The outer ring is not as important as the central group. Cernavoda was a mixed culture, yet what do we really have? You will find steppe lineages there, while you have not enough else in IE core groups.
This is not definitive, but it favours the EHG clan.

Rob said...

@ Zardos
Of course, that the main volume & most prolific of IE was exapnded by R1-lineages. But we' (or at least I) am discussing the genesis of the process- what got the ball rolling. I'm not interested in a pissing contest (in any case, I actually have no horse in the race)
Moreover, R1a -M417 & R1b-M269 might have been very different groups intially. To clump these into 'EHG clan' is rather pseudo-scientific w.r.t to the matter at hand.
Your claim that there isn't enough samples Mariupol samples should direct to you the published data.
So I ask for th 4th time w.r.t. your stated Lower Don group- which L/D group ? Do they have especially rich burials ?
I can certainly see their role as a link in the process of suffusion of EHG/CHG through the steppe, but what makes them them the most instrumental.

Ric Hern said...

If PIE originated around 4000 BC. from Languages spoken by Mesolithic or Pre-Mesolithic CHG/EHG mixed peoples then yes it was an isolate for most of its Proto-PIE developement...

Rob said...

Ric
In fact the EHG/ CHG language is extinct hehe. That’s pretty “isolated. “

Pine trees and mountains said...

@Archi

"LOL.
Palaeolithic Georgia Satsurblia cave [SATP] 11430-11180 calBCE
Mesolithic Georgia Kotias Klde [KK1] 7940-7600 calBCE"

Despite the greatly increased amount of foreign cultural influences that have caused both Svan and Georgian to changed, it is still not doubted by anyone knowledgeable about the subject that Svan and Georgian are part of the same language family.

I will repeat myself since your small dogmatic brain could not understand it from the first time, Svan split from the common proto-Kartvelian language 5000 years ago. 3000 BC is just 4000 years after Kotia. Western Georgia has been one of the most isolated regions in the world up until the Iron Age. Only significant foreign influence it could have received was from the ENF-Levant_N input, but considering that compared to autosomal percentage, farmer Y-DNA is very low, it is out of the question that Kartvelian has been significantly changed by the Anatolia_C group, because if it were significantly changed it would have similarities with NWC languages. So considering the isolation of Western Georgia it is entirely plausible that to the proto-Kartvelians in 3000BC the language of Kotia and perhaps even Satsurblia was quite recognizable and maybe even slightly mutually intelligible. Perhaps it would be recognizable even to a modern Svan.

Ric Hern said...

I think if you need a Linguist to point to similarities between Languages those Languages are prity much unintelligible to the general public....I think that is what most of us forget. Personally I can not hear any similarity between English and Hindi, but Linguists on paper will point it out. As a Germanic Language speaker it is even difficult to hear similarities between West and North Germanic Languages...so even a few hundred years can make a massive difference phonetically...

Ric Hern said...

Unless the art of Linguistics stretch back before the dawn of time...at best I think deep relations could only be found when writing came into the picture....

Rob said...

“Farmer” dna in Georgians is up to ~ 5o % (G2a; non -native J2). Sure they’re isolated; but not that isolated
Anyhow; I’d heard the “TMRCA” of Kartvelian was c 2000 BC

Ric Hern said...

@ Rob

"In fact the EHG/ CHG language is extinct hehe. That’s pretty “isolated. “"

Yes or Languages. So yes that is basically how I see it also.

Ric Hern said...

We can not even rely on Agglutinative or Non-Agglutinative as something that is fixed...just like many other previously proposed apparently unbendable rules...Change is constant and even things written in Stone have an expiry date...

Pine trees and mountains said...

@Rob
"“Farmer” dna in Georgians is up to ~ 5o % (G2a; non -native J2). Sure they’re isolated; but not that isolated
Anyhow; I’d heard the “TMRCA” of Kartvelian was c 2000 BC"

Autosomal farmer DNA (if we consider the ENF and Natufian component as farmer) does not reach 50%. It doesn't even reach 45% in any part of Georgia, and in most parts it is up to 40%.

I have talked about this in 2 separate posts and clearly you either have bad short-term memory or didn't bother reading either post. There is not one G2a1 in ENF and EEF samples, not a single one. Do consider how many EEF and ENF samples we have found, we have even found C among EEF, yet not even one G2a1. Meanwhile, late Neolithic Iranian has G2a1 with minimal ENF admixture.

https://umap.openstreetmap.fr/en/map/ancient-human-dna_41837#12/34.4795/48.0000

But you might say, "that can very well be from the ENF!". No it cannot, it cannot because the highest G haplogroup diversity we see in ancient groups is among Iran_N and Iran_C. We see both G2b and G1a in them, two subclades which today are extremely rare. Yet, the ENF are dominated almost entirely by G2a2 with no other subclades of G2 in sight, at least not in a remotely significant amount cause we haven't found one.

G2a2 in ENF itself almost definitely comes from Iran_N and perhaps CHG, because Pinarbasi was C, and later we see that C has gone almost extinct among farmers, we just find 1 or 2 Cs among the tens of samples of ENF and EEF.

If you consider Iran_C to be a farmer group, then we have around 35% farmer Y-DNA, but genetically speaking when we use the word farmer we use it as shorthand for ENF and Levant_N groups. So in reality in Georgia we have low Iran_N autosomal admixture but high Y-DNA - G2a1 specifically, and the Ls, R2as and Ts one might occasionally find among Georgians of different regions is also a trace of that Iran_C related expansion.

As for our J2a being a non-native element, I can only reply to that with "lol".

Oldest instances of both J1 and J2 are from Western Georgia, and we know they weren't recent migrants because they were blocked off from the rest of the world during the ice age, so they were in the region at least before the LGM. Claiming that a significant amount of J2 in Georgians is non-native is beyond foolish. Not to mention that we know very well that the J2 that is so common in modern Anatolia and Levant and Arabs was spread by expansion of KAC, expansion of Iran_N/C people and later expansion of Hurrians.

Here is one major J2a subclade of Georgians: https://www.yfull.com/tree/J-Y12378/

Here is another: https://www.yfull.com/tree/J-M67/

These are like 95%+ of our J2s, so you will have a hard time proving they are not native if you do try it.

CrM said...

@Rob, @KG
Proto-Kartvelians were isolated, this is a fact supported by Archeology and Archaegenetics. The Core Kartvelian region is Colchis, ie Western Georgia, and it was never a part of KAC like Eastern Georgia was(which was likely inhabited by various proto-Lezgic, proto-Nakh and Hurro-Urartian tribes). Kartvelians expanded to East Georgia at a much later date, well after KAC.

Genetically Georgians and pre Steppe admixed NWC are essentially the same, the only difference is that Georgians have a slightly higher % of Neolithic Iranian ancestry.

Kartvelians and Northwest Caucasian having a similar genetic background, but them speaking two unrelated languages is the most peculiar part.
The best Eneolithic proxy for proto-Kartvelians and proto-NWC is Darkveti-Meshoko, which was modeled as half CHG and half Anatolia_C in Wang et al.
You can assume that one Darkveti-Meshoko (Darkveti?) related group retained their original CHG language, while another (Meshoko?) switched to Anatolian. I'm leaning on NWC being the Anatolian speakers, it explains the prevelance of Anatolian G2a clades in NWC (Georgians are mostly G2a1, this clade was never found among ENF, so saying that Georgians are up to 50% "Farmer" is incorrect, the oldest G2a1 sample is from Iran_N, I also think it's possible that CHG carried G from the start/received it from Iran_N who had all the G varieties including G2a, G2b, G1 etc), and the Hattic-NWC similarities (https://www.academia.edu/1215069/The_Relation_of_Proto-West_Caucasian_to_Hattic).

The smoking gun for Kartvelian being a CHG-related language is its similarities with PIE, but things can get more complicated if you take Maykop into account, who had contacts with various Steppe peoples like Steppe Maykop, Yamnaya and Usatovo.

Rob said...


@ KG

“ have talked about this in 2 separate posts and clearly you either have bad short-term memory or didn't bother reading either post. There is not one G2a1 in ENF and EEF samples, not a single one. ”

Who said anything about european farmers ? It should be self evident that they don’t have anything directly to do with Georgia

G2a - whichever clade- isn’t native to Georgia . It arrived with “farmers” from whether S/S or some other group.
Please use more precise nomenclature in future
Yes I know J2a is found in Native caucasus hunter gatherers; but some could be from zagros or even northern Mesopotamia
Do you have an exact breakdown for J2a sub-Clades ?

The similarities of Kartvelian with PIE I take Arame view quite well- sprachbund effect due to Armenian etc
Otherwise CHG has nothing directly to do with PIE

So, Kartvelian might be a “CHG language”; but it’s so after the fact it’s something of a moot question, apart from appeal to romantic nationalism

Rob said...


@ CrM
Indeed. But I believe we can discern a difference , demonstrably, between NWC & Kartvelians

Pine trees and mountains said...

@Rob
There is no J2a found among Iran_N, only J2b so far.

Farmers are shorthand for ENF, EEF and Levant_N, stop being an autistic pedant and use common sense in these discussions. We don't even have a sample of SSC and I can already guarantee you that they are more Iran_N/CHG than ENF, you will see in time.

If it were a sprachbund effect due to Armenian, Georgian and Armenian would have more similarities than they currently have, yet most linguistic similarities we have are due to Persian influence.

Rob said...


@ KG

''Farmers are shorthand for ENF, EEF and Levant_N,''

Really ? According to which colloqium

'' autistic pedant and use common sense in these discussions. We don't even have a sample of SSC ''

Which are farmers.
Anyhow; check out the YDNA calls from Hotu cave

CrM said...

@Rob
The biggest difference, aside from Steppe ancestry in NWC(bulk of which should have been received after the Dolmen culture) is Iran_N ancestry (which Georgians have and NWC none or very little), CHG ancestry (NWC have a slightly bigger amount) and Anatolian/Natufian/Levant_PPN ancestry. But those differences are minimal.
https://i.imgur.com/rSX3xNN.png

epoch said...

@Knowledgeable Geneticist

I can't find any reference to the PIE words you mention in this PIE lexicon from the university of Helsinki.

http://pielexicon.hum.helsinki.fi/

Neither can I find them in this list, compiled by a linguist for his master thesis

http://loanwords.prehistoricmap.com/wp-content/uploads/2017/06/Bj%C3%B8rn-2017-Foreign-elements-in-the-Proto-Indo-European-vocabulary.pdf

The source you mention is by Carlos Quiles who wants to unite Europe under a revived Indo-European language.

https://carlosquiles.com/about/

Pine trees and mountains said...

@Rob

Imbecile, Hotu is J2a2.

Try finding a J2a2 here https://www.familytreedna.com/public/Georgia/default.aspx?section=yresults


Pine trees and mountains said...

Try checking the facts a little before putting out an idiotic answer, I am not your kindergarten teacher.

Pine trees and mountains said...

@epoch

Thanks for clearing that. I didn't expect fucking Carlos to be involved with this lexicon. I thought he just sticks to his own blog.

Pine trees and mountains said...

@Rob
No J2a that was relevant to the discussion, and the discussion in that specific case was about the native and non-native J in Georgians. At this point, the only reason I am talking to an underage ex mammoth hunter who out of embarrassment or slight maturement changed to a more normal name, is because it gives me slight joy humiliating morons who think they even have a slight understand of the topic they are discussing.

As for nationalism, if there are historical facts that grant more legitimacy to the status of an autochtonous nation and culture that my nation has, I will use such facts to my advantage, but only someone who received brain damage after a failed mammoth hunt will accuse me of skewing any facts and proposing improbable scenarios.

Rob said...

@ 'Knowledgeble' Geneticist

Let's review the facts here:

1) You stated ''but considering that compared to autosomal percentage, farmer Y-DNA is very low''

To repeat, you did not state what 'farmer' you're referring to'. I'm not trying to be difficult, nor was I necessarily disagreeing with you, or ''siding with Archie' .
G2a1 has not been found in Caucasian hunter-gatherers. which means it came with the neolithic/ farming, or later.
If you interpret 'farmer' selectively, then outline/ define it in the first place, instead of losing your shit that someone hasn't read your mind as to your own idiosyncratic definitions of it.


2) Then i pointed out ''Yes I know J2a is found in Native caucasus hunter gatherers; but some could be from zagros or even northern Mesopotamia
Do you have an exact breakdown for J2a sub-Clades ?'' (asking genuinely here)

To which you replied
''There is no J2a found among Iran_N''.

I asked about Hotu -Cave (technicaly a 'Mesolithic'') to which you replied ''Imbecile, Hotu is J2a2.''

How can you school people when you can't even follow your self?

& what exactly is wrong with "M - H" ? A throwback to ancient times, I certainly find better than an appeleative with grandiose claims. So settle down, you might blow a capillary.

Pine trees and mountains said...

@Rob
I just told you, I am not your kindergarten teacher, you can go do some basic research to instead of asking me everything, otherwise we are bringing down the level of discussion on the thread. When you told me to check the Y-DNA of Hotu, you said that as if it were relevant in any way. No, you were once again pedantically targetting just one tiny part of everything I have written, about J2a not being present in Iran_N, and when I wrote that I am sure you understood perfectly what I said, that the J2a that Georgians have is not present in Iran_N, but you had to go on using empty purposeless words, so I will follow along if that is what you want. I don't expect people to be extremely well-read about the topics that they discuss, but I can see you are not new to this topic, and you could have very quickly checked yourself that the J2a of Hotu and J2a of Georgians and other Caucasians is different. FamilyTreeDna and yfull are tools that every amateur knows.

I don't like empty baseless speculation and I never ruled out Iran_N having J2a1 but as long as we don't have such a sample and we have evidence for long isolation of CHG in the WG refugia odds are high that Iran_N did not have J2a and the Satsurblia J1 clade and the J2a and J1 of CHG were at the time CHG exclusive.

Archi said...

@ Knowledgeable Geneticist

You out of ignorance of linguistics do not understand what you are reflection. Kartvelian experienced the strongest influence of the North Caucasian languages. Languages ​​always change, even Old Georgian grammatically very different from modern. 4500 years is a gigantic period, it is approximately the decay time of Indo-European languages, you can imagine how Hindi differs from English, and even English from German, Svan from Georgian, etc. Whether Kotias spoke, let alone Satsurbliya, was unknown to the Protokartvelian language. Therefore, yes, linguistics laughs at your statements.

Rob said...

@ CrM

It's also useful to use more proximal samples for the source base
So if you include Majkop & K-Ax, there is no additional Iran _N needed in any of them.
Both NWC & Kartvellians can be modelled as predominantly Meshoko-type ancestry; with NWC having Catacomb-like admixture (makes sense) & some minor Hunic or Turkic or Mongol (whichever is most historically appropriate, didnt split hairs in the basic model)
Curiously some Abkhaz plot as if their from Georgia

https://imgur.com/GCxHfyP

Pine trees and mountains said...

@Archie
A large chunk of Linguistics is pseudo-science, so Linguistics can laugh as much as it wants. Anyway, I am fortunate enough to know Georgian, and to have read documents in Georgian. The grammar of Old Georgian, say 5th century Martyrdom of Shushanik, is exactly the same as modern Georgian. It is easily understandable to a modern speaker, and most of what is ununderstandable to modern speakers are technical terms. Once again, considering Kartvelians started moving to Eastern Georgia in MBA, the foreign influences on Georgian are quite apparent, primarily it would be Hurro-Urartian influence.

You comparing Hindi to English and English to German is only proving my point. These languages formed similar to how Georgian formed, as a result of unification of several different genetic and linguistic groups. Meanwhile, I am pointing out the obvious facts that there was no such event in Western Georgia as far as we know for the last 10k+ years, which means that it is highly probable that Svan or perhaps Megrel will be somewhat understandable to the people of Kotia.

zardos said...

@Rob: The problem with your idea is that you have to assume a small elite, which dominated for a short period of time a steppe-related group in a rather punctual event and then largely disappears again, with R1 steppe folks carrying on the torch after the transmission. I don't think that's parsimonious at all, but runs against the evidence we have (so far).

CrM said...

@Rob

I'm not denying that Iran_N in Kartvelians comes from Maykop, KAC or something similar. What I'm trying to point out is that NWC have very little or zero Iranian-related ancestry, thus KAC and Maykop is not a good source for them.

I don't have the coordinates for Dolmen, so I have to resort to Dzharkutan2 I5608 as a base for NWC, and NWC model extremely well with I5608, which incidentally has no Iran_N ancestry.

Target: UZB_Dzharkutan2_BA:I5608
Distance: 2.7122% / 0.02712236
63.6 GEO_CHG
27.0 Anatolia_Barcin_N
5.6 RUS_AfontovaGora3
3.8 Levant_Natufian
0.0 IRN_Ganj_Dareh_N
0.0 ITA_Grotta_Continenza_Meso

https://imgur.com/a/YHN3keX


zardos said...

Concerning haplogroup L3, I found the study which pointed to the L3 origin in Eurasia from 2018. Like the recent results from Shum Laka and with IBDmix, this study argues for an Eurasian origin of L3:
"The coalescence ages of all Eurasian (M,N) and African (L3 ) lineages, both around 71 kya, are not significantly different. The oldest M and N Eurasian clades are found in southeastern Asia instead near of Africa as expected by the southern route hypothesis. The split of the Y-chromosome composite DE haplogroup is very similar to the age of mtDNA L3. An Eurasian origin and back migration to Africa has been proposed for the African Y-chromosome haplogroup E. Inside Africa, frequency distributions of maternal L3 and paternal E lineages are positively correlated. This correlation is not fully explained by geographic or ethnic affinities. This correlation rather seems to be the result of a joint and global replacement of the old autochthonous male and female African lineages by the new Eurasian incomers."

https://www.ncbi.nlm.nih.gov/pubmed/29921229

This is in full agreement with the new paper and the IBDmix results which propose a major back-migration from Eurasia to Africa.

Archi said...

@KG "Svan or perhaps Megrel will be somewhat understandable to the people of Kotia"

"is pseudo-science"

Chad said...

Don't use autism as an insult. Grow the fuck up and have a bit of decency. As for anything being a CHG language, that's bull. CHG was barely visible in Chalcolithic Armenia and the same in Maykop. Anatolia ChL gets the same stuff that came into Armenia and Maykop. Not ancestral to Meshoko. CHG is almost nil since Maykop times. Not very knowledgeable indeed.

Romulus said...

All that's left is for Gimbutas to rise from the grave and disavow the Steppe theory, then that will be the lot of them.

TLT said...

Regarding the J2 discussion, there is something to note here: the Hotu J and the CHG J are diverged by 18,000 years. Kind of corresponds to the end of the Baradostian and the end of the LGM.

18,0000 years BP still means a near 8,000 year gap between Hotu and the split. So even the common ancestry was separated for a long time.
And that is the common ancestry alone, idk about the different ghost populations which would have separately contributed to both Hotu/Iran groups and CHG groups.

Andrzejewski said...

If NWC are descendants primarily from Anatolia farmers, then their languages must bear at least SOME basic connection to the languages spoken by the LBK, Etruscans, Raetians and/or Ötzi

JuanRivera said...

Don't EHG and Khvalynsk have some J?

Andrzejewski said...

No, but Karelia and Samara HG do

JuanRivera said...

EHG is the cluster formed by Karelia HG, Samara HG, and Sidelkino HG.

Rob said...

@ Zardos

“ the problem with your idea is that you have to assume a small elite, which dominated for a short period of time a steppe-related group in a rather punctual event and then largely disappears again, with R1 steppe folks carrying on the torch after the transmission. I don't think that's parsimonious at all, but runs against the evidence we have (so far).”

Huh ? Disappear ? Z, observe the data to avoid straying into la/-la -Land
W.r.t. lower Don - it seems you clutching at straws, so I'll inform you. In the pre-Yamnaya Eneolithic it was the Konstantinovka cultlure, significantly under Majkop influence (the limits of its penetration toward the steppe); so it might have harboured people like Ukr. Ozera. The big (eastern proto-) Yamnaya expansion is thought to have come from Repin, which is actually in the Dnieper-Don southern forest-steppe, not lower Don. And this was heavily Z2103, so certainly not the defining element of the kurgan cultures. So let's put your discussion to rest.

Rob said...

@ CRM

Yes but isn’t that because Dzharkutan2 is from the Caucasus ?

So the low Iran N in NWC could well be due to highly local ancestry ~ Meshoko (CHG+ ANF-like)
But until the SS genomes come out it’s hard to understand the “base “

@ TLT
Yes J2a must have a long split time. But that doesn’t equate with mutually exclusive distribution range

zardos said...

@Rob: As I told you before, I'm talking about LDC in the early phase, in which the EHG-CHG improved hunter-fisher culture, yet alone pastoralists, were not present in the East imho. They spread FROM the Azow Sea area upward and branched in different groups, of which some became important for later developments, others not as much. But all had their root in this North Pontic region.

capra internetensis said...

@TLT

CHG hsve J1b-Y6305 and J2a3-Z6046, which split by 29 000 years ago. Kotias and Satsurblia still manage to form a CHG cluster. You can't tell population split times from a small sample of uniparental markers like that.

TLT said...

@capra
Alright, not using uniparentals shows us that they were quite different regardless. About as different as modern day northern Europeans are from modern day Levantines.

For instance, Hotu has more basal and ANE than CHG does, while CHG is more Anatolia-like than Hotu is.

TLT said...

@Andrzejewski

When was J found in a Samara HG? I thought that there was one in Karelia and one in Popovo and that was it.

Davidski said...

There's a J1 in a Khvalynsk sample.

David Anthony thinks it might be from Iran, but the chances of that are slim to none IMHO.

Andrzejewski said...

What if the real land of the Indo-Europeans was under the Black Sea, inundated?

Andrzejewski said...

https://en.m.wikipedia.org/wiki/Black_Sea_deluge_hypothesis

The source of Noah, Gilgamesh legends

Rob said...

PM
@ Zardos

'' I'm talking about LDC in the early phase, in which the EHG-CHG improved hunter-fisher culture''

I see, so possibly an early diffusion of CHG to lower Don (which dates c. 6000 BC at least) from Caucasus, within hunter-gatherer context. My guess would be we'd have seen it permeating further afield, but it might have remained there relatively isolated.

@ Davidski

''There's a J1 in a Khvalynsk sample.''

? And Popovo.


@ Andrze
In the page you linked, the sea-level rise was cmplete by 7600 cal BP. That's considerably before what most would place when PIE formed and expanded

Rob said...
This comment has been removed by the author.
zardos said...

@Andre@Rob: The idea I heard and found reasonable is that the CHG migration happened along the coast before the sea level was raising. They reached the North Pontic that way and stayed in the LDC context. But their path North is now under the sea level.
In the Azow Sea area primarily the fusion with local EHG clans happened, which became the dominant part.
In this context I would assume a higher J frequency initially (8-7 mill), but decreasing afterwards, until they virtually disappeared in the clan wars.

So the J1 in Khvalynsk could be just one of the survivors in the late phase, when R1 was otherwise completely dominant.

This relates to the African situation, where the same happened with E in the Sahara context. The Green Sahara had the same role as the steppe for Europe and (now!) SSA E the role of R1 for Niger-Congo/Bantu. There too local mtDNAs (L0, L1) were taken up and the local paternal lineages (A, B) were largely replaced.

Its not that complicated, the same processes were repeated throughout the World.

Andrzejewski said...

@All https://phys.org/news/2020-02-hot-pots-ancient-siberian-hunters.amp

Does it also pertain to the EHG/ANE side of our PIE ancestors, or not?

Andrzejewski said...

What if the role of Indo-European females has been greatly underrated? Let’s presume for a minute that most U4 and U5 found in modern European populations could be attributed to PIE? In this case, then WHG contribution to Europeans is vastly exaggerated whereas PIE is pretty much underestimated.

Andrzejewski said...

@zardos “This relates to the African situation, where the same happened with E in the Sahara context. The Green Sahara had the same role as the steppe for Europe and (now!) SSA E the role of R1 for Niger-Congo/Bantu. There too local mtDNAs (L0, L1) were taken up and the local paternal lineages (A, B) were largely replaced.”

Idk about this “Green Sahara Pump” theory, for this reason: Bata/Twa etc Pygmy HG in Central Africa look almost exactly phenotype-wise to West African Niger-Congo, the “negroid” appearance. Therefore, the assumed 30% Eurasian contribution to black Africans seems way over the top. Besides, West African culture and languages, along with lifestyle differ completely from anything Eurasian, being it European or West Asian.

Vladimir said...

@Andrzejewski. In any case, thick-walled pots, suitable for cooking food in them on fire, appeared in the Trypillya culture precisely as an import from the Sredniy Stog culture. Tripoli's own ceramics were thin-walled and not suitable for cooking over a fire.

zardos said...

@Andre: According to the Neandertal introgression paper the Eurasian proper admixture could be estimated at about (low estimate) 30-40 percent in modern Subsaharans. This is not a full scale replacemnt, but a significant admixture. Replaced were the paternal lineages, not the maternal. Pygmies have less admixture, but they have it as well and are of course quite different from West Africans phenotype wise.
The original foragers of Western-Central Africa, in the Shum Laka paper labelled Ghost modern + archaic, don't exist in an unadmixed form any more. Add to that selection and time to form new phenotypes and the fact, that we don't deal with modern European's back migration.

When "the Eurasians" moved into the Sahara and Subsaharan Africa, that doesn't equal a modern European phenotype migrated and they moved in directly from the Near East anyway. Remind you on the time frame and what happened in Europe at the same time. We are not talking about modern Caucasoids in the first place, but Eurasian ancestry.

Later Eurasian migrations brought more clearly Caucasoid phenotypes into Africa, like we see them in North and East Africa. But that admixture even through the Sahara was much earlier and it was about one third plus of the total ancestry in moderns.

Paternal replacement doesn't equal autosomal and phenotype change, look at R-V88 in West Africa and N in Finland. Remember that for Europe as well, because even though R1 yDNA spread like wildfire, the original EHG phenotype was by far not as dominant in its daughter populations, rather Neolithic, CHG and new phenotypes emerged after the fusion. The classic Eastern forager type is rather rare in modern Europe.

YDNA is like a surname in a patrilinear society. It will survive regardless of with whom you mix and what kind of selection works on your descendents if the line being unbroken. This can mean that the opposite of your phenotype might be seen in a direct descendent of your lineage with the same surname in just a couple of generations.

Andrzejewski said...

@zardos “Remember that for Europe as well, because even though R1 yDNA spread like wildfire, the original EHG phenotype was by far not as dominant in its daughter populations, rather Neolithic, CHG and new phenotypes emerged after the fusion. The classic Eastern forager type is rather rare in modern Europe. ”

That’s interesting: how do you know what all these people looked like? We don’t know what WHG, ANF, CHG or EHG I’m their unadmixed form look like, and any reconstruction e.g. Ötzi, Cheddar Man or anything else is deemed by @Davidski as “useless”?

zardos said...

@Andre: The reconstructions are not useless, but they are unreliable indeed. However, we have the skulls, measurements and first genetic traits, especially on pigmentation. The craniometry and morphology of early Eastern hunter gatherers is very, very rare in moderns. They were extremely hypermorphic in the Russian terminology and showed details which too are rather rare in moderns. You see a decrease from the oldest, to DDC -> Yamnaya -> CWC -> Bronze Age. The biggest jump imho was from early HGs to Yamnaya and CWC already, so the intitial admixture with CHG and Neolithics, as well as the new lifestyle, made them different already.

TLT said...

@Andrzejewski

I doubt that flood myths have a single origin. Almost all major human settlements during the transition from the end of the ice age to the Holocene were located near water bodies that were in some ways connected to glaciers and/or oceans. As a result, almost all of them would have experienced a rise in sea levels.

capra internetensis said...

In many parts of the world you can see fossils of fish or shells in the rocks - so clearly water used to cover the mountains. Right conclusion, wrong mechanism.

Ric Hern said...

If I remember correctly the Black Sea mostly expanded towards the West and North. The Caspian Sea on the other hand more than doubled its size, mostly towards the North and East...

Vladimir said...

Interesting article: African climate response to orbital and glacial forcing in 140,000-y simulation with implications for early modern human environments.
Indeed, the model showed large increases in rainfall and vegetation at 125,000, at 105,000, and at 83,000 years ago, with corresponding decreases at 115,000, at 95,000 and at 73,000 years ago, when summer monsoons decreased in magnitude and stayed further south.

Between roughly 70,000 and 15,000 years ago, Earth was in a glacial period and the model showed that the presence of ice sheets and reduced greenhouse gases increased winter Mediterranean storms but limited the southern retreat of the summer monsoon. The reduced greenhouse gases also caused cooling near the equator, leading to a drier climate there and reduced forest cover.
https://www.pnas.org/content/early/2020/01/15/1917673117

Andrzejewski said...

@zardos “The biggest jump imho was from early HGs to Yamnaya and CWC already, so the intitial admixture with CHG and Neolithics, as well as the new lifestyle, made them different already.”

I’m starting to think that maybe Ötzi and his ilk were responsible for the Alpine cranial type of Central Europeans, and that the reason Greeks and Italians deviate from that particular phenotype is due to later admixture with non-Steppe CHG as well as with Levant_N.

And the reason Spaniards don’t look that way (for the most part) is because they have 25% WHG and a minor (5%) North African aDNA.

zardos said...

@Andre: "Ötzi" was not even brachycranic, but had a rather long mesocranic skull. He was not particularly typical for Neolithic variants anyway. Most of the cranial shape change happened later and was definitely not more correlated with European early Neolithic populations.

What you have to realise is that we talk about different eras. There happened a lot in between. Even if you descend from people X, after 5.000 years, a lot can change. Generally speaking, people looked similar to modern Europeans, but they were not the same. Even if they would pass in a modern population individually, if you would meet 5 WHG, EHG or even Neolithics as a group today, you would wonder where they are coming from. Especially in the case of the early EHG, you might actually prefer to change the sidewalk at first sight even, because transferred into modern living conditions, they would be just huge, very robust and musculous and look fairly intimidating with large heads, faces and hands.

Don't try to explain everything with ancestral components when we have thousands of years in between. New mixtures and selective trends emerged all the time. The idea of selection and evolution coming to a halt at some point, maybe in Neolithic times, is completely erroneous. Even on the contrary, the selective changes and turnarounds became more frequent, because the way of life changed all the time too. With more data you will see it in the genetic analyses too, first results appeared in different papers already.

Andrzejewski said...

@zardos “He was not particularly typical for Neolithic variants anyway.“

I read a study that found out that Ötzi’s group was the closest genetically to those of “Agean Farmers”.

Andrzejewski said...

@zardos “Most of the cranial shape change happened later and was definitely not more correlated with European early Neolithic populations.“

So what did they look like?

Spanked said...

@zardos "in the case of the early EHG, you might actually prefer to change the sidewalk at first sight even, because transferred into modern living conditions, they would be just huge, very robust and musculous"

There's a lot of info about EEF, WHG and even ANE phenotypes out there, but EHG, CHG and SHG are still a bit of a mystery to me. Particularly EHG. Do you guys have any good links for a layman's layman to read on the subject?

Andrzejewski said...

@Spanked “There's a lot of info about EEF, WHG and even ANE phenotypes out there, but EHG, CHG and SHG are still a bit of a mystery to me.”

To me too!

Plus, info regarding WHG/EEF are contradictory at best

Andrzejewski said...

I surmise that the ANE phenotype is close to what the Yenisseyan Kett look like but with all the genetic drift, who really knows?

AWood said...

My own view is that the EHG phenotype looked like Uralic/Russian people. Wider, slightly shorter skull with narrow eye slits. Most definitely the long head that is common in Europe came with the Neolithic people, as did thicker wavier curly hair. I'd anticipate the nose also tends to be more narrow and the bridge lower among EHG. Modern eastern Russians cannot be used for comparison, nor can Native Americans as they have substantial admixture from the East Asian populations who at this time would have looked quite different.

Modern Europeans have Caucasus and Western Middle Eastern ancestry which have most definitely adjusted the European face. I admit this is most likely conjecture but I'm looking at averages of modern populations to see how the face could have changed over time. Certainly no European is pure EHG, and those closest I guess are in the northern part of European Russia, west of the Urals.

zardos said...

We don’t know the soft parts too well obviously, but the skulls and bones of early Eastern hunter gatherers are informative. What most mean with "Uralic" equals Mongoloid affinities.
While these are present in EHG in some individuals and places, they stick out and are not representative. Whether those are the result of more ancient or recent admixture from the East, I don’t know, but it looks to me recent.

The typical EHG was strongly Caucasoid, but of a now rare phenotype. The most typical characteristic was the massiveness of the bones and the sheer size in virtually all measurements. E.g. their faces were quite broad and long in absolute numbers, most of todays Europeans are broad OR long. Some modern skulls look almost like those of kids in comparison.
Like in recent studies on genetic height, they proved that the steppe admixture on average increases body height. Now we know where this was coming from. The absolute cranial measurements in Northern Europe too are larger than in the South/Sardinians.
But this is the result after mixture and thousands of years of selection. The original EHG would be just larger and more robust than any modern European people.

Pine trees and mountains said...

@zardos

I have heard from some Russian anthropologist that Caucasians and especially North Caucasians have both one of the longest and one of the widest faces in the world. Indeed when you google Chechens or Georgians their heads are pretty big.

«Oldest ‹Older   1 – 200 of 233   Newer› Newest»