search this blog

Saturday, March 10, 2018

Was Ukraine_Eneolithic I6561 a Proto-Indo-European?

It's certainly a valid question, simply because the remains of this individual (sampled by Mathieson et al. 2018, see here) are from a cemetery of the Sredny Stog culture, which, based on historical linguistics and archaeological data, has already been posited to have been a Proto-Indo-European (PIE) culture, that gave rise to the supposedly Late Proto-Indo-European (LPIE) Yamnaya culture, that swept into Central Europe from the Pontic-Caspian steppe during the 3rd millennium BC. Moreover, consider the following points:

- whatever you might say about calling Y-Chromosome haplogroups "Proto-Indo-European", the fact is that Ukraine_Eneolithic I6561 is the oldest recorded individual belonging to Y-haplogroup R1a-M417, which is not a marker that can be reasonably linked to human expansions dating to the Paleolithic or even Neolithic, and yet today it peaks in frequency in modern-day Indo-European-speaking East and North European Europeans and South Asians, and is also recorded as the main Y-haplogroup amongst the ancient Scythians, who also were, in all likelihood, Indo-European-speakers, which strongly suggests that it was initially spread far and wide across Eurasia by the early Indo-Europeans

- following on from the last point, R1a-M417 can be divided into three main subclades: R1a-L664, R1a-Z293 and R1a-Z282, the first of which is almost exclusively confined to Northwestern Europe, while the latter two peak in frequency in South Central Asia and Eastern Europe, respectively, and the really interesting and important thing is that R1a-Z93 and R1a-Z282 are more closely related to each other than either is to R1a-L664, which mirrors the relatively close linguistic relationship between Balto-Slavs, who are rich in R1a-Z282, with Indo-Aryans, who are rich in R1a-Z93, (for instance, see here) and renders any arguments in this case based on isolation-by-distance practically useless

- Ukraine_Eneolithic I6561 is the oldest sample with UDG-treated genome-wide data to carry the 13910*T lactase persistence allele, which reaches its maximum frequency in Northwestern Europe, and is also relatively common amongst Indo-European-speaking South Asians, but not Middle Easterners (see here), suggesting that it spread from the Eastern European steppes both into Northwestern Europe and South Asia along with such ancient steppe markers as R1b-M269 and R1a-M417, and Indo-European speech

- based on historical linguistics data, the Proto-Indo-Europeans are generally regarded to have been foragers turned pastoralists, rather than farmers, but nevertheless, pastoralists familiar with farming, and indeed Ukraine_Eneolithic I6561 appears to be mostly a mixture of Eastern European and Caucasus Hunter-Gatherers (EHG and CHG, respectively), but with around 30% input from early European farmers.

Of course, we'll need many more ancient samples from Ukraine and surrounds to cement these findings, and prove, beyond any reasonable doubt, that the Sredny Stog people were indeed the Proto-Indo-Europeans, and that the Yamnaya people were the Late Proto-Indo-Europeans. It might also be necessary to develop new scientific methods that take into account multidisciplinary data to achieve this.

On a related note, the University of Leiden is currently seeking four historical linguists and one bioarchaeologist to take part in a new project titled The Linguistic Roots of Europe's Agricultural Transition. The principal investigator on the project is Guus Kroonen, whom I mentioned in a couple recent blog posts (see here, here and here). This is the project objective:

Today, Europe’s linguistic landscape is shaped almost entirely by a single language family: Indo-European. Even by the dawn of history, a patchwork of Indo-European subgroups, Germanic, Celtic, Italic, Baltic, Slavic and Greek, was covering the continent, and over the centuries, these subgroups evolved into the modern European languages, among which Russian, Italian, German, Lithuanian and Swedish, as well as the global lingua francas French, Spanish, and English.

The Indo-Europeanization of Europe was probably one of the most profound linguistic shifts ever to have taken place in the prehistory of Europe. The origin of the European languages, unsurprisingly, is therefore a matter of intense academic debate. There are currently only two prehistoric events that in the present academic debate are considered as likely driving factors behind the spread of Indo-European speech.

One the one hand, there are those historical linguists who by meticulous comparison of the different Indo-European languages have reconstructed a language and culture that is typical of the early Bronze Age. Terminology for horse-riding and wagon technology provides a possible link to the expansion of the Yamnaya culture on the Pontic-Caspian steppes, which was fueled by the invention of the wheel and the domestication of the horse. Others have suggested that the Indo-European languages diffused from Anatolia together with another major prehistoric event, the spread of agriculture to Europe between the 8th and 5th millennium.

The debate has remained unresolved for over two decades, but a new approach produces potentially decisive results. By studying prehistoric loanwords absorbed by the speakers of Indo-European when they entered Europe, and test the resulting cultural implications against the available archaeological record, new light can be shed on the language of Europe’s first farmers, and whether or not they spoke a form of Indo-European.

If you have the necessary passion and qualifications to apply for these positions, then please do so ASAP via these links:

PhD Candidate or Postdoctoral Researcher in the field of linguistics

Postdoctoral Researcher in the field of archaeology (specialization: bioarchaeology)


Chetan said...
This comment has been removed by the author.
Al Bundy said...

Why are they wasting their time and money on Anatolia and wondering if the early farmers spoke IE? Are they only talking about European IE or all IE?Do they know IE languages aren't only in Europe?Yes good post by Davidski but the terminology in that call for candidates seems sloppy.

Samuel Andrews said...

Wow, I didn't know he had 13910*T. That's a big deal. I would be PIEs were the common ancestor between Yamnaya and Corded Ware who could have lived in 5000 BC. But I don't know I'm not into lingustics.

If I6561 is representative of Sredny Stog, it is unlikely they contributed to Yamnaya. Because of his R1a M417 Sredny Stog must only be an ancestor of Corded Ware. To me it would make sense PIE language formed before R1a M417 and R1b L23 became popular.

Matt said...

He's a pretty interesting sample. Will say, having looked at him in G25 and in Mathieson's data, my opinion is skeptical of seeing him, and his specific autosomal mix, as actually directly ancestral to recent Europeans *instead* of Yamnaya-like people, for a few reasons:

1) Actually this sample is earlier than the other Ukraine_Eneolithic samples, presuming they've got the dating right and he belongs to the layer he's found at (always an issue with single samples).

He was postdated by the Ukraine Yamnaya, who seem more Yamnaya like, and also the other Ukraine_Eneolithic with their Ukraine_Meso affinity, which suggests that this person's specific autosomal mix did not last in Ukraine (whether it was a question of absorbing other cultures, or being absorbed or replaced).

(Of course it is possible that his population, if it was indeed a stable population with this mix, may have moved elsewhere, or population genetic structure may have persisted within Ukraine).

2) Later samples from across the Bronze Age European horizon seem to vary from closer to Yamnaya than he was, to being much closer to various European Neolithic groups, which does support admixture in situ between Yamnaya like, or at least Early Corded Ware, like and Neolithic cultures.

Plus those samples, at least from Beaker, have specific D-stat linkages to specific European_MN cultures (e.g. Poland LN generally best fit), which would support a fair amount of local absorption.

3) The specific position of I6561 does seem to me to be more likely to fit along a Samara_Eneolithic->Balkans_CA like cline, rather than exactly the same as cline as later Bronze Age samples from North-Central Europe... but this is very fine distinction and it's a single sample.

So I would think that this sample may be from early Indo-European speaking sample, just one that may not be directly ancestral to Europeans in Europe. (Rather a mixture that existed on the margin in his time, then was recreated on a more substantial scale later in history).

I would guess that this adds some evidence to support the idea that some cultural ideas and geneflow went "SE European 'Old Europe'"->Sredny Stog II->Yamnaya/Early Corded Ware (in two slightly varying directions), with some migration yet without much "mass migration", and then the end of the sequence reticulated back into Europe? At the same time, some southern, more CHG like source was feeding into what would become Yamnaya and early Corded Ware.

Some graphics I put together from resampling the Global 25 to try and visualise where this sample sits, in the context of other ancient Europeans and some recent ones:

supernord said...

In any case, it can be argued that the Dereivka culture was Indo-European.
Because this sample is from Dereivka culture.

Rob said...

I think a more pertinent question was how many languages were spoken in the Eneolithic - BA steppe.

Davidski said...

@Samuel Andrews & Matt

It's even unlikely that I6561 is a direct ancestor of Corded Ware, because, for one, his bam file suggests that his Y-DNA belongs to Z93+ and sitting on the branch leading to the "Indian" L657.

But this isn't the point. The point is that even though I6561 doesn't look like the perfect sample for modeling the PIE expansion (how can he be, if there was some diversity in the PIE community, and, for example, one person can only carry one Y-haplogroup?), he has a very peculiar set of qualities that many, including me, were expecting to be present in PIE samples.

One should expect some diversity in the PIE community, although obviously the right kind of diversity, before the LPIE expansions and founder effects.

So when other Eneolithic samples from Ukraine and nearby parts of south Russia come in, we might see individuals and groups with somewhat varying autosomal profiles, and Y-chromosome lines ancestral to the Corded Ware R1a-Z645 and R1a-L664, the Kalmykia/Samara Yamnaya R1b-Z2103, the Bell Beaker R1b-P312, and so on.

ryukendo kendow said...

Isn't it more rational to suppose that this R1a individual and his collaterals were assimilated into the expanding Yamnaya population? Or that his collaterals were present in the Yamnaya-ized population since the beginning? Than that this person specifically is a proto-IE proper. At best a para-IE methinks.

@ Davidski

Do you think you have enough res to perform haplotype analysis? We know for long periods in W Eurasia different HG groups, and HG and Agropastoral groups were extremely reproductively isolated, but we can see here that in SE Europe the boundaries were potentially breaking down (very novel social scenarios happening for sure), but we don't know for sure this is the case (aka what if it was just like CHG in the Caucasus which is in between Iran_HG and EHG but stayed that way for 10,000 years and is not a recent product of mixture?) Or, if this was the case, we don't know how recent this mixing was.

Haplotype analysis of all the SE Euro groups would be v, v interesting, would let us overlap artifacts and genealogies.

P.S. if you could keep track of some of these openings that come to you through the grapevine also for those who are much earlier down the academic track, it would be very appreciated indeed.

Davidski said...

@ryukendo kendow

Isn't it more rational to suppose that this R1a individual and his collaterals were assimilated into the expanding Yamnaya population? Or that his collaterals were present in the Yamnaya-ized population since the beginning? Than that this person specifically is a proto-IE proper. At best a para-IE methinks.

Sredny Stog is commonly regarded in academic works to have been PIE, or at least part of the PIE horizon, and also in part ancestral to Yamnaya. I'm just pointing this out, and also that this sample has qualities expected of someone from the PIE community.

The Yamnaya expansion doesn't preclude this sample from being PIE, since Yamnaya isn't considered to be PIE, but rather Late PIE, just like the more recent IE (Slavic) re-expansion into the region doesn't preclude Yamnaya from being Late PIE.

Do you think you have enough res to perform haplotype analysis?

I won't be running haplotype analyses of pseudo-haploid samples again, only real diploid genomes. The results for diploid genomes are much more reliable, even when most of their data are imputed.

namedguest said...

[1] "distance%=3.3889"



[1] "distance%=2.8995"



[1] "distance%=4.4146"



[1] "distance%=4.7776"



[1] "distance%=4.355"



namedguest said...

[1] "distance%=4.554"



[1] "distance%=4.2704"



Nirjhar007 said...

If he is not a result of contamination (unlikely) he can be related to PIER ...

Samuel Andrews said...

IMO, PIE must be a common ancestor between Sredny Stog & Yamnaya. UkraineHGs don't look like an important ancestor but they may be. That info along with CHG admixture, to me suggests PIE lived east of Ukraine and near the Caucasus.

It would make sense PIE lived around 5000-6000 BC. They would have had a mix of many yHGs including R1b & R1a. Then due to some cultural practice, R1a M417 became extremely popular in one clan and R1a L23 became extremly popular in another clan. The rest is history.

Btw, based on mtDNA it looks like Catacomb was of largely Ukraine HG origin. It'll be interesting to see what Y DNA they had.

Samuel Andrews said...

Think about it. Shouldn't PIE be the common ancestor between Corded Ware/Yamnaya, Yamnaya/Sredny Stog?

I can't imagine that separate R1a M417 and R1b L23 communities adopted one group's language without geneflow or that their PIE dialects evolved together due to interaction. If you're a die hard Steppe fanboy, all I6561 shows is PIE is older then people used to think. Maybe 2,000 years older than Yamnaya.

If so, it would make sense Corded Ware spoke something leading up to Balto Slavic and that Bell Beaker something leading up to Celtic or maybe even proto-Celtic. Maybe SIntashta & Andronovo spoke proto-Iranian or even the a line leading up to Sycthian.

Samuel Andrews said...


What do you think of the Kurgan hypothesis now that R1a Z93 may have just been found in Sredny Stog?

Davidski said...

Eneolithic/Early Bronze Age human and horse samples from around the Don, and just east of it, will be crucial in all of this, no doubt.

They should be published soon, hopefully this year.

Nirjhar007 said...

What do you think of the Kurgan hypothesis now that R1a Z93 may have just been found in Sredny Stog?

Concerning R1a it can be really interesting !.


What is the source of the claim there any confirmation?.

Davidski said...

It looks like it was scanned by Yfull. See here...

Nirjhar007 said...

Thanks bro .

epoch2013 said...


Could you do:

Mbuti GoyetQ116-1 Karela_HG AfontovaGora3
Mbuti GoyetQ116-1 Karela_HG MA1


Mbuti Loschbour GoyetQ116-1 Kostenki14

Nirjhar007 said...

Z-93 clade is amazing :) it is now perhaps ''proven'' to be 1000 years or so older than thought or predicted . Perhaps this means that none of the M417 clades of CWC were of M417 but of derived branches (Z-282,L664,Z-93?) and if they were just M417, then they went extinct .

This increase of age is also compatible for probablilty of finding M417+, Z-93 in IVC .

Chetan said...

If I6561 indeed turns out to be z93+, then that is an interesting turn of events since he is dated to around 4000 BC (? I'm not sure). But there can be no doubt that the people of Sintashta at least in the earliest stages spoke a specifically Proto Indo-Aryan language rather than Iranian. It is present in the Uralic loan word layer before Iranian and the only culture that is likely to have had significant contact with the forest steppe (where pUralic is supposed ti have been spoken) is the Sintashta (at the earliest date). Later the Andronovo extended into the forest steppe/forests, but by then the lingua franca was already Iranic?

I don't think we are as yet close to getting definitive answers as each new study keeps bringing out intricacies in the prehistory of PIE that we were previously unaware of.

Folker said...

Does PIE emerged before the population found to have carried it outside the Pontic Steppe?
This individual was probably part of a population which took part of PIE population ethnogenesis, with cultural contribution as well (so linguistics ones). To reach the point to identify them to PIE is probably too far reached. Several languages could have been spoken in Pontic Steppe prior to the emergence of PIE (with the same linguistic group or not).

To date, we can only connect diffusion of PIE, and more widely to IE languages, to the expansion of a population Yamna-like. Expansion which began in the late IVth millenium BC. All posterior IE populations did have some Yamna-like admixture. This is also meaning that PIE could have only emerged with ethnogenesis of this specific population, not before.

Alberto said...

So trying to figure out the chronology of Ukraine, this is more or less what I see (corrections or comments welcome):

- Mesolithic: 9000-6000 BCE. HGs that are closer to EHG than to WHG. Y-DNA R1a, R1b and I2a.

- Neolithic: 6000-5000(?) BCE. There's a migration from the west, bringing more WHG (Iron Gates?) admixture from the Balkans. Increase in R1b and I2a lineages. Only one R1a found.

- Early Eneolithic: ~5000-4200 BCE. In some old school archaeology might be called Sredny Stog I. We lack samples from this period, really. So some extrapolation is needed. Apparently (my guess) is that this period is still dominated by migration from the west (Balkans), but this time it's not HGs, but farmers. Admixture levels between the farmers and the HGs probably quite diverse. Y-DNA unkonwn.

- Later Eneolithic: ~4200-3100 BCE. Also loosely called Sredny Stog II. In this period things change, since we see a migration from the east by a Yamnaya-like population (by this I mean a population that was largely a 50/50 mix of EHG and CHG). This is clear from the genetic data (the sample featured in the post, I6561) and archaeology alike (for example, D. Anthony chapter: THE SREDNI STOG CULTURE: HORSES AND RITUALS FROM THE EAST, here). The exact origin is unknown, and often suggested to be the Khvalynsk culture in the Volga steppe. Though we have Khvalynsk samples from the time that are too CHG-poor to be the exact source. Probably somewhere south of Samara could be. This population brought R1a-M417. Speculatively, it could have brought R1b-L23-L51, but that's totally unclear. The admixture in that earliest sample (4000 BC, east Ukraine) shows mostly EEF admixture, but it's just one sample, and later ones show Ukraine_N too.

- EBA: ~3100-2400(?) BCE. The Yamnaya proper period. A second wave from the east being mostly EHG+CHG comes in, this time displacing the previous population (to CWC area?) and moving beyond Ukraine to the Balkans. Uniformly Y-DNA R1b-Z2013, with some outlier I2a. Maybe R1b-L51 from a specific group that made it to Western Europe?

Arza said...

@ epoch2013

Out of curiosity I ran earlier first two quadruples (with some other ones added):

epoch2013 said...


Mbuti.DG GoyetQ116-1 AfntovaGora3 MA1.SG 0.0276 2.939 7663 7252 162654

Wow, that is almost Z=3. GoyetQ116-1 is also having a relationship with Tianyuan.Thanks.

ryukendo kendow said...

Arza, matt, or any others, could you take advantage of the natural variability of the bell healer and corded ware samples and "back-project" to where their steppic precursors would be?

For corded ware I still think the presence of r1a carrying samples that are basically identical with yamnaya basically obviates the need for ukraine mesolithic contribution, for the bell heaker less sure.

Of course if IEs from Anatolia require Sredny Stog that's another thing entirely.

Anthro Survey said...


I mostly agree with your take.

re/Neolithic Ukr---Balkan HGs are probably it, but I wouldn't rule out WHG-shifted migrants from the forest zone upstream/north of the sampled location.

Sredny Stog 1---My suspicion also. In fact, that Neolithic Ukraine outlier probably wasn't such an outlier at all, but represented a major stream of ancestry for relevant parts of SE Ukraine.
I.E region was a checkerboard.

As for the CHG dosage: have you considered a formed hybrid EEF-CHG(~2:1) population circa Azov/Donets with which Samara_Eneo-like Khvalynsk mixed to produce someone like 6561?

If such a population existed, the question is whether it contributed to Yamnaya through reflux by folks like 6561 or not.

supernord said...


Your periodization of Ukrainian's cultures is not correct.

Early Eneolithic
1. Sredniy Stog I/Kvitian: ~5000-4500 BC.
2. Sredniy Stog II: ~4500-4200 BC.
Middle Eneolithic:
3. Dereivka: ~4250-3750 BC
Late Eneolithic:
4. <~3750 BC - A lot of different cultures.

Chetan said...

@Alberto Good summary Alberto. The late Eneolithic phase I think is the most crucial (Sredny Stog II by your terminology). It is proposed that this culture could have been Indo-Europeanized by influence from the east, the Suvorovo-Novadanilovka migrations that show up in the Balkans at around this time and which could have even branched off the groups that would become Anatolian speakers. After this event, the east and west could still have maintained linguistic contact until the formation of the Yamnaya culture under increasing contacts with the west and south.

a said...

" whatever you might say about calling Y-Chromosome haplogroups "Proto-Indo-European", the fact is that Ukraine_Eneolithic I6561 is the oldest recorded individual belonging to Y-haplogroup R1a-M417,..............spread far and wide across Eurasia by the early Indo-Europeans"

What do sample Ukraine_Eneolithic I6561[culture] have in common with Corded Ware culture and Swat Valley Aryan invasion[culture] burials?
No Aśvaḥ (अश्व)-aka horse
No Anas (अनस्)-aka cart
No Ratha (रथ)-aka spoked wheel chariots.

Why not use Eneolithic I6561 sample instead of R1b-Z2103{found in all three above regions} in constructing your calculations[in fact just to really narrow it down-don't use any R1b Kurgan Steppe-samples]; if you have firm conviction in your proving your theories.

Matt said...

Ryu, yeah, I'm with you on that; leveraging main vectors within group would make sense.

Here's one way of doing it.

Drop all Global 25 for all of Beaker samples into PAST PCA:

So that gives you a PC1 that describes the main vector of change within the Beaker samples, and a bunch of other vectors that are less meaningful / important, or describe the way geographically diverse Beakers differ from each other (PC1: 72.8%, PC2: 6.97%, PC3: 6%, etc.).

Then you can just use the scores and loadings on that to output a simulated sample for each "end" of the dominant Beaker cline that should be neutral, using the loadings to get something that is, in this case, probably -0.15 and +0.15 on PC1 and neutral on all other dimensions.

I'll have a go at the last step later if I have the time.

(It feels like a massive pain, but poss we could write a script to just do it at some point.)

Matt said...

Sample datings sequence for Ukrainian ancients from Mathieson's supplement (accompany Alberto's summary, with which I see wrong really):

Ukraine_Mesolithic - Various - 9000-6703calBCE (I5876 is the latest sample)

Ukraine Neolithic - Various - Earliest 6500 calBCE, latest 4343 calBCE

Ukraine_Eneolithic - I6561 - 4045-3974 calBCE (Sredy Stog II)

Trypillia - I3151, I2110, I2111, I1926 - 4000 calBCE (earliest) - 3640 calBCE (latest date)

Ukraine_Eneolithic - I4110 - 3634 - 3377 calBCE (Dereivka I, lower HG)

Ukraine_Eneolithic - I5882 - 3264-2929 calBCE (Dereivka I, higher WHG)

Ukraine_Eneolithic - I5884 - 2890-2696 calBCE (Dereivka I, higher WHG)

Yamaya_Ukraine - I2105 and I3141 - 3300 - 2700 BCE

Yamnaya_Ukraine_Outlier - I1917 - 3095-2915 calBCE

Globular Amphora Ukraine - ILK001, ILK002, ILK003 - 2900 calBCE (earliest) - 2694 calBCE (latest)

Note the Trypillia Verteba cave and GAC Ilyatka are pretty west of the other samples, which are relatively pretty close to each other.

Matt said...

@RK: Here are two pairs of zombie coordinates in Global25 (scaled) using that above procedure:

Both are a bit like imaginary composites. The "steppe_rich" simulation looks to be relatively richer in EHG+EEF (specifically) than the Early Corded Ware, this said. (See:

Alberto said...

@Anthro Survey

Yes, I guess that possibility of I6561 having its AN ancestry from the Caucasus cannot be discarded, though all the models I've seen seem to prefer European Neolithic. That, with the fact that we do have a population that lacks such admixture (Yamnaya), makes me prefer the other option. But we still lack samples to say anything definitive.


Labels are not very well defined. I used them in a very loose way, but I don't mind calling something Dereivka instead. The dates, I'm more or less guessing from the aDNA and adjusting the archaeological ones. Still, also the dates are quite loose. More samples will give us a more precise chronology.


Are we looking in the same place? I'm using the excel file in the supplementary data. What I see there is:

- Ukraine Mesolithic: latest sample I5885, 6392-5927 calBCE

- Ukraine Neolithic: Earliest sample I1736, 6248-6070 calBCE. Latest sample I4111, 4722-4548 calBCE (the only sample that is after 5000 BCE, so I think that with archaeology we can say that Eneolithic probably started earlier, though it didn't occupy the whole area of Ukraine. Then there are 2 samples without C14 dating that just say 6500-4000 BCE. and one sample I1378 that has a C14 date of 4519-4343 calBCE, but he's labelled as "son of I1732", which does have a New Date (unlike I1378) and it's 5364-5213 calBCE, so the first one seems wrong.

Matt said...

To be honest I glanced over it and captured it by eye, so possibly I skipped over a sample or two with younger / older dates than the youngest / oldest I saw. Feel free to correct. Youngest labeled date for Ukraine_N is I378, given 4519-4343 calBCE.

Screenshot of table here, all populations containing the word Ukraine or Trypillia, sorted in date order (using the average date):

a said...

" whatever you might say about calling Y-Chromosome haplogroups "Proto-Indo-European", the fact is that Ukraine_Eneolithic I6561 is the oldest recorded individual belonging to Y-haplogroup R1a-M417,..............spread far and wide across Eurasia by the early Indo-Europeans"

Why waste time, [endless hours debating/heated vexatious posts-unless you have nothing better to do ] on speculating. Just tweak the Ukraine_Eneolithic I6561 sample with a R1a zombie clone and add it to the powerful Global 25 that fits your model/s?


Alberto said...


Thanks. Yes, we're looking at the same table then. That I1378 is the one labelled as "son of I1732", which has a new c14 date of 5364-5213 calBCE. And with the Ukraine_Neolithic_Outlier (essentially a Balkan farmer) dated to 4949-4799 calBCE I think that 5000 BCE is in the ballpark of the transition to Neolithic farmers moving onto the steppe.


I don't know if modelling Bell Beakers and Corded Ware to see what kind of steppe source they take is exactly what you meant (maybe you mean creating a steppe "ghost" from the average of all Bell Beakers and another from the average of all CWC samples?). Since I did some runs with Global 25 using the real populations (against all individuals from Bell Beakers, CW and others), I'll leave the link here in case it's useful:

Essentially the steppe sources are Yamnaya_Samara, Yamnaya_Kalmykia and the 4 Ukraine_Eneolithic individuals (which are picked in the models almost unanimously). Removing the Ukraine_Enelithic samples mostly increases Yamnaya (Samara or Kalmykia) and adds Globular Amphora + Koros_HG (with models getting slightly worse).

ryukendo kendow said...

Aaah I see, thanks Matt and Alberto.

It seems that the earliest Corded Ware in the Baltics (which are like NW-shifted Yamnaya and less like later German CW) are shifted in the "impossible" NW direction with both WHG/EHG and EEF input from some unknown group, but with higher WHG ratio than any existing European Late Neolithic or Chalcolithic population.

It seems Ukraine_Eneolithic with maybe a bit less EEF will fit the bill nicely, which will also explain CW R1a. The EHG and CHG will then become "invisible" in the Yamnaya background but the EEF:WHG will still be visible.

How does Ukraine_Eneolithic look like if we get rid of the EEF? That may give us some clues.

Jaap said...

In the riverine landscape of origins there are always things that came before. From two or three different rivulets. Yamnaya is late PIE, there's no question. Celtic and Latin. Predominantly R1b. Greek and Armenian are a bit of a poser, but definitely steppe. But what about Anatolian? Steppe? If so Shredny Stog will do as your PIE-homeland.
But just imagine the questions this would raise! And just one specimen too! Why not call Shredny Stog late PIE, too? Just a bit earlier, and a different dialect ...
The common denominator seems to be CHG, there's no getting away from that ... Just stating the obvious. But the narrative to go with this is fascinating!
I like this post as it narrows down the Slav-Indian thing: and I must stop here as all the bubbles about trade/brides/metallurgy/liturgy/literature need to settle down to be scrutinised in earnest.

Davidski said...


The common denominator seems to be CHG, there's no getting away from that.

The common denominator in the context here that matters is steppe Y-DNA. CHG is just admixture mediated via women. See here...

Romulus said...

Steppe Y DNA isn't the common denominator because there are numerous populations high in steppe y dna, Basques, Central Asian Turkic speaking groups, etc. that have steppe Y DNA and don't speak indo european languages.

Rob said...

About B.B., I think there might be structure there (if I’m not confusing myself )

- the barnyard L51s who form a tightcuster therefore might represent the post-bottleneck expansion phase from CE

- a few Outlier (male L51) BBs

- the Z2103 BBs

- auxiliary evidence from contemporary non-BB R1bs (eg Nagyrev culture)

I put this though an iterative PCA and get a possible solution

@ RK
Early Baltic CWC is very similar to Yamnaya as are some of the early German CWC. It is later german cwc and Czech cwc which differ by acquiring local EEF.
Baltic BA subsequently acquires both EEF and WHG, possibly coming from Ukraine late Eneolithic- type group; or a Narva / Hungary combination .

Nirjhar007 said...

PIE is perhaps older than most of here imagine....

Ric Hern said...

Is there any method by which the Contribution (if any) of CHG languages to the formation of PIE can be accurately measured ?

If CHG did not contribute much Linguistically then any or all on the Steppe before Yamnaya could already have been Proto-Indo-European.

Apparently some Linguists determined that links between Languages within a Specific Language family can be as old as 10 000 years. So anything on the Steppe from 8000 BCE could have been identifiably related to PIE.

Davidski said...


Steppe Y DNA isn't the common denominator because there are numerous populations high in steppe y dna, Basques, Central Asian Turkic speaking groups, etc. that have steppe Y DNA and don't speak indo european languages.

Turkic R1a is the same as Andronovo, Sintashta, Srubnaya and Scythina R1a, so I'm guessing you're claiming now that these groups were Proto-Turkic.

If so, good luck with that.


PIE is perhaps older than most of here imagine....

Nah, the consensus is that PIE is an Eneolithic language. This of course matches the ancient DNA record, which shows massive expansions from the steppe from the Eneolithic onwards.

Nirjhar007 said...


Arza said...

@ ryukendo

First I modelled them (Beakers and others) as EHG, Ukraine_Mesolithic, CHG, Armenia_EBA, Globular_Amphora:Poland, TDLN, Blatterhole_MN. Then I recalculated fitted points from EHG, Ukraine_M, CHG and Armenia_EBA scores.

small + - Beaker ghosts
large + - CWC ghosts
X - Czech_EBA ghosts
[ ] - CWC_Baltic_early
O - Yamnaya

PC2,3,5 rotated to provide best view with GAC->Central Europe->Steppe MLBA as a vertical line. Matt's consensus points included.

Two CWC + that deviates heavily towards Ukraine Mesolithic are Spiginas2 and I1540. Both can be safely ignored, as they require WHG that I intentionally didn't include to not trigger WHG-CHG mix for most of the samples. Beaker ghosts that follow them possibly also have more WHG than Blatterhohle_MN can provide.

There are also samples heavily shifted towards Armenia_EBA, most probably via Balkans (the one between CHG and Armenia is Sicily_no_steppe).

Majority of the ghosts form one big steppe cluster which shows that Yamnaya was in fact a peripheral population that did not contribute much to Beakers, CWC nor Czech_EBA (Hungary, especially outlier may be an exception).

At the same time it's obvious why CWC_Baltic_early is a better proxy for steppe ancestry than Yamnaya.

PS A pic that I posted long time ago, based on some observations from G10:

Vara said...


Do you deliberately try to misunderstand my comments?

Maykop could've been all J2 and that still won't disprove a male migration. It has been explained many times how the CHG Y-DNA could've went extinct after a few centuries in the comments, so I don't think you ignored all that and I don't want to repeat it anyway. In short it probably became extinct the same way Corded Ware R1a1a1 and Karelia J1 went extinct in Eastern Europe.

Also, I'm not sure why you are terrified of the idea of R1b in Maykop. They probably had some EHG ancestry.

Davidski said...


The idea that CHG admixture made it onto the steppe via female mediated gene flow makes the most sense, because we have samples from the Eneolithic/Bronze Age steppe with low level CHG but with southern mtDNA.

Pretty simple but robust inference. And even R1b in Maykop won't contradict it, unless it's M269+.

And of course all Corded Ware R1a lineages didn't go extinct. What an idiotic thing to claim.

It seems like you missed the most recent scientific literature, which shows that modern-day Eastern European R1a (Z282+) derives from Corded Ware via Trzciniec.

Vara said...


"PIE is perhaps older than most of here imagine...."

I disagree. The way PIE is defined makes it less than 7kyo. Many people have skipped over the basics in the PIE debate, so I will explain.

I think the easiest way to explain the age of PIE is with the myth of the Dragon Slayer World-King. The Dragon Slayer myth is common in all IE cultures including Anatolian. He is worshiped as the Sky God in all IE cultures except Iranian and some versions of the IA myths. But how does a snake killer helps us determine the age of PIE? The answer is simple: he doesn't, his weapon and helper(s) do. In all versions of the story he wields a mace or an axe (Anatolian and Slavic) that later becomes thunderbolt, and is helped by a smith or group of them or fathers a smith god.

Lets start with the Indo-Aryan version:

"When the clever Tvashtar had turned the well-made, golden, thousand-edged vajra, Indra takes it to perform his manly deeds ; he slew Vritra, he forced out the stream of water." RV (1.85.9)

Indra's golden vajra(thunderbolt) is clearly a bronze mace same as the Avestan wazra which is the mace used to kill a 3 headed dragon. In another version it is Trita Aptya who kills the three headed snake which is more similar to the Iranian version.

"The prince took a pair of compasses and drew a picture of an ox-headed mace in the dust. The blacksmiths set to work, and when they’d forged the heavy mace, which glittered like the sun in the heavens, they took it to Feraydun." - Shahnameh

Feraydun/Thraetona Athwya, the son of Trita Aptya above, is the first Iranian dragonslayer and is equated to the IE Sky God, and his story is the closest to Zeus'. Like Indra his right hand man, Kaveh, is a blacksmith.


"And they remembered to be grateful to him for his kindness, and gave him thunder and the glowing thunderbolt and lightning: for before that, huge Earth had hidden these." - Hes. Th. 492

Here the blacksmith cyclops forge thunderbolts for Zeus by extracting them from the earth.


"Perkūnas is the creator of the weapons (Akmeninis kalvis, "the stone smith") or he is helped by the heavenly smith Televelis (Kalvelis)."


In the first picture we see the Sky God Tarhunt depicted with an axe and in the second we see him fighting a sea serpent with a helper similar to Indra.

OFC, there is everyone's favorite Avenger, Thor and the Devil and Blacksmith story as well. In my opinion, based these identical myths which have their origin in PIE make it clear that PIE, including Anatolians, knew of metallurgy. The most accepted version of the IE tree model is Ringe-Warnow's and it dates the split of Anatolian from PIE around 4000 BCE. No steppe culture knew how to make bronze weapons before 3300BCE, which basically means that Anatolian is not that old and almost every linguist is incorrect or that Early PIE wasn't spoken in the steppes.

Vara said...


Not R1a (Z282+) but R1a1a1 and R1a-z93, which is almost non existent in modern day East Europe despite it being found in the BA.

Davidski said...


R1a-Z282 derives from R1a-M417, and there's a clear trail of these mutations from early Corded Ware to Trzciniec to modern-day Balto-Slavs. And there's no Z93 in Corded Ware.

So Corded Ware R1a isn't extinct in Eastern Europe. You're probably talking right now to a direct paternal descendant of Eastern European Corded Ware.

So what was your point again?

epoch2013 said...


"It has been explained many times how the CHG Y-DNA could've went extinct after a few centuries in the comments, so I don't think you ignored all that and I don't want to repeat it anyway. In short it probably became extinct the same way Corded Ware R1a1a1 and Karelia J1 went extinct in Eastern Europe"

Y-DNA indeed can go extinct. But I reckon chances for that to happen in a rapidly expanding population are far smaller than in a stable population.

Ric Hern said...

@ Davidski

But if the massive expansion happened during the Eneolithic then surely there had to be a Pre-Expansion formation point and period of PIE which will put most of its development in the Neolithic ? Or did PIE only form due to the Expansion ?

Expansion tells me that the they had less to do with the core...

Anthro Survey said...


You misunderstood me somewhat. :-) I don't favor a Caucasus origin for the majority of 6561's ANF ancestry. Rather, I think SE Ukraine was an intersection whereby where Balkan EEFs coming from the West encountered and hybridized with CHGs. Late Rakuschenii Yar culture perhaps?

So, I take Chad's most negative f3 with Armenia_EBA to be an expected artifact but one that says more about low extra WHG content in Yamnaya.

Nirjhar007 said...

Interesting Perspective Vara :) .

Davidski said...

Vara appears to have some serious mental problems.

There were a few papers released recently showing a clear progression from Corded Ware to modern-day Balts and Slavs in terms of genome-wide DNA, Y-DNA and mtDNA.

Vara said...


Calm your bust before it combusts. Nobody is talking about Slavs here. Where is the Indo-Iranian Y-dna R1a-Z93 in East Europe today? Based on this post it formed in Ukraine yet almost non-existent now, meaning it's extinct in that region now.

I'm basing it off this heat map:

Nirjhar007 said...


You can disagree with commentators here , its your blog, but personal insults as such is unnecessary and disappointing for the spirit IMO.

We do see a few basal Z-93 snps folks popping up in E Europe today (Perhaps also Ukraine not sure) .But we can consider that the Indian Subcontinent is still not properly sampled and of course the aDNA will be a must see .

Davidski said...

There are Z93 lineages in Poland and western Russia that can't be explained by Turkic or Roma ancestry, because they're ancestral to Turkic lines. So they're probably related to the Z93 that we're seeing in samples from the Bronze Age steppe.

It doesn't matter that they're not common in Eastern Europe today. That's just the luck of the draw. For one reason or another, Z282-derived lineages were recently more successful in Eastern Europe.

And like I said, this has nothing to do with Corded Ware anyway, since there's no Z93 in Corded Ware. But there is plenty of M417 and even some Z282, and this is where modern-day Eastern European R1a comes from.

Davidski said...


You can disagree with commentators here , its your blog, but personal insults as such is unnecessary and disappointing for the spirit IMO.

I expect people commenting here to do some background reading and at least be aware of the basics.

Claiming nonsense such as that Corded Ware R1a lineages are now extinct is unacceptable considering recent scientific literature and not much better than trolling.

Vara said...


Sintashta is derived from Corded Ware, so R1a-Z93 is to be expected no?

Since R1a makes you sensitive, how about Khvalynsk Q and Karelia J1?

In fact Folker has given a nice scenario in a recent post of how farmer's Y-DNA could have died out.

Nirjhar007 said...


Even if we take what you said as totally the case , personal attacks/insults as such is truly disappointing and unwanted . I request you to add a rule like''no personal attacks/insults '', IMO that will help conversations here more smoooth,fruitful and scientific....

Davidski said...


Sintashta is derived from Corded Ware, so R1a-Z93 is to be expected no?

It's not yet guaranteed that Sintashta is derived from Corded Ware.

Doesn't the discovery of Z93 in Eneolithic Ukraine and Bronze Age Bulgaria in Sintashta-like individuals put a spanner in that model? What if Sintashta, at least in part, derives from a population on the North Pontic steppe, and that's where its Z93 comes from?

In any case, like I said, there is some unusual Z93 in Eastern Europe, so this might derive from the Bronze Age North Pontic steppe and/or an as yet unsampled Corded Ware population.

The currently sampled Corded Ware populations, however, look good as ancestral pops to modern-day Eastern and Northern Europeans, with their shared M417 and Z282.

I can't see Maykop being responsible for the M269 or M417 in modern-day Europeans, and if I'm right, then this means it's extremely unlikely that Maykop was Indo-European, because the Indo-European expansion was one driven by highly patriarchal and patrilineal groups, which is what we're seeing in the ancient DNA record.

I'm sure that you can come up with all sorts of excuses if you like why typically Caucasus Y-haplogroups didn't take part in these massive expansions and failed to make their mark on modern-day Indo-European populations, but it's all just special pleading to me.

vacuouswastrel said...

A linguistic reminder: "PIE" is defined as the latest common ancestor of all indo-european languages.
(or minus Anatolian. There's two terminologies: "PIE" and "Late PIE", vs "PIH" (Proto-Indo-Hittite) and "PIE". Depends how deep you think the Anatolian/Non-Anatolian split is.)

So we can't say things like "maybe PIE was spoken 2000 years before the expansions", because 2000 years before the expansions it wouldn't be the latest common ancestor. Then you're talking about "PIE".

Likewise, if we say "Sredny Stog was PIE, and Yamnaya was Late PIE", that has a specific linguistic meaning: it means we think that Sredny Stog spoke a language ancestral to Anatolian, and that Yamnaya didn't. Unless we can show a clear migration from Sredny Stog to Anatolia, or at least the Balkans, before (or contemporaneous with) the formation of Yamnaya, we have no grounds to say that. And even if we did have that migration, the claim would still be tendentious. Given the linguistic lack of clarity, the genetic lack of evidence regarding Anatolian, and the archaeological lacuna (we see Anatolian cultures in Anatolia, but not how they got there), I think it's sensible to remain agnostic about the details of Anatolian's split from the rest of PIE.

vacuouswastrel said...

Sorry, meant to say "then you're talking about 'pre-PIE'".

Anyway, I'm not sure why it's clear that we can ascribe the pre-PIE language specifically to Sredny-Stog?

Given that archaeologists seem to agree that Khvalynsk was the primary source of Yamna, and that Yamna then overwhelmed Sredny Stog, would it not make more sense to ascribe pre-PIE to Khvalynsk? Particularly given the great cultural and genetic influence from the south and the east? E.g. kurgans developed in the east, probably through influence from the south, and only later spread to the west. So I'm not sure why we'd assume that the language of the kurgan-builders spread from the west to the east.

Also worth noting: "Yamnaya" is only the western end of the culture - Afanasievo, far to the east, is archeologically and genetically effectively the same, AIUI. Wouldn't it be simpler to assume a language expanding, with its culture, from a southern Volga or North Caspian location in both directions, and eventually overrunning Sredny Stog in the west?

Incidentally, and maybe this is a stupid idea, I don't know enough to say, but how would this model work as an idea for Europe:
- R1 is common on the steppe; R1b is more common further east, and R1a further west (Sredny Stog)

- the R1b culture, Yamnaya, expands in all directions. In the west, it expands THROUGH Sredny Stog (a culturally dominant or just migratory group that doesn't replace the local population) and R1a, into the Balkans

- from the Balkans, the R1b people expand up the Danube into Central Europe, where they adopt the local Bell Beaker culture (recently expanded), and with the BB culture go on to expand across all of Western Europe. The western parts of this continuum have no surviving linguistic descendents, but the central parts give us Germanic, Italic and Celtic, and probably also Albanian (and Illyrian, if that's not Albanian). In other words, the "Northwestern" languages

- meanwhile, back in what used to be Sredny Stog, the locals have adopted the culture and language of their neighbours/invaders and, being a more fertile area, have come to outnumber them - and developing chariots. They have their own migration directly into central europe, forming Corded Ware. From Corded Ware, or perhaps directly from the steppe, these R1a-dominant groups migrate with their chariots back across the steppe and central asia into India and Iran, and into Tocharia (perhaps Tocharian is directly from the steppe and not from Corded Ware?). Also into the Balkans, Greece, and Anatolia (Phrygian and maybe Armenian).

- A branch of Northwestern back-migrates along the Baltic coast, picking up the Satem sound-shift, and forming the Balto-Slavic languages (the big Slavic expansion is thousands of years later).

This would seem to line up pretty excellently with the linguistics. Is it genetically possible?

Davidski said...


Anyway, I'm not sure why it's clear that we can ascribe the pre-PIE language specifically to Sredny-Stog?

No one said here that it was clear. Certainly not me in my blog entry.

There are some key Eneolithic samples missing from the area between eastern Ukraine and Kalmykia, around the Don River, that are making it very difficult to infer what really happened on the PC steppe just before the great migrations by the ancestors of the Corded Ware people and Yamnaya (although they may have been one and the same, we don't know yet).

Ukraine_Eneolithic I6561 from Alexandra is a very interesting sample for now, because he's from a culture that has been suggested to have been PIE-speaking, he belongs to a very common Indo-European haplogroup, carries the lactase persistence (LP) allele, which is common at the two ends of the Indo-European expansion, and is largely Yamnaya-like in terms of genome-wide genetic structure, despite the fact that he precedes Yamnaya by a thousand years or so.

But he might not look so interesting after we see what a few Eneolithic individuals from just east of Alexandria, around the Don, offer to the debate. These people might be basically like Yamnaya, with both R1a-M417 and R1b-M269, as well as the LP allele. And if their horses cluster at the base of the domesticated horse phylogenetic tree, then that will be that.

Vara said...


"I'm sure that you can come up with all sorts of excuses if you like why typically Caucasus Y-haplogroups didn't take part in these massive expansions and failed to make their mark on modern-day Indo-European populations, but it's all just special pleading to me."

The same can be said about your women kidnapping theory, despite the fact that we have burial practices, art and metallurgy making its way from the Caucasus to the steppe. Also, see my comment @Nirjhar.

Anyways, we'll see when the Maykop data comes out.

Chetan said...

@Vara Non-Anatolian PIE is well established to have been spoken 4000-3000 BCE because numerous strong arguments exist for that dating. This dating is actually the least disputed in the whole of Indo-European studies, unless you are an Anatolian Origin theorist, which I think has been more or less abandoned even by its original proponent.

Chetan said...

@David This is confusing. So there is z93 found in Dereivka which dates to before ALL the m417s?

Do you think there could have been a mistake?

Vara said...


That's what I was alluding to in my reply to Nirjhar. PIE was spoken in the BA.

Alberto said...

@Anthro Survey

Ok, I see now what you meant. Well, that's another possibility, yes. It kind of implies that the Yamnaya-like admixture in Eneolithic Ukraine is an effect of a different process of admixture between different populations (what we refer to as pseudo-Yamnaya effect), and that Yamnaya proper formed in a different place, different way.

It's definitely possible, given the different people moving around. Though I guess it also suggests the level of language diversity that such a vast area with different people moving around would have had (especially before Yamnaya proper, and certainly during 4500-4000 BCE which is posited as the time of PIE language).

supernord said...

According to all the data of linguistics, archeology, culturology, anthropology, genetics, the Yamnaya culture were not PIE or Late PIE at all. It was clearly a dead-end side branch of IE which have not left descendants, though perhaps participated in the formation of the Hittites, Tocharians and part of the Greeks.

PIE was not language of the Bronze age, this language was the Copper age.

Anthro Survey said...

Here is my wild guess about a possible scenario:

1)A hypothetical population in SE Ukraine. Mostly EEFs, some CHG and maybe some Ukr_N in certain pockets.
2)Early-Mid Khvalynsk mixes with 1), producing 6561 and others similar to him.
3)Meanwhile, Khvalynsk continues to pick up some additional CHG from lower Volga or so.
4)People like 6561 in 2) then reflux back into Late Khvalynsk area and make a decisive contribution to Yamnaya's ethnogenesis(possibly introducing IE speech and husbandry).

Essentially, I envision a back-n-forth interaction zone in the western Pontic-Caspian steppe.

Al Bundy said...

@Rob What about PIE around 5000? I understand different linguists say different things, but if Anatolian diverged as early as the middle 4th century which some linguists claim, PIE would have to be born atleast a few hundred years before.

Ric Hern said...

Where does Apples or Crapapples grow naturally within the Pontic-Caspian-Steppe area ?

ryukendo kendow said...

Thanks Matt and Arza. It does seem the Yamnaya are a close proxy of three contributing population, but does not represent the actual source.

Hopefully Mathieson resequences the current set of genomes to release again with new genomes from the south of Ukraine. I can tell you conclusions drawn from broad and nonspecific componental shifts (e.g. massive migration from the east) falls on harder ground among archaeologists, but they will probably go gaga over haplotype networks, that would reveal a high degree of spatial and social structure among sites within and between cultures in this region.

Looks like we who are more interested in genetics of the issue are starting to realise that complex scenarios require high res data and a deep dive into social dynamics too.

Rob said...

@ Al
Yes o think 5000 BC is a correct figure for break up of indo Hittite , but still waiting in details of directions and paths
Hopefully future studies will bring more data from relevant places instead of 1000s of British Beakers

vacuouswastrel said...

Ric: PIE speakers did not have apples. At least, there's no clear evidence that they did.

The "apple" word is:
a) found only in Northwestern (specifically Celtic, Germanic and Balto-Slavic)
b) features *b, which probably didn't exist in PIE (it's only found in obvious onomatopoeias and in words suspected to be regional loanwords)

Thus, it's likely that the "apple" word is a regional loanword picked up somewhere in central or northern europe. The fact that it's in three branches, and displays ablaut, suggests it's a relatively early borrowing, however. Meanwhile, a different (possibly related?) word seems to have been borrowed independently into Latin and Greek. The word is also missing in Indo-Iranian.

Thus the parsimonious (though certainly non-binding) guess is that PIE originated somewhere that lacked apples, like the western steppe, and then migrated into areas that possessed apples, picking up local words for them in the process.

vacuouswastrel said...

(it's particularly interesting that "apple" appears in Celtic, but not Italic, given that Celtic and Italic look to form a genetic clade. Although it's certainly possible that Italic just lost the word randomly, this celtic/italic split further suggests that "apple" was taken into Celtic, Germanic and Balto-Slavic independently in central europe somewhere, from a local language or language family, rather than being inherited - Italic, being related to Celtic but outside the north/central european cultural sphere and in contact with different language families, would then have borrowed a different word from a mediterranean culture, in the same way that Greek did).

Ric Hern said...

@ vacuouswastrel

Thanks. So it could have been an PIE word that lost favour amongst Indo-European people who migrated into different climate zones with different flora that apples do not favour growing in...

Samuel Andrews said...

I second what David said. It's important we remember that other than Yamnaya, no early IE people from the Steppe has had their DNA thoroughly sequenced. A handful of Khvalynsk genomes isn't thorough.

So, the claim Rob has made "Steppe ancestry appears in the Balkans first" is very dishonest. There isn't an ancient DNA database in the Steppe comparable to what we have from Balkans (Bulgaria). There very well may have been people like Yamnaya living in Russia in 5500 BC.

UkraineHG genomes debunks denalists like Maju. "Steppe" ancetry can't be confused with genric ancient eastern European ancestry. Steppe might descend from a very specific PIE community who originally lived in southern Russia. Northern Russia may have been EHG, then ukraine EHG/WHG.

Rob said...

@ Sam

"So, the claim Rob has made "Steppe ancestry appears in the Balkans first"

I'm stating that steppe admixture appeared earlier in Balkans than in central & northern Europe, not steppe admixture magically originated in the middle of the Balkans. Any one with half a brain can follow that, but apparently not you. See Skolund/ Mathieson "Steppe ancestry appeared in southeastern Europe by 6000 BP (72), northeastern Europe around 5000 BP 250(47)and central Europe at the time of the Corded Ware Complex around 4600 BP'

Do yourself a favour, a don't worry about what I said. Stick to your half of the bell curve.

Matt said...

Though Ryu, though just as a comment on my earlier exercise, for what it's worth, I repeated the exercise (use PCA to extract a steppe->farmer continuum among Beaker samples, then project to obtain the "ends" of that) with the NW European Beaker samples, and then the whole CAEBA/Beaker->LBA NW European sequence.


That actually did produce a "Steppe" side that sat much closer to the Early Corded Ware samples than my earlier version (something with avoiding the Hungarian/Bavarian/Iberian/Italy samples helped do this).

The drawback though was amplification of noise in dimensions where Europeans are relatively noisy (looks like dimensions 16-17 it's less robustly able to cut through the noise without as many samples to estimate from), and that the "EEF" particularly side became exaggerated in some subtle dimensionality that distinguishes present day NW Europeans (looks like 24), I'd guess because slight NW European drifts tend to correlate with the small increases of EEF over time.

Samuel Andrews said...

"I'm stating that steppe admixture appeared earlier in Balkans than in central & northern Europe, not steppe admixture magically originated in the middle of the Balkans."

Ok, Rob this is not the first time this has happened. Several times now, you have made a claim then months later when I bring it up you deny it. For now on, either stop playing this game or be more direct. But of course, being the sneaky asshole you are, you won't do that.

I never remeber you saying the data shows Steppe people migrated into the Balkans before into northern Europe. Never. You just simply stated (paraphrase) "Oldest incident of Steppe ancestry is in the Balkans." What conclusion do you expect people to come to when you say stuff like that?

Like, I just said this is not the first time you've done this. You strongly hint towards an idea with indirect comments then deny completely deny what you were obviously hinting towards.

Samuel Andrews said...

""Steppe ancestry appeared in southeastern Europe by 6000 BP (72), northeastern Europe around 5000 BP 250(47)and central Europe at the time of the Corded Ware Complex around 4600 BP'"

Yes, but Steppe ancestry existed in the Balkans before in northern Europe. However, Steppe ancestry became persuasive in northern Europe before it did in the Balkans. The Balkans is right at the doorstep of the Steppe. Random intermarriages can explain what we see not mass migration like what happened in northern Europe. Important difference.

Samuel Andrews said...

"Do yourself a favour, a don't worry about what I said. Stick to your half of the bell curve."

Your insults don't weaken me like you intend them to. All you do is motivate me. You're bully man. Your attitude is no different than playground bullies. If I were David I'd ban you.

I won't waste my time fighting you on the internet. I'll focus my energy on fighting real-life bullies. Thank you for the encouragement.

vacuouswastrel said...

Ric: theoretically, sure, it could just be that the word was lost in Indo-Iranian and Greek and Anatolian (and probably Armenian and Albanian and Tocharian). I wouldn't say that's because people moved into an area without apples, though, because of course a lot of Indo-Iranian is/was spoken in apple-growing areas. And indeed, Indo-Iranian must have picked up apples pretty quickly, because there's a single root shared between Iranian and Indo-Aryan.

The unusual shape of 2ebol, however, also independently suggests a loanword. L-stems are pretty rare to begin with, and, as I say, *b is widely thought to have been non-existent.

When you add the fact that the word looks non-IE to the fact that it's only found in a certain geographically-contiguous set of branches, borrowing certainly seems the most parsimonious solution. Although it's certainly not certain.

Rob said...

@ Sam- don’t throw stones when you live a glass house. As I said- don’t adress me , it’s that simple.

Ric Hern said...

@ vacuouswastrel

Yes could be. This map however is interesting:

Anthro Survey said...


Let's say we ignore the presence of EEF in Yamnaya and the almost certain absence of it in Russia at the date you've mentioned:
It isn't just a question of Yamnaya-like DNA, but whether "people like Yamnaya living in [the steppes of southern] Russia in 5500 BC" left an important genetic(and cultural) legacy or were just a dead-end clade, a "false start".

I tend to think that consequential steppe ancestry was fully formed and ready to roll a couple of centuries after 4000BC, not before.

The "Steppe" admixture in the Varna outlier was most likely a precursor component to what was to become "proper steppe" or the precursor's sister clade.

Chetan said...

@vacuouswastrel The world originally could have meant just "the fruit of a tree" or some other general sense term like that. In fact that meaning is preserved even in English today.

But like you said, the Indo-Iranian geographical area (Khazakhstan and Pamir mountains) are the natural habitat of the apple tree. So I expect them to have been familiar with the apple but used a different word for it. Which again brings us to the conclusion that the apple word in PIE was some general purpose word for a fruit of some kind.

If Celtic, Germanic and Balto-Slavic have the same word with ablaut, then I think that is a strong indication that the word was in PIE. You must have heard of the Temematic substrate theory which is the theory that Balto-Slavic had a centum substrate similar to Celtic. Maybe the word entered that way.

Or it could be that the NWIE including Celtic,Germanic and BaltoSlavic is actually true and the later satem developments in BS are result of a common substrate with Indo-Iranian. I see far less evidence for the last one but it is a possibility.

Joukowski Transform said...

@Vara, since Tarhunt seems like he may be related to older Hatti Taru, it raises some questions.

Vara said...


You're probably right. Teshub/Tarhunt is most likely related to Taru. However, the root of the names Tarhunt and Illuyanka, his enemy, is clearly Indo-European, making it likely likely to have been borrowed from IE rather than Hattians. Besides, I do not see how Earth-worship leaning Hattians could have influenced all other sky-worship leaning Indo-Europeans, especially when the myth is almost identical in all versions.

The dragonslayer myths became common in the Iron Age among non-IE and even Semitic gods became dragonslayers, but the story was different.

Jijnasu said...

which word? To my knowledge the most common Indo-Aryan word for apples is a loan from persian

Chetan said...

@Jinjnasu Yes exactly. Central Indo-Aryan spoken on the Ganges plains must have lost the word and borrowed it from Persian later.. But I don't know if there was a word for it in the Vedic stage Indo-Aryan though.

Ric Hern said...

Interesting is the huge gap between the wild distribution of Malus Sylvestris and Malus species found in Mountains of Southeastern Kazakhstan who was the main contributor to the domesticated Apple.

Also interesting is that according to DNA annalisis the European Crabapple (Malus Sylvestris) is the second largest contributor to the Modern Apple Genome.

The huge gap between where these two Malus Species are found could be where Indo-Iranian lost touch with the original Proto-Indo-European word for Apple....

What is also interesting is the High concentration of the Crab Apple around the Lower Don River and the visible low concentration along the Middle Dnieper....

vacuouswastrel said...

Of course, both "it's a loanword" and "it's in PIE" are compatible: it could be a loanword into PIE, that was then lost in some branches. And it's not implausible it could have been randomly lost and replaced in Greek.

Chetan: the position of Balto-Slavic is indeed ambiguous. It could have been a 'Graeco-Aryan' language that partook in satemisation, but split off (became peripheral) by the time of the rise of the augment - perhaps we could date satem to Corded Ware, and the augment to a Sintashta-like group (if we can get away with a late date for the greek migration). And then borrowed features from Germanic, like the -m- conjugational forms. Or, it could have been a Northwestern language that strayed into contact with Graeco-Aryan and picked up some satemisation. Probably impossible to tell.

Ric Hern said...

Interesting that he had the same MtDNA Haplogroup H2a1 as the Khvalynsk R1b guy....So I think Sredny Stog 2 and Khvalynsk were somehow related even if only via some female lines.